ArticlePDF Available

Women’s weight: Findings from the Australian Longitudinal Study on Women’s Health :Report prepared for the Australian Government Department of Health and Ageing

Authors:
Womens weight:
Findings from the Australian Longitudinal
Study on Womens Health
Report prepared for the Australian Government Department of Health and Ageing
June 2007
Authors (in alphabetical order):
Lyn Adamson
Wendy Brown
Julie Byles
Catherine Chojenta
Annette Dobson
David Fitzgerald
Richard Hockey
Deborah Loxton
Jennifer Powers
Melanie Spallek
Bree Waters
Melanie Watson
ii
Table of contents
1. EXECUTIVE SUMMARY...................................................................................1
1.1. AIMS OF THIS REPORT.............................................................................................................1
1.2. SUMMARY OF MAJOR FINDINGS .............................................................................................. 2
1.2.1. Trends in women’s weight, height and body mass index ..................................................2
1.2.2. Predictors of weight change.............................................................................................3
1.2.3. Weight, weight change, and health and wellbeing............................................................4
1.2.4. Weight, weight change, and health care usage.................................................................5
1.3. DISCUSSION ...........................................................................................................................5
2. WOMEN’S WEIGHT CHANGE........................................................................7
2.1. KEY FINDINGS ........................................................................................................................7
2.2. INTRODUCTION.......................................................................................................................7
2.3. WEIGHT, HEIGHT AND BMI IN 1996 ....................................................................................... 7
2.4. CHANGES IN WEIGHT, HEIGHT AND BMI BETWEEN SURVEYS 1 SURVEY 4 ............................. 8
2.4.1. Changes in weight ............................................................................................................8
2.4.2. Changes in height.............................................................................................................9
2.4.3. Changes in BMI................................................................................................................9
2.5. CHANGES IN BMI CATEGORIES OVER TIME ..........................................................................11
2.6. WOMENS COMMENTS.......................................................................................................... 12
2.6.1. Increased weight among Younger women......................................................................12
2.6.2. Increased physical activities among Mid-aged women..................................................13
2.7. DISCUSSION .........................................................................................................................14
2.8. REFERENCES ........................................................................................................................15
3. PREDICTORS OF WEIGHT CHANGE..........................................................16
3.1. KEY FINDINGS ......................................................................................................................16
3.2. PHYSICAL ACTIVITY .............................................................................................................18
3.3. TIME SPENT SITTING .............................................................................................................23
3.3.1. Changing sitting time and weight gain...........................................................................25
3.4. ENERGY INTAKE AND DIET ...................................................................................................26
3.4.1. Women's comments.........................................................................................................29
3.5. PREGNANCY .........................................................................................................................30
3.5.1. Women's comments.........................................................................................................31
3.6. SMOKING .............................................................................................................................32
3.6.1. Women's comments.........................................................................................................34
3.7. HYSTERECTOMY ..................................................................................................................34
3.8. AREA OF RESIDENCE ............................................................................................................35
3.9. EDUCATIONAL LEVEL........................................................................................................... 37
3.10. DISCUSSION .........................................................................................................................39
3.11. REFERENCES ........................................................................................................................39
4. WEIGHT CHANGE, HEALTH AND WELL-BEING....................................41
4.1. KEY FINDINGS ......................................................................................................................41
4.2. INTRODUCTION.....................................................................................................................41
4.3. PHYSICAL AND MENTAL HEALTH BY BMI CATEGORIES FOR THE YOUNGER, MID-AGED AND
OLDER COHORTS ..................................................................................................................42
4.4. CHANGES IN PHYSICAL AND MENTAL HEALTH OVER TIME FOR WOMEN GROUPED ACCORDING
TO THEIR BMI AT SURVEY 1. ...............................................................................................42
4.5. WEIGHT CHANGE BY PHYSICAL AND MENTAL HEALTH FOR THE YOUNGER, MID-AGED AND
OLDER COHORT.................................................................................................................... 45
4.6. WEIGHT, WEIGHT GAIN AND CHRONIC DISEASE ....................................................................46
4.7. WOMEN'S COMMENTS ..........................................................................................................47
4.8. DISCUSSION .........................................................................................................................47
4.9. REFERENCES ........................................................................................................................48
5. WEIGHT, PHYSICAL ACTIVITY AND HEALTH CARE USAGE............49
5.1. KEY FINDINGS ......................................................................................................................49
iii
5.2. INTRODUCTION.....................................................................................................................50
5.2.1. Health care usage and BMI............................................................................................50
5.2.2. Total charges and BMI...................................................................................................50
5.2.3. Number of GP visits by BMI...........................................................................................51
5.2.4. Total Medicare claims by BMI category........................................................................53
5.2.5. Health care costs and weight change .............................................................................54
5.2.6. Total charges, total Medicare claims and weight change ..............................................54
5.3. NUMBER OF GP VISITS AND WEIGHT CHANGE ......................................................................55
5.4. HEALTH CARE USAGE AND PHYSICAL ACTIVITY ...................................................................56
5.4.1. Total charges, total claims and physical activity............................................................56
5.4.2. Number of GP visits and physical activity......................................................................58
5.5. CASE STUDY......................................................................................................................... 59
5.6. DISCUSSION .........................................................................................................................60
5.7. REFERENCES ........................................................................................................................61
6. APPENDICES.....................................................................................................62
APPENDIX 1: PAPERS...............................................................................................................63
APPENDIX 2: THE AUSTRALIAN LONGITUDINAL STUDY ON WOMEN’S HEALTH..
..............................................................................................................................64
A 2.1: PARTICIPATION AND RETENTION .....................................................................................65
A 2.2: CASE STUDIES ................................................................................................................. 70
APPENDIX 3: ABBREVIATIONS AND DEFINITIONS ........................................................71
A 3.1: AREA OF RESIDENCE (RRMA).........................................................................................71
A 3.2: BODY MASS INDEX (BMI) ..............................................................................................71
A 3.3: MENTAL HEALTH COMPONENT SCORE (MCS) AND PHYSICAL HEALTH COMPONENT
SCORE (PCS)................................................................................................................... 71
A 3.4: PHYSICAL ACTIVITY........................................................................................................72
A 3.5: REFERENCES....................................................................................................................72
APPENDIX 4: TABLES...............................................................................................................74
APPENDIX 5: GRAPHS..............................................................................................................77
iv
List of Figures
Figure 2-1: Average self-reported weight, height and calculated BMI with
95% confidence intervals for each cohort group (Younger, Mid-
aged, Older) for Surveys 1, 2, 3 and 4 from 1996 to 2006. ........................10
Figure 2-2: Proportions of women in each of the BMI categories based on
self-reported height and weight for each age cohort for Surveys 1,
2, 3 and 4 from 1996 to 2006......................................................................12
Figure 3-1: Median and inter quartile ranges for physical activity
(MET.mins) in each BMI category (U=underweight; H= healthy
weight; V= overweight; B= obese) for Surveys 2, 3 and 4 for the
Younger and Older cohort and for Surveys 3 and 4 for the Mid-
aged cohort..................................................................................................19
Figure 3-2: Physical activity categories for Younger women at Surveys 2, 3
and 4, by BMI category at Survey 2. ..........................................................20
Figure 3-3 Physical activity categories for Mid-aged women at Surveys 2 and
3, by BMI category at Survey 3..................................................................21
Figure 3-4: Physical activity categories for Older women at Surveys 2, 3 and
4, by BMI category at Survey 2..................................................................21
Figure 3-5: Odds ratios and 95% confidence intervals for gaining more than
5kg (N=2046), compared with maintaining weight (± 2.25kg,
N=3077), for Mid-aged women in each physical activity group;
adjusted for menopause, hysterectomy, smoking, BMI, energy
intake and sitting time.................................................................................22
Figure 3-6: Average weight with 95% confidence intervals by time spent
sitting group at Survey 2 for the Younger cohort. ......................................23
Figure 3-7: Average weight with 95% confidence intervals by time spent
sitting group at Survey 3 for the Mid-aged cohort......................................24
Figure 3-8: Average weight with 95% confidence intervals by time spent
sitting group at Survey 3 for the Older cohort. ...........................................24
Figure 3-9: Average weight with 95% confidence intervals by changes in
time spent sitting between Surveys 2 and 3 in the Younger cohort............25
Figure 3-10: Average weight with 95% confidence intervals by changes in
time spent sitting between Surveys 3 and 4 in the Younger cohort............26
Figure 3-11: Average weight with 95% confidence intervals by changes in
time spent sitting between Surveys 3 and 4 in the Mid-aged
cohort. .........................................................................................................26
Figure 3-12: Average weight with 95% confidence intervals by energy intake
for the Younger cohort at Surveys 1, 2, 3 and 4. ........................................27
Figure 3-13: Average weight with 95% confidence intervals by energy intake
for the Mid-aged cohort at Surveys 1, 2, 3 and 4........................................28
Figure 3-14: Weight change for most commonly occurring smoking patterns
for the Younger cohort from 1996 to 2006.................................................33
Figure 3-15: Weight change for most commonly occurring smoking patterns
for the Mid-aged cohort from 1996 to 2004. ..............................................34
Figure 3-16: Average BMI with 95% confidence intervals at Surveys 1, 2, 3
and 4 by hysterectomy status for the Mid-aged cohort from 1996
to 2004. .......................................................................................................35
v
Figure 3-17: Average BMI with 95% confidence intervals by area of
residence for the Younger cohort, at Surveys 1, 2, 3 and 4, from
1996 to 2006. ..............................................................................................36
Figure 3-18: Average BMI with 95% confidence intervals by area of
residence for the Mid-aged cohort, at Surveys 1, 2, 3 and 4, from
1996 to 2004. ..............................................................................................36
Figure 3-19: Average BMI with 95% confidence intervals by area of
residence for the Older cohort, at Surveys 1, 2, 3 and 4, from 1996
to 2005. .......................................................................................................37
Figure 3-20: Average weight at Surveys 1, 2, 3 and 4 by education status at
Survey 4 for the Younger cohort. ...............................................................38
Figure 3-21: Average weight at Surveys 1, 2, 3 and 4 by education status at
Survey 1 for the Mid-aged cohort...............................................................38
Figure 3-22: Average weight at Surveys 1, 2, 3 and 4 by education status at
Survey 1 for the Older cohort. ....................................................................39
Figure 4-1: Average mental health component scores (MCS) at Surveys 1, 2,
3 and 4 by BMI category at Survey 1 for the Younger cohort. ..................43
Figure 4-2: Average mental health component scores (MCS) at Surveys 1, 2,
3 and 4 by BMI category at Survey 1 for the Mid-aged cohort..................43
Figure 4-3: Average mental health component scores (MCS) at Surveys 1, 2,
3 and 4 by BMI category at Survey 1 for the Older cohort. .......................44
Figure 4-4: Average physical health component scores (PCS) at Surveys 1, 2,
3 and 4 by BMI category at Survey 1 for the Younger cohort. ..................44
Figure 4-5: Average physical health component scores (PCS) at Surveys 1, 2,
3 and 4 by BMI category at Survey 1 for the Mid-aged cohort..................45
Figure 4-6: Average physical health component scores (PCS) at Surveys 1, 2,
3 and 4 by BMI category at Survey 1 for the Older cohort. .......................45
Figure 5-1: Total Charges at Surveys 2 and 3 by BMI category at Survey 1
for Younger women. ...................................................................................51
Figure 5-2: Total Charges at Surveys 2, 3 and 4 by BMI category at Survey 1
for the Mid-aged cohort. .............................................................................51
Figure 5-3: Number of GP visits at Surveys 2 and 3 by BMI at Survey 1 for
the Younger cohort. ....................................................................................52
Figure 5-4: Number of GP visits at Surveys 2, 3 and 4 by BMI at Survey 1
for the Mid-aged cohort. .............................................................................52
Figure 5-5: Total Medicare claims by at Surveys 2 and 3 BMI at Survey 1 for
the Younger cohort. ....................................................................................53
Figure 5-6: Total Medicare claims at Surveys 2, 3 and 4 by BMI at Survey 1
for the Mid-aged cohort. .............................................................................53
Figure 5-7: Total charges at Surveys 2, 3 and 4 by weight change trajectory
for the Mid-aged cohort: S=no weight change G=weight gain
L=weight loss between two consecutive surveys. ......................................55
Figure 5-8: Total Medicare claims at Surveys 2, 3 and 4 by weight change
trajectory for the Mid-aged cohort: S=no weight change G=weight
gain L=weight loss between two consecutive surveys. ..............................55
Figure 5-9: Number of GP visits at Surveys 2, 3 and 4 by weight change
trajectory for the Mid-aged cohort: S=no weight change G=weight
gain L=weight loss between two consecutive surveys. ..............................56
Figure 5-10: Total Charges at Surveys 2 and 3 according to physical activity
group at Survey 3 for the Younger cohort. .................................................57
vi
Figure 5-11: Total Charges at Surveys 2, 3 and 4 according to physical
activity group at Survey 4 for the Mid-aged cohort....................................57
Figure 5-12: Total Medicare claims at Surveys 2, 3 and 4 in the ‘low’
physical activity group by BMI category at Survey 3 for the Mid-
aged cohort..................................................................................................58
Figure 5-13: Number of GP visits at Surveys 2 and 3 according to physical
activity group at Survey 3 for the Younger cohort. ....................................59
Figure 5-14: Number of GP visits at Surveys 2, 3 and 4 according to physical
activity group at Survey 4 for the Mid-aged cohort....................................59
Figure A1: Mean mental health component scores (MCS) and 95%
confidence intervals in the Younger cohort by BMI category for
Surveys 1, 2, 3 and 4: U=Underweight H=Healthy weight
V=Overweight B=Obese.............................................................................77
Figure A2: Mean mental health component scores (MCS) and 95%
confidence intervals in the Mid-aged cohort by BMI category for
Surveys 1, 2, 3 and 4: U=Underweight H=Healthy weight
V=Overweight B=Obese.............................................................................77
Figure A3: Mean mental health component scores (MCS) and 95%
confidence intervals in the Older cohort by BMI category for
Surveys 1, 2, 3 and 4: U=Underweight H=Healthy weight
V=Overweight B=Obese.............................................................................78
Figure A4: Mean physical health component scores (PCS) and 95%
confidence intervals in the Younger cohort by BMI category for
Surveys 1, 2, 3 and 4: U=Underweight H=Healthy weight
V=Overweight B=Obese.............................................................................78
Figure A5: Mean physical health component scores (PCS) and 95%
confidence intervals in the Mid-aged cohort by BMI category for
Surveys 1, 2, 3 and 4: U=Underweight H=Healthy weight
V=Overweight B=Obese.............................................................................79
Figure A6: Mean physical health component scores (PCS) and 95%
confidence intervals in the Older cohort by BMI category for
Surveys 1, 2, 3 and 4: U=Underweight H=Healthy weight
V=Overweight B=Obese.............................................................................79
Figure A7: Mental health component scores (MCS) for the Younger cohort
for most commonly occurring weight change patterns from 1996
to 2006: S=no weight change G=weight gain L=weight loss
between two consecutive surveys. ..............................................................80
Figure A8: Mental health component scores (MCS) for the Mid-aged cohort
for most commonly occurring weight change patterns from 1996
to 2004: S=no weight change G=weight gain L=weight loss
between two consecutive surveys. ..............................................................80
Figure A9: Mental health component scores (MCS) for the Older cohort for
most commonly occurring weight change patterns from 1996 to
2005: S=no weight change G=weight gain L=weight loss between
two consecutive surveys. ............................................................................81
Figure A10: Physical health component score (PCS) for the Younger cohort
for most commonly occurring weight change patterns from 1996
vii
to 2006: S=no weight change G=weight gain L=weight loss
between two consecutive surveys. ..............................................................82
Figure A11: Physical health component scores (PCS) for the Mid-aged
cohort for most commonly occurring weight change patterns from
1996 to 2004: S=no weight change G=weight gain L=weight loss
between two consecutive surveys. ..............................................................82
Figure A12: Physical health component scores (PCS) for the Older cohort for
most commonly occurring weight change patterns from 1996 to
2005: S=no weight change G=weight gain L=weight loss between
two consecutive surveys. ............................................................................83
Figure A13: Relationship between BMI (in intervals of 1 kg/m2) and
percentage of Younger women reporting medical problems,
surgical procedures, syptoms and GP visits at Survey 1. Data from
Younger women with BMI > 30kg/m2 are included in the BMI
category labeled 30. ....................................................................................84
Figure A14: Relationship between BMI (in intervals of 1 kg/m2) and
percentage of Mid-aged women reporting medical problems,
surgical procedures, symptoms and health care utilizations at
Survey 1. .....................................................................................................85
Figure A15: Total charges at Surveys 2 and 3 by most commonly occurring
weight change patterns for the Younger cohort from 1996 to 2006:
S=no weight change G=weight gain L=weight loss between two
consecutive surveys. ...................................................................................86
Figure A16: Total number of Medicare claims at Surveys 2 and 3 by most
commonly occurring weight change patterns for the Younger
cohort from 1996 to 2006: S=no weight change G=weight gain
L=weight loss between two consecutive surveys. ......................................86
Figure A17: Number of GP visits at Surveys 2 and 3 by most commonly
occurring weight change patterns for the Younger cohort from
1996 to 2006: S=no weight change G=weight gain L=weight loss
between two consecutive surveys. ..............................................................87
Figure A18: Total number of Medicare claims at Surveys 2 and 3 by physical
activity group at Survey 3 for the Younger cohort. ....................................87
Figure A19: Total number of Medicare claims at Surveys 2 and 3 by BMI
category at Survey 3 for Younger women in the 'no activity' group
at Survey 3. .................................................................................................88
Figure A20: Total number of Medicare claims at Surveys 2 and 3 by BMI
category at Survey 3 for Younger women in the 'low activity'
group at Survey 3........................................................................................88
Figure A21: Total number of Medicare claims at Surveys 2 and 3 by BMI
category at Survey 3 for Younger women in the 'moderate activity'
group at Survey 3........................................................................................89
Figure A22: Total number of Medicare claims at Surveys 2 and 3 by BMI
category at Survey 3 for Younger women in the 'high activity'
group at Survey 3........................................................................................89
Figure A23: Total number of Medicare claims at Surveys 2, 3 and 4 by
physical activity group at Survey 4 for the Mid-aged cohort. ....................90
viii
List of Tables
Table 1-1: Schedule of Surveys for the Australia Longitudinal Study on
Women’s Health. ..........................................................................................1
Table 2-1: Mean and 95% confidence intervals for weight, height and BMI
for the Younger, Mid-aged and Older cohort in 1996. .................................8
Table A1: Participation and retention of Younger women..........................................65
Table A2: Participation and retention of Mid-aged women. .......................................66
Table A3: Participation and retention of Older women...............................................67
Table A4: Completion of surveys by Younger women (n=14247). ............................68
Table A5: Completion of Surveys by Mid-aged women (n=13716)...........................68
Table A6: Completion of Surveys by Older women (n=12432). ................................69
Table A7: Number of women in each BMI category at Surveys 1, 2, 3 and 4
for the Younger, Mid-aged and Older cohort. ............................................74
Table A8: Number of women in each physical activity (PA) category at
Surveys 2, 3 and 4 by BMI category at Survey 2 for the Younger
cohort. .........................................................................................................74
Table A9: Number of women in each physical activity (PA) group at Surveys
3 and 4 by BMI category at Survey 3 for the Mid-aged cohort..................75
Table A10: Number of women in each physical activity (PA) group at
Surveys 2, 3 and 4 by BMI category at Survey 2 for the Older
cohort. .........................................................................................................75
1
1. Executive summary
1.1. Aims of this report
The Australian Longitudinal Study on Women’s Health (ALSWH) is a longitudinal
population-based survey funded by the Australian Government Department of Health
and Ageing. The project began in 1996 and involves three large, nationally
representative, cohorts of Australian women representing three generations:
Younger women, aged 18 to 23 years when first recruited in 1996 (n=14247)
Mid-aged women, aged 45 to 50 years in 1996 (n=13716)
Older women, aged 70 to 75 years in 1996 (n=12432) (Lee et al. 2005).
The women have now been resurveyed at least three times over the past 10 years
providing a large amount of data on women’s lifestyles and health outcomes.
Table 1-1: Schedule of Surveys for the Australia Longitudinal Study on Women’s Health.
Survey 1 Survey 2 Survey 3 Survey 4 Survey 5 Survey 6 Survey 7
Younger (1996)
18-23 yrs
(2000)
22-27 yrs
(2003)
25-30 yrs
(2006)
28-33 yrs
(2009)
31-36 yrs
(2012)
34-39 yrs
(2015)
37-42 yrs
Mid-aged (1996)
45-50 yrs
(1998)
47-52 yrs
(2001)
50-55 yrs
(2004)
53-58 yrs
(2007)
56-61 yrs
(2010)
59-64yrs
(2013)
62-67 yrs
Older (1996)
70-75 yrs
(1999)
73-78 yrs
(2002)
76-81 yrs
(2005)
79-84 yrs
(2008)
82-87 yrs
(2011)
85-90 yrs
(2014)
88-93 yrs
This report has been prepared on the basis of discussions between the ALSWH
research team and staff of the Department of Health and Ageing and presents findings
on women’s weight from four surveys of the three cohorts. The following research
questions are addressed:
What are the trends in women’s weight, height and body mass index (BMI)
among the three age groups of participants in the ALSWH over the first eleven
years of the study?
What factors are predictive of weight change?
What are the effects of weight and weight change on women’s health?
What are the effects of weight and weight change on health care usage?
The report includes summaries of published and unpublished papers, as well as
primary analyses. Case studies of individual women who have commented in their
surveys on the topic of weight change are included to illustrate the findings of this
report. All ALSWH publications relating to this report are provided as a supplement
(see Appendix 1). Additional Appendices are included to provide current information
about ALSWH data (i.e., the study design, attrition and retention rates) and some of
the definitions and measurements used in the report.
This report uses the cut-off for BMI categories (underweight, healthy weight,
overweight and obese) as defined by the WHO (see Appendix 3, Section A3.2). To
classify area of residence, the Rural, Remote and Metropolitan Areas classification
scheme (RRMA) was used (see Appendix 3, Section A3.1). As a measurement for
mental and physical health, the two summary measures mental health component
2
score (MCS) and physical health component score (PCS) were used (see Appendix 3,
Section A3.3).
1.2. Summary of major findings
1.2.1. Trends in women’s weight, height and body mass index
Over the period of the study there has been an overall increase in the women’s weight,
particularly for women in the Younger cohort. While these women had the lowest
BMI at the start of the study, they have gained an average of 6.32 kg over ten years
(between Survey 1 in 1996 and Survey 4 in 2006). Mid-aged women have gained
3.43 kg in 8 years (between Survey 1 in 1996 and Survey 4 in 2004). In contrast,
Older women have lost an average of 1.67 kg in nine years (between Survey 1 in 1996
and Survey 4 in 2005). These data are only provided for women in each cohort who
answered questions on height and weight on all four surveys. Women were more
likely to be categorised as “healthy weight” at Survey 1 than women who did not
provide data for all four surveys.
Changes for Younger women
At Survey 1, Younger women had the lowest average weight and BMI. At this time,
less than 1 in 10 of the Younger women were underweight, almost 7 in 10 were in the
healthy weight range, and 2 in 10 were classified as overweight or obese. By Survey
4, less than 4 in 100 were classified as underweight, and almost 4 in 10 were
classified as overweight or obese.
With an average weight of 62.7 kg at Survey 1, the Younger women were almost 6kg
lighter than women in the Mid-aged cohort. They were also taller, and so their BMI
was considerably lower than the Mid-aged cohort, with the Younger women having
an average BMI of 22.8 which is well within the healthy weight range. Over
subsequent surveys there have been clear increases in weight among women in the
Younger cohort, and the average BMI for these women increased to 25.03 which is
just into the overweight range. The proportion of women in the healthy weight range
was reduced from 67.7% to 56.9%, and the proportion of women in the obese range
more than doubled, from 6.0% at Survey 1 to 15.8% at Survey 4.
The rapid increase of weight in this cohort means that the BMI pattern for the
Younger women aged 28-33 years of age is fast approaching the pattern seen for the
Mid-aged cohort when they were 45-50 years of age. If this rate of weight gain
continues, the cohort of Younger women will be substantially heavier than the Mid-
aged women by the time they reach 45.
Changes for Mid-aged women
At Survey 1, Mid-aged women had the highest average weight and BMI. At Survey
1, almost 4 in 10 Mid-aged women were classified as overweight or obese. This
prevalence increased to almost 6 in 10 by Survey 4. The rate of weight gain for the
Mid-aged women was not as rapid as observed among the Younger cohort, but there
was a steady increase in the proportions of women classified as overweight or obese
3
at each Survey, and a corresponding decrease in the proportions in the healthy weight
range.
Changes for Older women
The Older women showed an average decrease in weight over the four Surveys.
However, they also showed a decrease in height (around 1.85 cm). As a consequence
of these two anthropometric changes, the average BMI for the Older women did not
change greatly across the first three Surveys, although there was a reduction in
average BMI at Survey 4. Between Surveys 1 and 4, the main changes in BMI
categories for women in the Older cohort were a slight increase in the proportion
classified as obese (increasing from 12.4% to 13.4%), a reduction in the proportion
classified as healthy weight (from 51.1% to 49.2%) and an increase in the proportion
classified as underweight (from 2.2% to 4.3%).
1.2.2. Predictors of weight change
Energy balance, the net effect of energy intake (through diet) and energy expenditure
(through physical activity), is one of the major factors affecting weight and weight
change among women in the cohorts. However, the prevalence of these factors varies
according to sociodemographic, lifestyle and other personal factors. Findings also
suggest that strategies aimed at changing eating behaviours and physical activity
should be age-group specific and be related to other lifestyle factors.
Energy intake and diet
Energy intake is one of the major predictors of weight and weight change. In the
Younger and Mid-aged cohorts, women with lowest energy intake had lowest weight
and women with highest energy intake had highest weights throughout the study
period. Among Younger women, those with greater energy intake gained more
weight from Survey 1 to Survey 4 (7.4 kg), than did women with lower energy intake
(5.7 kg). Diet and energy intake were not assessed among women in the Older cohort.
Intakes of food and nutrients varied significantly across socio-demographic groups,
with unmarried women, and women in “labouring occupations” (e.g. cleaner, factory
worker, kitchen hand) having poorer nutrition intake.
Physical activity
Throughout the study period, physical activity levels have declined among Younger
women, particularly among those women who were already overweight at Survey 1.
In contrast, physical activity levels have increased among Mid-aged women and
particularly among those who were overweight at Survey 1. Physical activity levels
have declined among Older women. Compared with women in the ‘high’ physical
activity group, Mid-aged women who reported doing less than recommended levels of
(‘moderate’) physical activity were about 1.5 times more likely to gain weight at
twice the average rate. High physical activity did not appear to carry additional
advantage when compared with moderate activity.
Sitting time is used as an indicator of physical inactivity, and is strongly associated
with weight. The difference in average weight of women who reported sitting less
4
than three hours per day and those who reported sitting for more than eight hours a
day at Survey 3 was 2.61kg among Younger women, 5.36kg among Mid-aged
women, and 6.64kg among Older women. Younger and Mid-aged women who
increased their sitting time gained most weight and those who decreased their sitting
time by more than three hours per day gained weight at the slowest rate.
Other personal and lifestyle factors associated with weight and
weight change
A number of other lifestyle factors were associated with weight and weight change:
Among Younger women, weight was associated with ever having a baby at
Survey 1, and changes in weight were associated with having a baby between
Survey 1 and subsequent surveys. Younger women who have had children
gained 2-3kg in addition to the 4kg weight gained by women who had not had
children up to Survey 3.
Younger and Mid-aged women showed greater weight gain in the period
around quitting smoking than women who did not change their smoking habits
during the study period.
Mid-aged women who had a hysterectomy before 1996 had higher BMI than
those who have not had a hysterectomy. However, there is no evidence that
having had a hysterectomy leads to increased weight gain.
Sociodemographic factors associated with weight and weight change
In the Younger and Mid-aged cohorts, women in rural areas showed higher weight
gain than women in urban areas. However, in the Older cohort there was very little
difference in BMI or weight according to area of residence.
In all three cohorts, women with a university degree had lowest weights and BMI
throughout the study period compared to women with no formal qualifications, who
had highest weights and BMI. However, there is no evidence of differences in weight
change over time between women with different qualifications.
1.2.3. Weight, weight change, and health and wellbeing
Composite scores for physical and mental health were obtained from the Short-Form
36 (SF-36) health profile, and were used to provide general health and wellbeing.
According to these scores, women in all three cohorts experienced declining physical
health over the course of the study, however, those who were underweight and
women with healthy weight had better physical health than overweight and obese
women at Survey 1. Weight loss between two surveys was associated with improving
physical health for the Mid-aged cohort. Among the Younger women, those who
gained weight between two surveys had deteriorating physical health, whereas those
with stable weight did not show a significant change in their physical health.
Mental health improved over time for the Younger and Mid-aged cohorts. In all three
cohorts, mental health for Survey 1 was poorest for women who were underweight.
Among the Mid-aged and Older women, women in the healthy weight range had
better mental health than women in the obese category. Best mental health was
5
reported by women in all three cohorts who had a stable weight throughout the study
period. Weight loss between two surveys resulted in deteriorating mental health for
the Mid-aged and Older cohort. Mid-aged women with unstable weight patterns also
reported poorer mental health. Among Younger women, weight loss between surveys
was not a common trajectory. Younger women who gained weight between surveys
had poorest mental health in this cohort.
1.2.4. Weight, weight change, and health care usage
There are clear associations between weight and health care usage (total charges,
number of Medicare claims, number of GP visits). Among Younger women, these
associations are less strong than among the other age groups, and there is no
difference in total charges (aggregate total cost in dollars incurred by each participant)
according to BMI category. Younger women in the obese group made more Medicare
claims for GP visits than Younger women in the healthy weight group. Among Mid-
aged women, women in the obese group had higher total charges, higher total
Medicare claims, and more GP visits at each Survey when compared to women in the
healthy weight range. Women in the obese group also had more Medicare claims and
more GP visits than women in the overweight group at all surveys, and had higher
total charges than overweight women at Survey 4. In 2004, total charges for women
in the obese group were around $130 higher per woman (on average) than charges for
women in the overweight group. Mid-aged women who maintained a stable weight
across all surveys tended to have lower charges, fewer GP visits and fewer total
Medicare claims.
There were also associations between physical activity and health care usage, and
there was some evidence that these associations varied according to BMI category.
Younger women in the ‘none’ physical activity group had higher total charges than
women in the ‘moderate’ and ‘high’ physical activity categories. The difference in
total charges between the ‘none’ and the ‘high’ physical activity groups was around
$200 per woman for 2003.
Total charges for the Mid-aged cohort were higher at all surveys for women in the
‘none’ physical activity group compared to women in the ‘high’ physical activity
group. Women in this group also had more GP visits and more total Medicare claims
at all surveys compared to all other physical activity groups. Mid-aged women in the
‘low’ physical activity group made more claims if they were obese compared to
women with healthy weight.
1.3. Discussion
This report emphasises the growing problem of obesity among Australian women.
The longitudinal data provided by the study show the rapid increase in weight among
Younger women. This problem is underestimated by simple cross-sectional
comparisons. Indeed, cohort differences in weight and BMI at Survey 1 would
suggest the Younger women had healthier weight profiles than the Mid-aged women.
As the Younger women age, however, their weight is increasing rapidly and their
weight profiles now resemble those of the Mid-aged cohort at the start of the study.
Unless there is a significant reduction in the rate of weight increase in this Younger
6
cohort, they will have a much higher prevalence of obesity and overweight when they
reach 45 years of age.
The report also demonstrates the relationship between overweight and obesity and
poorer mental and physical health and higher health care costs. These conditions
contribute significantly to poor health among women in Australia and there is
potential for considerable cost savings, at a population level, if trends in overweight
and obesity could be reversed.
An exploration of the factors contributing to overweight and obesity suggests that
while energy balance is important, through attention to diet and physical activity,
other contextual factors must also be taken into account.
There are also key life events that signal times when women may be more susceptible
to weight gain (such as the periods following childbirth). Women’s health promotion
may need to emphasise the particular importance of healthy eating and adequate
physical activity following these events. Quitting smoking is also another key event
when women seem to gain weight, and the case studies reveal the tensions women
feel about these competing health risks. Strategies are needed to help women quit
smoking, and receive the benefits of this healthy change, without trading the risks of
smoking for risks associated with increasing weight.
7
2. Women’s weight change
2.1. Key findings
At Survey 1, Younger women had the lowest average weight (62.7kg)
and BMI (22.8). Mid-aged women had the highest average weight
(68.6kg) and BMI (25.7).
Younger and Mid-aged women experienced weight and BMI increases
over time (Younger women: 10%, Mid-aged women 5%). At Survey 1,
9 in 100 of Younger women were underweight but this proportion
declined to less than 4 in 100 by Survey 4.
At Survey 1, 2 in 10 Younger women and more than 4 in 10 Mid-aged
women were classified as overweight or obese. This increased to almost
4 in 10 and 6 in 10 respectively by Survey 4.
Between the last two surveys, the average rate of weight gain in
Younger women (1.7kg) was almost double that of the Mid-aged women
(0.9kg).
Older women lost weight and became shorter over time, resulting in a
relatively stable average BMI (25.0-25.5).
2.2. Introduction
Women in all three cohorts have reported their height and weight in every survey since
the Study began in 1996. These data give important insights into changes in weight and
body mass index (BMI), which uses the cutoff for underweight, healthy weight,
overweight and obese as defined by the WHO (see Appendix 3, section A 3.2 for the
definition and categorization of BMI) which are associated with the onset and outcomes
of many chronic health problems.
Data are presented here as means and 95% confidence intervals (CI) for the Younger
(N= 5609), Mid-aged (N= 7507) and Older women (N= 6264) who have reported their
height and weight at all surveys. Mid-aged and Older women who have reported their
height and weight at all surveys were more likely to be categorised as 'healthy weight'
at baseline than those women who did not answer these questions at all surveys and
those who did not continue responding to the surveys.
2.3. Weight, height and BMI in 1996
In 1996, the Younger women had the lowest average weight (62.7kg) while the Mid-
aged women were heaviest (68.6kg), a difference of almost 6 kg. The average weight
of the Older women was between that of the other two cohorts (65.8kg). There were
also considerable differences in height; with the Younger women being tallest (165.8
cm) followed by the Mid-aged (163 cm) and Older (161.5 cm) women. Consequently,
average BMI was highest in the Mid-aged women (25.7), followed by the Older (25.2)
and Younger (22.8) women (Table 2-1).
8
Table 2-1: Mean and 95% confidence intervals for weight, height and BMI for the Younger, Mid-aged
and Older cohort in 1996.
Mean 95% CI
Weight 62.7 [62.4; 63.0]
Height 165.8 [165.7; 166.0]
Younger
BMI 22.8 [22.6; 22.9]
Weight 68.6 [68.3; 68.9]
Height 163.0 [162.9; 163.1]
Mid-aged
BMI 25.7 [25.6; 25.8]
Weight 65.8 [65.5; 66.1]
Height 161.5 [161.3; 161.6]
Older
BMI 25.2 [25.1; 25.4]
In 1996, the highest proportion of women in the healthy weight range was found in the
Younger cohort (69.7%, compared with 52.5% of the Mid-aged and 51.1% of the Older
women). The Younger cohort also had a much higher proportion of underweight
women (9.3%, compared with only 1.5% and 2.2% of the Mid-aged and Older women
respectively). In contrast, the proportion of women categorised as obese was highest in
the Mid-aged cohort (17.6%), followed by the Older (12.4%) and Younger (6.0%)
cohorts. More than one third of the Older cohort were overweight at Survey 1 (34.3%,
compared with 28.4% of the Mid-aged women and 15.0% of the Younger women)
(Table A7).
2.4. Changes in weight, height and BMI between
Surveys 1 Survey 4
Changes in weight, height and BMI for each cohort from 1996 to 2004 are shown in
Figure 2-1.
2.4.1. Changes in weight
Between the first and fourth surveys, there were clear increases in weight in the
Younger and Mid-aged cohorts and a decline in the Older cohort (Figure 2-1). Average
weight increased most in the Younger cohort. Over the ten year period between
Surveys 1 and 4 , the average weight of the Younger women increased by 6.32kg, and
the rate of weight gain decreased from 588g/year between Surveys 1 and 2 to 567g/year
between Surveys 3 and 4 . In contrast, the average weight of the Mid-aged women
increased by 3.43kg during the eight years from Survey 1 to Survey 4, and the rate of
weight gain decreased from 460g/year between Surveys 1 and 2 to 303g/year between
Surveys 3 and 4. Over nine years from Survey 1 to Survey 4, the average weight of the
Older women decreased by 1.67kg, and between Surveys 1 and 2 they lost weight at a
rate of 137g/year compared with 360g/year between Surveys 3 and 4.
9
2.4.2. Changes in height
The average reported height of the Young and Mid-aged women did not change over
the first four surveys (Figure 2-1). However, the average height of the Older women
decreased by 1.85 cm over the nine years between Surveys 1 and 4.
2.4.3. Changes in BMI
Changes in BMI between Surveys 1 and 4 are also shown in Figure 2-1. Between
Surveys 1 and 4, the average BMI of the Younger women increased by 2.2 (from 22.8
at Survey 1), with a steady rate of increase between Surveys 3 and 4 compared with
between Surveys 1 and 2. Average BMI changed more slowly in the Mid-aged cohort,
increasing by 1.3 (from 25.7 at Survey 1) over the eight years between Surveys 1 and 4.
In the Older women, average BMI remained largely unchanged over the first three
surveys, reflecting the decreases in both weight and height in this cohort; but it declined
between Surveys 3 and 4.
10
Figure 2-1: Average self-reported weight, height and calculated BMI with 95% confidence intervals for
each cohort group (Younger, Mid-aged, Older) for Surveys 1, 2, 3 and 4 from 1996 to 2006.
11
2.5. Changes in BMI categories over time
Changes in the proportions of women in each of the BMI categories since the beginning
of the Study are shown in Figure 2-2.
In the Younger cohort the proportion of women in the healthy weight range decreased
from 67.7% at Survey 1 to 56.9% at Survey 4. While the proportion of underweight
Younger women also decreased markedly (from 9.3% to 3.9%), the proportion in the
overweight category increased from 15.0% to 23.5%. In the same period, the
proportion of Younger women categorised as obese more than doubled, from 6.0% at
Survey 1 to 15.8 % at Survey 4. It is evident from Figure 2-2 that the pattern of BMI in
the Younger women at Survey 4 (when they were 28 – 33 years old) is approaching that
of the Mid-aged cohort at Survey 1 (who were 45 – 50 years old in 1996).
The changes in the distribution of BMI categories in the Mid-aged women show a slow
decrease in the proportion of healthy weight women at each survey. By Survey 4, only
40.3% of this cohort was in the healthy weight range, with 33.7% overweight and
24.9% obese.
There was much less change in the distribution of BMI categories in the Older cohort.
At Survey 1 just over half the Older women (51.1%) were in the healthy weight range
and 12.4% were categorised as obese. By Survey 4 the proportion of healthy weight
Older women had declined to 49.2% and the proportion who were obese had increased
to 13.4%. The proportion of overweight Older women has remained stable throughout
the first four surveys (34.3% – 33.0%) while the proportion of underweight Older
women has increased from 2.2% to 4.3% (Figure 2-2, Table A7).
12
Figure 2-2: Proportions of women in each of the BMI categories based on self-reported height and
weight for each age cohort for Surveys 1, 2, 3 and 4 from 1996 to 2006.
2.6. Women’s comments
2.6.1. Increased weight among Younger women
The weight gain among the Younger cohort over the four surveys is reflected in
comments made by Younger participant Sally1. At Survey 1 Sally had a BMI of 23,
which is within the healthy weight range. At this time her life ‘lacked routine’ and she
was stressed by the need to live away from home in order to pursue study. By Survey 2
Sally had become overweight with a BMI of 28 and chose not to write any comments
on the survey form.
At Survey 3 Sally had experienced significant weight gain and had a BMI of 32,
indicating that over a seven year period she had moved from being a healthy weight to
being obese. She wrote about being diagnosed with Polycystic Ovarian Syndrome and
had moved from a ‘stressful’ job to a less rewarding but also less stressful position.
Sally commented on her weight for the first time at this survey:
Have become overweight in the last three years. I have put on about 15kgs.
1 Pseudonyms are used in the description of qualitative data.
13
By Survey 4 Sally had experienced more weight gain, with a BMI of 33. She again
commented on her weight:
I am overweight and very unhappy about it but I recently joined a gym for the
first time ever. I am working part time in a clerical position while furthering my
studies. I could earn a lot more and have a much easier financial situation but I
don’t want the travel and excessive hours that come with that.
2.6.2. Increased physical activities among Mid-aged women
Helen, a Mid-aged participant, made free-text comments at all four surveys and
exemplifies some of the quantitative findings for the Mid-aged cohort. Helen had been
steadily gaining weight across the first three surveys, but by Survey 4 she had increased
her physical activity levels and consequently experienced a significant weight loss.
At Survey 1, with a BMI of 35, Helen wrote:
I consider that the majority of my health problems are related to my obesity. I
have always found it difficult to loose and keep weight off. I am on the true cycle
of losing weight and then gaining more. I have been to all of the usual weight
loss programs. Because of my weight I am breathless on exertion. The key to
better health for me is weight loss and manageable work hours.
At Survey 2, with a BMI of 37 Helen again wrote that her health problems were due to
obesity. By Survey 3, her BMI had risen to 38 and she was using a prescribed weight
loss product in an attempt to lose weight and was exercising sporadically. However, her
busy lifestyle prevented her from ‘exercising as much as (she) should.’
At Survey 4 Helen reported that she had started a new weight loss program 18 months
beforehand. Her BMI at Survey 4 was 34, lower than it had been at any survey time
point:
18 months ago I commenced a weight loss/fitness program exercising on a
treadmill morning and night (half an hour each) and in nine months lost 30 kgs
in weight. I stopped medication for hypertension. I maintained the weight loss
with minimal exercise for 6 months.
Despite these positive results, Helen reported that she has gained 10 kilograms in the
past six months, due to:
…loss of motivation, increased workload/intensity, extreme tiredness associated
with workload…The problem is trying to balance them all (work, family,
friendships, study) I want the lot- but my age is catching up.
This case study points to time pressure and difficulties in work-life balance as being
potential barriers to sustained behavioural change and uptake of regular physical
activity. Nevertheless, and as pointed out in this report, as Mid-aged women experience
life changes such as children leaving home, and changes in paid work, more time might
become available for women to increase their levels of physical activity.
14
2.7. Discussion
The data presented here show a trend of increasing weight in both the Younger and
Mid-aged women. This trend is most marked in the Younger cohort, who gained an
average of 567 g/year between Surveys 3 and 4.
An earlier analysis (Ball et al., 2002) of the factors associated with weight maintenance
between Surveys 1 and 2 found that fewer than half (44%) of the Younger cohort
reported their BMI at follow-up to be within 5% of their baseline BMI, while 41% had
gained weight (more than 5% of baseline) and 15% had lost weight. Weight
maintainers were more likely to be in managerial or professional occupations; to have
never married; to be currently studying; and to not be mothers. Controlling for socio-
demographic factors, weight maintainers were more likely to be in a healthy weight
range at baseline; and to report that they spent less time sitting, and consumed less
takeaway food, than women who gained weight. Weight gain was more marked among
those who were already overweight, including women living in rural areas, though
location was not independently associated with weight gain.
As more of the Younger cohort have become mothers between Surveys 2 and 4,
ongoing analyses are exploring the impact of this life event on weight gain (see Section
3). The weight gain data presented here suggest that strategies for maintenance of
healthy weight and prevention of further weight gain at this life stage are now urgently
required, if these Younger women are to avoid the early onset of weight-related chronic
health problems.
Weight gain was also notable in the Mid-aged cohort, although the decline in the rate of
weight gain between Surveys 3 and 4 is encouraging. An analysis of the factors
associated with weight gain between Surveys 1 and 3 found that on average the women
gained almost 0.5 kg per year (average 2.42 kg [2.29-2.54] in the first five years of the
Study (Brown et al., 2005)). In multivariate analyses, variables associated with energy
balance (physical activity, sitting time and energy intake), as well as quitting smoking,
menopause / hysterectomy, and baseline BMI category were significantly associated
with weight gain, but other behavioural and demographic characteristics were not.
There were independent relationships between the odds of gaining more than 5kg and
lower levels of habitual physical activity; more time spent sitting; energy intake, (but
only in women with BMI > 25 at baseline); menopause transition and hysterectomy.
Ongoing analyses are exploring the factors associated with increasing physical activity
and decreasing rate of weight gain in the Mid-aged cohort. It is hypothesised that
factors relating to the availability of time for physical activity, such as children leaving
home, changes in patterns of paid work may be associated with increasing physical
activity and decreasing rate of weight gain at this life stage. Other analyses are
exploring the associations between weight gain and health outcomes such as high blood
pressure, diabetes and back pain in this cohort.
The pattern of weight change was different in the Older cohort, with a decrease in
average weight over the first nine years of the survey. Ongoing analyses are exploring
the associations between weight, weight change and health outcomes in this cohort.
15
2.8. References
Ball K, Brown W & Crawford D. Who does not gain weight? Prevalence and predictors
of weight maintenance in young women. International Journal of Obesity, 2002 (26):
1570-1578
Brown W, Williams L, Ford J, Ball K & Dobson A. Identifying the energy gap:
magnitude and determinants of 5-year weight gain in midage women. Obesity
Research, 2005, 13(8): 1431-1441
WHO Consultation on Obesity (1999: Geneva, Switzerland) Obesity: preventing and
managing the global epidemic: report of a WHO consultation. (WHO technical report
series; 894)
16
3. Predictors of weight change
3.1. Key findings
Physical activity (PA)
At all surveys and in all three cohorts, there is a clear downward trend
in median PA values across the healthy, overweight and obese groups.
Underweight women had also having lower physical activity levels
than women of healthy weight.
Among Younger women the proportion who are adequately active
declined between Surveys 3 and 4. The greatest decline was found
among those who were overweight at Survey 1, where the percentage
of active women declined from 59.6% at Survey 3 to 51.2% at Survey
4.
In contrast PA levels increased over the same period among Mid-aged
women. The greatest increase was found among those who were
overweight at Survey 1, where the percentage of active women
increased from 45.5% at Survey 3 to 55.1% at Survey 4.
Throughout the whole study period the percentage of active women
among Older women has declined, approximately 4% between Surveys
2 and 3 and 6% between Surveys 3 and 4. By Survey 4 fewer than one
in five of the Older obese women were active.
Time spent sitting
Younger women who reported sitting for more than 8 hours a day at
Survey 4 weighed 3.40 kg more than those who reported sitting less
than or equal to 3 hours per day.
Mid-aged women who reported sitting for more than 8 hours a day at
Survey 4 weighed 5.92 kg more than those who reported sitting less
than or equal to 3 hours per day.
Older women who reported sitting for more than 8 hours a day at
Survey 3 weighed 5.64 kg more than those who reported sitting less
than 3 hours per day.
Each additional hour spent sitting at Survey 3 was associated with 227
grams more weight in the Younger cohort, and with 747 grams more
weight in the Mid-aged cohort.
Between Survey 2 and Survey 3 for the Younger cohort and between
Survey 3 and Survey 4 for the Mid-aged cohort there is a clear pattern
that women who increased their sitting time gained most weight
(Younger cohort: 0.7 kg/year, Mid-aged cohort: 0.5 kg/year) and those
who decreased their sitting time by more than 3 hours per day gained
weight at the slowest rate (Younger cohort: 0.5 kg/year, Mid-aged
cohort: 0.3 kg/year).
17
Women who sit longer are heavier, or women who are heavier sit
longer.
Energy Intake and Diet
In the Younger and Mid-aged cohorts, women with lowest energy
intake had the lowest weight and women with the highest energy
intake had the highest weights throughout the study period.
Younger women with greater energy intake gained more weight
(7.4kg) from Survey 1 to Survey 4 than women with lower energy
intake (5.7kg).
Findings suggest that strategies aimed at changing eating behaviours
should be age-group specific and take account of other life style factors
such as physical activity and cigarette smoking.
Intakes of both foods and nutrients varied significantly across socio-
demographic groups, with unmarried women, and women in
‘labouring’ occupations (e.g. cleaner, factory worker, kitchen hand)
having poorer nutrition intake.
Pregnancy
Younger women who have had children during the study period gained
2-3kg in addition to the 4kg weight gain by women who had not had
children by Survey 3.
Smoking
In the Younger and Mid-aged cohorts, women gained more weight in
the period around the time of quitting smoking than women who did
not change their smoking habits during the study period.
Hysterectomy
Mid-aged women who had a hysterectomy before 1996 had
significantly higher BMI than those who have not had a hysterectomy
throughout the study period. (mean BMI in 1996 for women with and
without a hysterectomy: 26.6 [26.3; 26.8]; 25.3 [25.2; 25.5]).
There is little evidence that having had a hysterectomy leads to
increased weight gain (1.3units increase in BMI over the study period
for women who have not had a hysterectomy and for women who had
a hysterectomy before 1996).
Area of Residence
In the Younger cohort, women in rural and remote areas show more
weight gain (+2.9BMI units) than women in urban areas (+2.4BMI
units) throughout the study period.
For the Younger and particularly the Mid-aged cohort show that initial
BMI increased with rurality. The initial BMI for rural and remote
women in the Mid-aged cohort was significantly higher than those in
18
other areas (urban: 25.4, large rural: 25.6, small rural: 25.8, other rural:
26.1).
In the Older cohort there is very little difference in BMI or weight
change for different areas of residence.
Education Level
In all three cohorts, women with a university degree had the lowest
weights and BMI throughout the study period compared to women
with no formal qualifications, who had highest weights and BMI.
However, there is no evidence for differences in weight change over
time between women with different qualifications. In the Younger
cohort, women with no formal qualifications gained 3.4kg throughout
the study period while women with a university degree gained 2.6kg.
Corresponding values for the Mid-aged cohort are 5.7kg and 5.2kg.
Older women with no formal qualifications lost 1.3kg and those with a
university degree lost 1.4kg.
3.2. Physical activity
Figure 3-1 shows medians and inter quartile ranges (IQR, which is the range between
the third and first quartiles) of physical activity (PA) within each BMI category at
each survey for all three cohorts. Physical activity is reported in MET.mins, which is
a measure of energy expenditure with 600 MET.mins being equivalent to 150 minutes
of moderate intensity physical activity per week. (for further details about the
calculation and categorisation of physical activity levels see Appendix 3, Section
A3.4)
The large IQRs are indicative of wide variation in physical activity levels among
women within the different BMI categories, but there is a clear downward trend in
median values across the healthy, overweight and obese groups at all surveys and in
all three cohorts (with underweight women also having lower physical activity levels
than women of healthy weight).
In the Younger cohort, physical activity levels were lower in 2006 for women in
every BMI category, than in the previous survey. In contrast, in the Mid-aged cohort,
women in every BMI category reported more physical activity in 2004 than in 2001.
(Data from the 1998 survey are not used for comparison in this cohort because the
question about physical activity was asked slightly differently.) In the Older cohort,
physical activity levels in women in each BMI category decreased over time at each
survey from 1999 to 2005.
600 MET.min: equivalent to 150 minutes of moderate intensity
physical activity per week
19
Figure 3-1: Median and inter quartile ranges for physical activity (MET.mins) in each BMI category
(U=underweight; H= healthy weight; V= overweight; B= obese) for Surveys 2, 3 and 4 for the
Younger and Older cohort and for Surveys 3 and 4 for the Mid-aged cohort.
20
Figure 3-2 to Figure 3-4 shows changes in physical activity over time based on BMI
at Survey 2 for the Younger and Older cohorts, and on BMI at Survey 3 for the Mid-
aged cohort. For the Younger women there was a tendency for the proportion of
women categorised as 'active' (i.e., having ‘moderate’ or ‘high’ levels of physical
activity sufficient to meet the guidelines for recommended levels of physical activity)
to increase between Surveys 2 and 3 and to decrease between Surveys 3 and 4, with
the same pattern evident in women in every BMI category (Figure 3-2, Table A8).
In contrast, in the Mid-aged cohort, the proportion of women categorised as 'active'
increased between Surveys 3 and 4, by about 7% in the underweight and healthy
weight women, and by 9-10% in the overweight and obese women. The proportion of
'active' women was higher in the healthy BMI category than in the overweight or
obese groups at both surveys (Figure 3-3, Table A9).
In the Older cohort the proportions of 'active' women were highest in the underweight
and healthy BMI categories at Survey 2, and decreased with each subsequent survey
in all BMI categories. By Survey 4 it was evident that fewer than one in five of the
Older obese women were 'active' (Figure 3-4, Table A10).
Figure 3-2: Physical activity categories for Younger women at Surveys 2, 3 and 4, by BMI category at
Survey 2.
‘active’: having ‘moderate’ or ‘high’ levels of physical activity sufficient to meet
the guidelines for recommended levels of physical activity.
21
Figure 3-3 Physical activity categories for Mid-aged women at Surveys 2 and 3, by BMI category at
Survey 3.
Figure 3-4: Physical activity categories for Older women at Surveys 2, 3 and 4, by BMI category at
Survey 2.
22
Figure 3-5 shows data from a paper by Brown et al (2005) depicting the relationship
between total physical activity over five years (determined as the sum of the physical
activity scores at Survey 1, Survey 2 and Survey 3, and categorised so that 'moderate'
activity equates with meeting current guidelines) and the odds of gaining weight at
twice the average rate over this period, in the Mid-aged cohort. After adjusting for all
other factors associated with weight gain, the data show that women who reported
doing less than recommended levels of activity were about 1.5 times more likely to
gain weight at twice the average rate than women in the 'high' activity category.
'High' activity equates with about one hour per day of moderate intensity activity.
Women who met the current activity guidelines ('moderate' physical activity) were not
more likely to gain weight at the higher rate than the women in the 'high' activity
category.
Figure 3-5: Odds ratios and 95% confidence intervals for gaining more than 5kg (N=2046),
compared with maintaining weight (± 2.25kg, N=3077), for Mid-aged women in each physical activity
group; adjusted for menopause, hysterectomy, smoking, BMI, energy intake and sitting time.
‘high’ activity: equates with about one hour per day of moderate intensity activity.
‘moderate’ activity equates to meeting the current guidelines.
The National Physical Activity Guidelines suggest that, for health benefit, all
Australians should accumulate at least 30 minutes of at least moderate intensity
physical activity on most, if not all, days of the week.
23
3.3. Time spent sitting
Indicators of physical inactivity are also known to be independently associated with
weight gain (Proper et al., 2007; Mummery et al., 2005; Brown et al, 2003).
Questions about sitting time have now been included in Surveys 2 to 4 for the
Younger women, Surveys 3 and 4 for the Mid-aged women and Survey 3 four the
Older women.
Figure 3-6 to Figure 3-8 show cross-sectional data from Survey 2 for the Younger
cohort and Survey 3 for the Mid-aged and Older cohort with mean weight on the
vertical axis (on the same scale for all graphs) and five categories of average time
spent sitting on the horizontal axis; the categories are 0 – <= 3, 3 - <=4.5, 4.5 – <=6, 6
– <=8 and > 8 hours per day. At Survey 2 for the Younger cohort there was no
difference in mean weight in relation to time spent sitting. While for the Mid-aged
and Older women there was a clear positive association between weight and time
spent sitting.
The difference in average weight of Younger women who reported sitting less than or
equal to 3 hours per day and those who reported sitting for more than 8 hours a day at
Survey 2 was only 0.19 kg (Figure 3-6); but at Survey 3 this difference was 2.61 kg;
and at Survey 4 it was 3.40 kg (Figures not shown).
Corresponding data for the Mid-aged cohort at Survey 3 were 5.36 kg (Figure 3-7)
and at Survey 4 5.92 kg (Figure not shown). For the Older women at Survey 3, the
difference in average weight of those who reported less than or equal to 3 hrs and
more than 8 hours sitting per day was 5.64 kg (Figure 3-8).
Figure 3-6: Average weight with 95% confidence intervals by time spent sitting group at Survey 2 for
the Younger cohort.
24
Figure 3-7: Average weight with 95% confidence intervals by time spent sitting group at Survey 3 for
the Mid-aged cohort.
Figure 3-8: Average weight with 95% confidence intervals by time spent sitting group at Survey 3 for
the Older cohort.
Further analysis (Figures not shown) of the relationship between time spent sitting
each day and weight found that:
each additional hour spent sitting at Survey 3 was associated with 227 grams
more weight in the Younger cohort.
each additional hour spent sitting at Survey 3 was associated with 747 grams
more weight in the Mid-aged cohort.
These data show that women who sit longer are heavier, or that women who are
heavier sit longer. The prospective data in the following section will help to establish
the direction of this relationship.
25
3.3.1. Changing sitting time and weight gain
The following graphs show changes in weight between Surveys 2 and 3 (Figure 3-9),
and between Surveys 3 and 4 (Figure 3-10), for the Younger cohort, and from Survey
3 to Survey 4 in the Mid-aged cohort (Figure 3-11) in relation to their changes in
sitting time which were categorised as: decrease of more than 3 hours, any change
between decreasing 3 and increasing 3 hours, and increase of more than 3 hours per
day. In the Older cohort the sitting time questions have only been asked in one survey,
therefore no graphs of changes in sitting time and weight can be included.
The main patterns observed are:
Between Surveys 2 and 3, Younger women:
who increased their sitting time gained most weight.
who decreased their sitting time by more than 3 hours per day gained weight
at the slowest rate.
who did not change their sitting time by more than 3 hours initially had the
lowest weight and their rate of weight gain was between those for the other
two groups.
This same pattern is also apparent for the Mid-aged women between Surveys 3 and 4.
However there were almost no differences in the rates of weight gain among the
Younger women between Surveys 3 and 4 indicating that other factors (including
pregnancy) were probably more important determinants of weight gain during this
period.
Figure 3-9: Average weight with 95% confidence intervals by changes in time spent sitting between
Surveys 2 and 3 in the Younger cohort.
26
Figure 3-10: Average weight with 95% confidence intervals by changes in time spent sitting between
Surveys 3 and 4 in the Younger cohort.
Figure 3-11: Average weight with 95% confidence intervals by changes in time spent sitting between
Surveys 3 and 4 in the Mid-aged cohort.
3.4. Energy intake and diet
The major determinant of weight change is energy balance, the net effect of energy
intake (through diet) and energy expenditure (through physical activity). Diet, and
hence energy intake, has only been assessed once in the Younger and Mid-aged
cohorts, using the Food Frequency Questionnaire of the Cancer Council of Victoria.
The assessments were made at Survey 3 in 2001 for the Mid-aged women and at
Survey 3 in 2003 for the Younger women.
To show the main effects of energy intake on weight gain we grouped the women into
three groups according to their energy intake measured at Survey 3.
27
Figure 3-12 and Figure 3-13 show the mean weight over time of Younger and Mid-
aged women in these groups. Women with the lowest intake had the lowest mean
weight for the whole period from Survey 1 to Survey 4. Those with the highest intake
weighed most at Survey 1 and at all subsequent surveys. Women in the three
intermediate groups had initial mean weight in the same rank order as their energy
intake.
For the Younger women, those with greater energy intake gained more weight
between Surveys 1 and 4 (see Figure 3-12). In contrast, weight gain from Survey 1 to
Survey 4 was fairly constant among the Mid-aged women for all categories of energy
intake (see Figure 3-13).
There are some difficulties interpreting these data, because energy intake has only
been measured once, but the data appear to suggest that energy intake is a better
predictor of weight (for example at Survey 1) than of weight change (especially for
the Mid-aged women).
Figure 3-12: Average weight with 95% confidence intervals by energy intake for the Younger
cohort at Surveys 1, 2, 3 and 4.
28
Figure 3-13: Average weight with 95% confidence intervals by energy intake for the Mid-aged
cohort at Surveys 1, 2, 3 and 4.
Brown, et al (2005) conducted an analysis of the magnitude and determinants of
weight gain over the five-year period between Surveys 1 and 3 for the Mid-aged
women. On average, the women gained almost 0.5 kg per year [average 2.42 kg [2.29
- 2.54] over five years]. In multivariate analyses, variables associated with energy
balance (physical activity, sitting time and energy intake), as well as quitting
smoking, menopause and hysterectomy, and baseline BMI category were significantly
associated with weight gain, but other behavioural and demographic characteristics
were not. After adjustment for all the other biological and behavioural variables, the
odds of gaining weight at about twice the average rate (more than 5 kg over five
years) were highest for women who quit smoking (OR = 2.94 [2.17, 3.96]). There
were also independent relationships between the odds of gaining more than 5 kg and:
lower levels of habitual physical activity; more time spent sitting; energy intake (but
only in women with BMI > 25 at baseline); menopause transition; and hysterectomy.
Average weight gain was commensurate with an energy imbalance of only about 10
kcal or 40 kJ per day, which suggests that small sustained changes in the modifiable
behavioural variables could prevent further weight gain.
Some analyses of dietary composition have also been conducted. The earliest study
involved data collected in a pilot study in 1995 (Dobson, et al, 1997). Younger and
Mid-aged women living in urban and rural areas of New South Wales completed a
very brief food frequency questionnaire. The results are listed below:
Urban women in both age groups consumed meat less frequently than women
in rural areas.
Women in the less populated rural areas were more likely to eat green and
yellow vegetables and least likely to eat dried beans. There were few other
geographic differences in food habits.
Mid-aged women consumed reduced-fat milk, fruit, vegetables, fish, biscuits
and cakes significantly more frequently than Younger women.
29
Smokers in both age groups consumed fresh fruit, vegetables and breakfast
cereals significantly less frequently than non-smokers.
Women with low levels of habitual physical activity consumed fresh fruit and
cereals less frequently than more active women.
These results shed light on the subsequent data on weight and weight gain. Urban
Mid-aged women who reported the healthiest diet have weighed less than rural
women throughout the study and the Mid-aged women have gained less weight than
the Younger women. These findings suggest that strategies aimed at changing eating
behaviours should be age-group specific and be related to other life style factors such
as physical activity and cigarette smoking.
Two major papers have been published from the dietary data from the Mid-aged
women. In the first Ball and colleagues (2004) investigated the proportion of Mid-
aged women meeting the national dietary recommendations. Only about one third of
women complied with more than half of the 13 commonly promoted dietary
guidelines. In fact only two women in the entire sample met all 13 guidelines
examined. While guidelines for meat, fish, poultry, eggs, nuts, legumes and ‘extra’
foods (e.g., ice cream, chocolate, cakes, potatoes, pizza, hamburgers and wine) were
met well, large percentages of women (68 - 88%) did not meet guidelines related to
the consumption of breads, cereal-based foods and dairy products, and intakes of total
and saturated fat and iron. Women working in lower status occupations and women
living alone or with people other than a partner and/or children were at significantly
increased risk of not meeting the guidelines. From these results the current national
guidelines appear unachievable for many women.
The second paper, by Mishra et al (2005), examined socio-demographic inequalities
in the diets of the Mid-aged women. Intakes of both foods and nutrients varied
significantly across socio-demographic groups, with unmarried women, and women
in ‘labouring’ occupations (e.g. cleaner, factory worker, kitchen hand) having poorer
nutrition intake. As well as helping to address the dearth of current data on dietary
intakes in the Australian population, these results highlighted the need for continued,
targeted public health strategies aimed at improving diet of women, particularly those
from more disadvantaged socio-economic backgrounds.
3.4.1. Women's comments
The journey from ‘I’ll die from overeating but I don’t care’ – to
weight loss
Throughout the study period, Mid-aged participant Elizabeth has been in the obese
BMI category and has written about her weight at all four surveys. At Survey 1 with a
BMI of 39, Elizabeth explains she is content with her size:
I have been overweight since I had my first child. I have 5 or 6 times gone on
diets and lost some weight but always ended up heavier than when I started. I
was diagnosed with mature onset diabetes 2 years ago but it is very mild and I
keep it well under control (gave up soft drink for the diet variety and cut down
30
on cakes, etc). I have been content with my size for some time now and feel
“skinnyness” is like winning lotto – it only happens to other people.
At Survey 2, still with a BMI of 39, weight continued to be an important issue in
Elizabeth’s life, and although she had started to exercise she reiterated that she did
‘not care’ about her weight:
I found that I didn’t feel as well when I didn’t get any exercise at all and have
started feeling much better since I started playing sport again, so I will take
up a bit of extra walking as well. I have always been a pretty active person,
but I will not be trying actively to lose weight. I know I am overweight and
have been for years but one of my mottos (rightly or wrongly) is –“I don’t
smoke and I don’t drink. I’ll probably die of overeating but I don’t care!”
At Survey 3 Elizabeth had experienced some weight loss with a BMI of 36. She still
felt as if trying to lose weight would fail and that her weight did not impede her life.
However, she acknowledged that her diabetes ‘would be even better if (she) lost
weight.’ Elizabeth repeated her statements from Survey 2, ‘I will probably die from
overeating but I don’t care.’
In 2004, eight years after completing her first survey, Elizabeth has experienced some
major life changes, and has a BMI of 35, lower than at any other survey. She was
living alone, had started a new job and commented, ‘I am happier than I have been for
years’. Elizabeth has started actively trying to lose weight:
So far I have lost 6 kg but it is very hard going as I have been this weight for
30 years. If I sit down I am quite stiff when I first stand up but soon loosen up
after walking around a bit. I walk to work (one km each way) and play sport
once a week. Diabetes is well under control.
This case study shows that although women can be aware of the negative impact of
being overweight or obese, moving to a stage where weight loss is actively
undertaken can take many years and might involve other life changes. However, diet
and exercise are key elements, as demonstrated by the quantitative data.
3.5. Pregnancy
It is well-known that many women have difficulty regaining their pre-pregnancy
weight after having a baby. In the Younger women, examination of childbearing
patterns in relation to weight gain showed the following results:
Younger women, who have not been pregnant at any time (nulliparous
women), gained the least weight, approximately 4 kg between Surveys 1 and
3.
Younger women, who had their first baby before Survey 1, were the heaviest
group (by about 3 kg) at that stage. Their subsequent weight gain has been
similar to nulliparous women.
31
Younger women, who had babies between Surveys 1 and 2 gained most
weight, during that period, which they did not subsequently lose. Their
subsequent weight gain has been similar to nulliparous women.
Younger women, who first had a baby between Surveys 2 and 3, gained most
weight, approximately 4 kg during this period. By Survey 3 they were about
the same weight as the other two groups of women who had had babies earlier.
By Survey 3, Younger women who had babies between Surveys 1 and 2 were
heaviest, with an average weight of approximately 71.5 kg; and nulliparous
women were lightest, with an average weight of approximately 67.5 kg.
Younger women who have had children during the study period gained 2 – 3
kg in addition to the 4 kg weight gained by women who had not had children
up to Survey 3.
At this stage of the study, insufficient women had had more than one pregnancy to be
able to distinguish clearly the effect of subsequent pregnancies on weight gain.
3.5.1. Women's comments
Soggy stomach, sagging breasts and stretch marks…
Daniela is a member of the Younger cohort and has contributed free-text comments at
three of the four surveys. At Survey 1 Daniela had just given birth and her BMI was
25, putting her at the high end of the healthy weight range. Nevertheless, Daniela
noted that she had gained weight compared to before her pregnancy and was generally
unhappy with her body shape:
I still haven’t lost the last of my pregnancy weight gain and the new shape i.e.
soggy stomach, sagging breasts and stretch marks everywhere are hard to get
used to.
At Survey 2 Daniela did not mention her weight but had a BMI of 27, which is
classified as overweight. At Survey 3, she had a BMI of 31, reflecting a steady weight
gain over the study period, and echoing the results of the quantitative data that show a
consistent increase in BMI among the Younger cohort. It is Survey 4 before she again
comments on her weight. With a BMI of 37 Daniela has reached the obese category
and has identified factors that are leading her to feel depressed, and a cyclic
relationship between her weight and feelings:
Any depression or frustration is largely centred around worrying if I am
successful in parenting and how the boys will grow up. Will they be well
balanced respectable people? Could I be doing things differently or better?
Most of my health concerns relate to my weight and body shape. I hate being
this way but it’s so hard to make change, bad cycle, overweight - feel bad -
eat to console self and then put on more weight.
For Daniela, weight gain started with her first pregnancy and has continued over the
following decade. Her experience exemplifies the quantitative findings that show an
increase in weight associated with childbirth.
32
3.6. Smoking
The graphs below show weight change by smoking group. The smoking groups are
the main smoking patterns found with at least 100 women included. The letters N, X
and S represent change in smoking status between two surveys. N represents never
smokers, S represents smoker and X represents ex (or former) smoker at both surveys.
Q (quitting) represents a change in smoking status from smoker to ex (or former)
smoker between two surveys. Therefore, for example SQX represents the group of
women who were smoker at Surveys 1 and 2, quit smoking between Surveys 2 and 3
and stayed ex-smoker between Surveys 3 and 4. The group denoted N-S represents all
those women who took up smoking during the study period between Surveys 2 and 3
(NSS) or between Surveys 3 and 4 (NNS). Weight change in kilograms is on the
vertical axis with means representing mean weight change between Survey 1 and
Survey 2, Survey 2 and Survey 3, and Survey 3 and Survey 4 (inclusive 95%
confidence intervals). The mean weight change points are joint to make it easier to
see changes over time in the same smoking group.
Figure 3-14 shows weight change by smoking group for the Younger women.
Women in all groups gained about two kilograms, so that almost all the lines are
above zero (which corresponds to no weight change). There was very little difference
in weight change between the women had never smoked (NNN) and the smokers
(SSS). There is some evidence that women who quit smoking gained weight during
the same period as they quit smoking; for example, the group QXX had the highest
weight gain (2.9kg) between Surveys 1 and 2, which is when they quit smoking;
similarly, the group SQX had the highest weight gain (3.9kg) between Surveys 2 and
3.
33
NNN = non-smoker between Surveys 1 and 2, 2 and 3, and 3 and 4.
SSS = smoker between Surveys 1 and 2, 2 and 3, and 3 and 4.
QXX = quit smoking between Surveys 1 and 2, continued as an ex-smoker between surveys 2 and 3, and 3 and 4.
SQX = smoker between Surveys 1 and 2, quit smoking between Surveys 2 and 3, an ex-smoker between Surveys 3 and 4.
SSQ = smoker between Surveys 1 and 2, and 2 and 3, quit smoking between Surveys 3 and 4.
N-S = non-smoker between Surveys 1 and 2, a smoker between Surveys 3 and 4. Either smoker or non-smoker between
Surveys 2 and 3.
Figure 3-14: Weight change for most commonly occurring smoking patterns for the Younger cohort
from 1996 to 2006.
Figure 3-15 shows the same graphs for the Mid-aged women. The confidence
intervals for the mean weight change for the last three smoking groups are very large
because of the small numbers of women who changed smoking status. Among
women who were never smokers (NNN) and those who were smokers (SSS) the
average weight gain was about 1 kg between each survey. Once again, weight gain
was higher in the period coinciding with quitting, i.e., between Surveys 1 and 2 in the
QXX group and between Surveys 2 and 3 in the SQX group.
As few of the Older women smoke, data for this group are not shown in this section of
the report.
34
NNN = non-smoker between Surveys 1 and 2, 2 and 3, and 3 and 4.
SSS = smoker between Surveys 1 and 2, 2 and 3, and 3 and 4.
QXX = quit smoking between Surveys 1 and 2, continued as an ex-smoker between surveys 2 and 3, and 3 and 4.
SQX = smoker between Surveys 1 and 2, quit smoking between Surveys 2 and 3, an ex-smoker between Surveys 3 and 4.
SSQ = smoker between Surveys 1 and 2, and 2 and 3, quit smoking between Surveys 3 and 4.
N-S = non-smoker between Surveys 1 and 2, a smoker between Surveys 3 and 4. Either smoker or non-smoker between
Surveys 2 and 3.
Figure 3-15: Weight change for most commonly occurring smoking patterns for the Mid-aged cohort
from 1996 to 2004.
3.6.1. Women's comments
The relationship between the cessation of smoking and weight gain was commented
on by Mid-aged participant Ebony. Ebony did not make any free-text comments at
Survey 1, when her BMI was 21.
At Survey 2, Ebony felt that most of the answers to the survey related in some way to
her stopping smoking. She attributed her weight gain and resultant feelings of
discomfort with movement, lack of self esteem and a general feeling of being unfit, to
ceasing smoking 6 months previously. Her BMI at this point was 26, just inside the
overweight category.
By Survey 3 Ebony had started smoking again and reported losing weight, reflected in
her BMI of 21. She commented:
The loss of weight was due to resuming smoking more than diet.
At Survey 4 Ebony again commented that her weight was directly related to whether
she was smoking or not. At the time of the survey, she had a BMI of 22 and was
currently smoking.
3.7. Hysterectomy
In some of the ALSWH analyses of weight change, having had a hysterectomy has
been identified as a potential risk factor for weight gain. Changes in weight for Mid-
aged women who have had and not have had a hysterectomy are illustrated in Figure
35
3-16. The vertical axis denotes BMI, with vertical lines denoting the means and 95%
confidence intervals at each survey. The solid line represents those women who have
not had a hysterectomy; the dotted line represents women who had a hysterectomy
before the beginning of the study in 1996; the other dashed line represents those
women who had a hysterectomy some time between Surveys 1 and 4.
Figure 3-16 shows clearly that increase in BMI is about the same for all three groups
of women. It also shows that those women who had a hysterectomy before the start
of the study had higher BMI at the beginning of the study than those who have not
had a hysterectomy. The implication of these results is that higher BMI is a risk
factor for having a hysterectomy, rather than that having had a hysterectomy leads to
increased weight gain.
Figure 3-16: Average BMI with 95% confidence intervals at Surveys 1, 2, 3 and 4 by hysterectomy
status for the Mid-aged cohort from 1996 to 2004.
3.8. Area of residence
The next three graphs show the average BMI at each survey for women living in
urban areas, large rural centres, small rural centres and other rural areas (including
remote areas). The definition of these areas is given in Appendix 3, Section A3.1.
The data shown are mean BMI with 95% confidence intervals, and lines joining the
means to denote changes across surveys.
Figure 3-17 and Figure 3-18 for the Younger and Mid-aged cohorts respectively show
that initial BMI increased with rurality. The initial BMI for rural and remote women
in the Mid-aged cohort was significantly higher than those in other areas (urban: 25.4,
large rural: 25.6, small rural: 25.8, other rural: 26.1). Additionally in the Younger
cohort weight gain over the four surveys was also higher among women in rural and
remote areas (+2.9BMI units) than among urban women (+2.4BMI units). Figure
3-19 for the Older women shows very little difference in BMI for different areas of
residence.
36
Figure 3-17: Average BMI with 95% confidence intervals by area of residence for the Younger
cohort, at Surveys 1, 2, 3 and 4, from 1996 to 2006.
Figure 3-18: Average BMI with 95% confidence intervals by area of residence for the Mid-aged
cohort, at Surveys 1, 2, 3 and 4, from 1996 to 2004.
37
Figure 3-19: Average BMI with 95% confidence intervals by area of residence for the Older cohort,
at Surveys 1, 2, 3 and 4, from 1996 to 2005.
3.9. Educational level
For this report level of educational attainment is used as a marker of socio-economic
status. Figure 3-20 to Figure 3-22 show weight changes over time for groups of
women defined by levels of education. The vertical axis shows weight in kilograms
and the horizontal axis shows time. Women in the Younger cohort were grouped by
their level of educational attainment by Survey 4 because at earlier surveys, especially
Survey 1 a large proportion was still studying full-time or part-time. Mid-aged and
Older women were grouped by the level of education they reported at Survey 1, as
very few have increased their level of educational attainment since then.
For the Younger cohort at Survey 1 women with no formal qualifications (shown by
the open circles) had the highest weight (and also the highest BMI, not shown here),
and those with university qualification (open squares), had the lowest weight (and
BMI). They were no differences among the other three educational groups (school
certificate only, higher school certificate only, trade/apprenticeship/certificate/
diploma) who had mean weights between those of the two extreme groups. Over the
10 years between Surveys 1 and 4 mean weight increased by similar amounts for all
five groups (i.e. the lines in the figure are approximately parallel) (see Figure 3-20).
The pattern for the Mid-aged women is essentially the same as for the Younger
women, except those with university education have gained weight more slowly over
the study period so far (see Figure 3-21).
Among the Older women, the predominant pattern is loss of weight over time,
especially between Surveys 3 and 4. At Survey 1 differences in weight between the
educational groups were similar to those in the Mid-aged women with a clear gradient
from those with no formal qualifications (who were heaviest) to those with university
level education (least weight). Over time all five groups experienced similar declines
in weight (see Figure 3-22).
38
Figure 3-20: Average weight at Surveys 1, 2, 3 and 4 by education status at Survey 4 for the Younger
cohort.
Figure 3-21: Average weight at Surveys 1, 2, 3 and 4 by education status at Survey 1 for the Mid-
aged cohort.
39
Figure 3-22: Average weight at Surveys 1, 2, 3 and 4 by education status at Survey 1 for the Older
cohort.
3.10. Discussion
The results in this Section illustrate two important features. Firstly weight and weight
gain reflect energy imbalance both in the long-term and over the 3 year periods
between surveys. The estimate of 10 kcals (40 kJ) per day as the average excess of
energy intake over expenditure provides a basis for considering, at a population level,
how the “obesity epidemic” may best be controlled. Increased energy expenditure
would involve more time devoted to physical activity to an extent that might prove
difficult for many women (Costanza et al, 2007). Reduced energy intake may be
more feasible. However for sustained weight loss or weight maintenance a
combination of lower energy intake and increased physical activity may be optimal.
The importance of time spent sitting as a contributor to energy imbalance is
particularly apparent from the ALSWH results shown here.
Secondly of the other factors examined here in relation to weight gain, the most
important one is pregnancy in the Younger women. This may warrant specific public
health strategies to assist young mothers to regain pre-pregnancy weight.
3.11. References
Ball K, Mishra GD, Thane CW & Hodge A. How well do Australian women comply
with dietary guidelines? Public Health Nutrition, 2004; 7(3): 443-452.
Brown WJ, Williams L, Ford JH, Ball K & Dobson AJ. Identifying the 'energy gap':
Magnitude and determinants of five year weight gain in Mid-aged women. Obesity
Research, 2005; 13(8); 1431-1441.
Brown WJ, Miller Y, Miller R. Sitting time and work patterns as indicators of
overweight and obesity in working Australians. International Journal of Obesity,
2003; 27: 1340-1346.
40
Constanza MC, Beer-Borst S, Morabia A. Achieving energy balance at the population
level through increases in physical activity. American Journal of Public Health, 2007;
97(3), 520-524.
Dobson A, Mishra G, Brown W & Reynolds R. Food habits of young and middle-
aged women living outside the capital cities of Australia. Australian and New Zealand
Journal of Public Health, 1997; 21(7): 711-715.
Mishra G, Ball K, Patterson A, Brown W, Hodge A & Dobson A. Socio-demographic
inequalities in the diets of Mid-aged Australian women. European Journal of Clinical
Nutrition, 2005; 59(2): 185-195.
Mummery K, Brown W, Schofield G, Eakin E, Steele R. The relationship between
occupational sitting time and overweight and obesity in Australian workers.
American Journal of Preventive Medicine, 2005; 29(2); 91-97.
Proper KI, Cerin E, Brown WJ, Owen N. Sitting time and socio-economic differences
in overweight and obesity. International Journal of Obesity, 2007; 31, 169-176.
41
4. Weight change, health and well-being
4.1. Key findings
Mental health improved over time for the Younger and Mid-aged
cohorts and tended to be poorer in those women who were underweight
at Survey 1.
Mental health for the Older cohort was poorest for women who were
underweight at Survey 1.
Best mental health was reported by women in all three cohorts who
had a stable weight throughout the study period.
Weight loss between two surveys resulted in deteriorating mental
health for the Mid-aged and Older cohort.
Women in all three cohorts experienced deteriorating physical health,
however, those who were underweight and women with healthy weight
started off with better physical health than overweight and obese
women.
Weight loss between two surveys resulted in improving physical health
for the Mid-aged cohort.
Weight gain between two surveys resulted in deteriorating physical
health for the Younger cohort.
4.2. Introduction
The ALSWH uses a well-known instrument SF-36 to provide general measures of
health and well-being. Thirty-five of these questions are used to construct an 8-scale
profile of functional health and well-being scores as well as psychometrically-based
physical and mental health summary measures. The 36th item measures health
transition (Ware JE, 2000). Higher scores represent better physical or mental health,
lower scores represent poorer physical or mental health.
For this report two composite scores, Physical Health Component Score (PCS) and
Mental Health Component Score (MCS) were used, each calculated from the 8-scale
profiles measuring respectively physical and mental health (see Appendix 3, Section
A3.3). The scores are standardised separately for each cohort (Mishra G, Schofield
MJ, 1998). This means that the results are not comparable across cohorts so they are
shown in different figures which depict trends over time and across groups defined by
BMI.
The data are summarised as means and 95% confidence intervals. The scores are
sensitive to small differences between groups. This characteristic, together with the
large sample sizes, means that differences may be statistically significant (which can
be assessed visually if the confidence intervals do not overlap) even if they are too
small to be of clinical or public health importance. Typically differences of 3-4 units
would be regarded as clinically important.
42
This sensitivity of PCS and MCS to small differences in physical and mental health
makes them suitable for assessing possibly subtle effects of aspects of women’s lives
that may impact in often non-specific aspects of health. In this report trends in PCS
and MCS are reported for women who experience stability or changes in their weight
over time, for example those who lost or gained weight during the first ten years of
the study compared with those whose weight remain stable.
It is important to emphasise that these data were only for women for whom data about
MCS, PCS and BMI were available in all four surveys, so changes in scores over time
were not due to less healthy women dropping out of the study.
4.3. Physical and mental health by BMI categories for
the Younger, Mid-aged and Older cohorts
For the Younger and Mid-aged women, there is a clear pattern at each survey:
underweight women have poorest mental health, followed by the obese women,
whereas overweight women and women with healthy weight have the best mental
health (Figure A1 and Figure A2). The pattern for the Older cohort at each survey is
different; underweight women have poorest mental health followed by women with
healthy weight while overweight and obese women scored highest on the mental
health scale (Figure A3).
In all three cohorts physical health was better for underweight women and women
with healthy weight than for overweight and obese women, with obese women having
poorest physical health at each survey. The differences in physical health between
women who were of healthy weight, those who were overweight and those who were
obese were statistically significant at all surveys except for Survey 1 and were large
enough to be clinically important (Figure A4, Figure A5 and Figure A6).
4.4. Changes in physical and mental health over time
for women grouped according to their BMI
at Survey 1.
Figure 4-1 to Figure 4-6 show changes in mean MCS and PCS over the study period
for women grouped according to their BMI at Survey 1. In the Younger cohort
(Figure 4-1) mental health improved overall, after a slight drop at Survey 2 and the
pattern was similar for all BMI categories. Mental health also improved in the Mid-
aged cohort, with lower scores for women who were underweight at Survey 1 (Figure
4-2). Among the Older cohort mental health showed evidence of deterioration,
especially at Survey 4, but heavier women remained better mental health than women
who were underweight at Survey 1 (Figure 4-3).
43
Figure 4-1: Average mental health component scores (MCS) at Surveys 1, 2, 3 and 4 by BMI category
at Survey 1 for the Younger cohort.
Figure 4-2: Average mental health component scores (MCS) at Surveys 1, 2, 3 and 4 by BMI category
at Survey 1 for the Mid-aged cohort.
44
Figure 4-3: Average mental health component scores (MCS) at Surveys 1, 2, 3 and 4 by BMI category
at Survey 1 for the Older cohort.
Physical health showed very different results to mental health. In each cohort physical
health at Survey 1 is lower in the overweight and obese group than in the healthy
group (Mean PCS for the healthy, overweight and obese group in the Younger cohort:
51.4, 49.8, 48.8; the Mid-aged cohort: 52.3, 50.5, 47.3; the Older cohort: 53.7, 51.0,
45.8). Overall, physical health deteriorated in each cohort within each BMI category,
except for underweight women in the Younger cohort where physical health remained
stable throughout the study period and Younger women with healthy weight whose
physical health remained stable for the first three surveys (Figure 4-4 to Figure 4-6).
Figure 4-4: Average physical health component scores (PCS) at Surveys 1, 2, 3 and 4 by BMI
category at Survey 1 for the Younger cohort.
45
Figure 4-5: Average physical health component scores (PCS) at Surveys 1, 2, 3 and 4 by BMI
category at Survey 1 for the Mid-aged cohort.
Figure 4-6: Average physical health component scores (PCS) at Surveys 1, 2, 3 and 4 by BMI
category at Survey 1 for the Older cohort.
4.5. Weight change by physical and mental health for
the Younger, Mid-aged and Older cohort
Trends in physical health and mental health were observed among women classified
by their weight change over the study period to date. Therefore most common
trajectories of weight changes were selected for analysis. Weight change between two
surveys was defined as average percent change in weight in kilograms per year,
compared to the weight reported at the previous survey. This change was categorised
as:
Stable (S) (gained or lost less than 2.5% per year)
46
Gain (G) (gained greater than or equal to 2.5% per year)
Lost (L) (lost greater than or equal to 2.5% per year).
Mental health among the Younger cohort was best in women who kept their weight
stable throughout the study period. The group with poorest mental health were those
who gained weight between Surveys 2 and 3 and between Surveys 3 and 4. A
trajectory which included losing at a rate of 2.5% per year of body weight in between
two surveys was not one of the most common trajectories and was therefore not
observed (Figure A7).
A large number of the Mid-aged women had a stable weight throughout the period.
As for the Younger cohort, these women reported the best mental health. In contrast
women with an unstable weight pattern reported poorer mental health. Those women
who lost weight between two surveys experienced a decline in mental health around
the time of the weight loss (Figure A8).
For the Older cohort, the best mental health was reported among those who stayed the
same weight. Weight gain between two surveys consistently lead to better mental
health, in contrast to weight loss which was more likely to result in poorer mental
health. Weight loss in Older women is often related to disease, so that the appearance
of a disease may have led to the weight loss and poorer mental health (Figure A9).
The results for physical health showed different patterns from those for mental health
except for the Younger cohort. Younger women who did not change their weight had
the best physical health and those who experienced most weight gain over the period
reported the poorest physical health. Additionally a weight gain between two surveys
resulted in a deterioration of physical health. Furthermore, women with stable weight
did not experience major changes in physical health, in contrast to all the other groups
of women who showed deterioration in physical health (Figure A10).
Physical health for the Mid-aged cohort showed opposite pattern to mental health.
Women in all groups showed poorer physical health in each subsequent survey
(whereas mental health increased). Mid-aged women who lost weight between two
surveys improved their physical health around the time of weight loss (Figure A11).
Older women showed the same pattern as the Mid-aged women with a deterioration in
physical health in all groups. Older women who had a stable weight throughout the
period had better physical health at Survey 1 but experienced similar decline in
physical health as women in the other groups (Figure A12).
4.6. Weight, weight gain and chronic disease
Overweight and obesity have long been known to be associated with a number of
chronic conditions, including cardiovascular disease, lung disease, and bone and joint
conditions. For ALSWH this is illustrated in Figure A13 for the Younger cohort
(from WJ Brown et al, 2000, (included in Appendix 1) and in Figure A14 for the Mid-
aged cohort (from WJ Brown et al, 1998 (included in Appendix 1). What is less well
known is the extent to which a weight change can affect the risk of these chronic
conditions. The Australian Longitudinal Study on Women's Health offers the
opportunity to investigate the relative importance of weight gained in early life,
47
including young adulthood, and gradual weight gain throughout middle age and how
these factors contribute to risk of chronic disease. So far we have investigated the
incidence of diabetes and our early results show that body mass index at Survey 1 was
a much stronger predictor of subsequent diagnosis of diabetes than weight change in
any of the intervening times (Mishra et al., 2007).
4.7. Women's comments
Although physical health is deteriorating in the Older cohort and weight changes did
not seem to influence this deterioration, the following case shows that positive
changes in a women’s lifestyle can contribute to weight loss even in older age.
I began doing water aerobics…I threw my walking stick away
While the quantitative results highlight the health problems associated with weight
gain, this case study offers some observations on the benefits of healthy weight loss.
Older participant Meredith had a BMI of 34 at Survey 1, placing her in the obese
category. Despite her weight she feels ‘healthy for (her) age’ and attributes ‘aches and
pains’ to normal ageing:
I am overweight and would like to weigh three stone less, but have tried over the
years without success. I am not a large eater and do not eat anything that is fattening.
At Survey 2, with a BMI of 32, Meredith continued to feel healthy for her age and
commented on her preference for alternative medical practitioners. She believed her
good diet, regular flu injection, and vitamin intake resulted in vitality and good health.
By Survey 3 Meredith had started regular physical activity and experienced an
improvement in health and wellbeing:
My health has changed dramatically since April 2001, when I began doing water
aerobics three times a week for three hours. After a very short period (three visits) I
was able to throw my walking stick away and I haven’t used it since. After this I went
on a diet and lost 20kg. I was originally 90kg.
Meredith’s BMI was then 26, placing her just inside the overweight category, she
commented:
I am so much more active and I am very healthy. I haven’t felt this well for many
years.
By Survey 4 Meredith continued to report good health and wellbeing. Unfortunately
she did not report her weight at this survey, so her BMI at this time is unknown. This
case study clearly demonstrates the health benefits of a reduction in BMI brought
about by increased physical activity and the introduction of a weight reducing diet.
4.8. Discussion
PCS and MCS provide general measures of physical and mental health, not related to
specific medical conditions.
48
A previous publication using data from the Mid-aged cohort examined weight change
between the first two surveys (Survey 1 in 1996, Survey 2 in 1998), and assessed the
relationship between weight change and physical and mental well-being. Weight gain
( 2.25 kg) was found to be negatively associated with physical well-being, whereas
both weight loss and weight gain were associated with poorer mental well-being
(Williams, Young & Brown, 2006).
Even in the Younger cohort there were large differences between mental and physical
health for obese women compared to women with healthy weight. Subscales of the
SF-36 instrument have been used previously to examine the relationships between
body mass index and well-being in young Australian women. Mean scores for
physical functioning, general health and vitality were highest for women with BMI in
the range 18.5 – 25. Additionally, women in the highest BMI category (>30) were
more likely to report hypertension, asthma, headaches, backpain, sleeping difficulties,
irregular periods, and more visits to their medical practitioner (Brown, Mishra,
Kenardy & Dobson, 2000). On the other hand, underweight women reported irregular
periods and low iron. These findings illustrate the adverse effects of overweight that
can be seen at a comparatively young age.
Further research on well-being and weight change was conducted using data from the
Younger cohort, looking at weight change and changes in life satisfaction and
aspirations between Survey 1 (1996) and Survey 2 (2000) (Ball, Crawford &
Kenardy, 2004). Obese women reported more dissatisfaction with work/career/study,
family relationships, partner relationships, and social activities. The results show that
being overweight or obese early in life can have lasting effects on a women’s life
satisfaction and their future life aspirations.
4.9. References
Ball K, Crawford D & Kenardy J. Longitudinal relationships among overweight, life
satisfaction and aspirations in young women. Obesity Research, 2004; 12(6): 1019-
1030.
Brown WJ, Dobson AJ, Mishra G. What is a healthy weight for middle aged women?
International Journal of Obesity 1998, 22. 520-528.
Brown WJ, Mishra G, Kenardy J & Dobson AJ. Relationships between body mass
index and well-being in young Australian women. International Journal of Obesity,
2000; 24(10): 1360-1368.
Mishra G, Carrigan G, Brown W, Barnett A, & Dobson A. Short-term weight change
and the incidence of diabetes in midlife: results from the Australian Longitudinal
Study on Women’s Health. Diabetes Care, In Press.
Williams L, Young A & Brown W. Weight gained in two years by a population of
Mid-aged women: How much is too much? International Journal of Obesity, 2006;
30: 1229-1233.
49
5. Weight, physical activity and health care
usage
5.1. Key findings
Total charges - the aggregate total cost in dollars incurred by each participant
Among the Younger women total charges increased in all BMI
categories, but there was no difference in total charges according to
BMI category.
Total charges increased at each survey in Mid-aged women in the
healthy, overweight and obese groups, and women in the obese group
had higher total charges at each survey compared with women in the
healthy weight group. In 2004, the mean annual cost for obese Mid-
aged women was $760 [$710, $810], whereas the mean cost for
women in the healthy weight BMI range was $636 [$610, $662].
Younger women in the ‘none’ physical activity group had significantly
higher total charges than in all other groups at Survey 2. In 2003, the
mean annual cost for Younger women in the ‘none’ physical activity
group was $621 [$524, $718]. This compares with a mean cost of
$417 [$387, $446] per woman in the ‘high’ physical activity group.
Total charges for the Mid-aged cohort were higher at all surveys for
women in the ‘none’ physical activity group compared to women in
the ‘high’ physical activity group. In 2004, the mean cost for Mid-
aged women in the ‘none’ physical activity group was $818 [$752,
$885].
GP visits
In the Younger and Mid-aged cohorts, obese women had more GP
visits than women in the healthy and overweight group.
Number of GP visits tended to be lower for Mid-aged women who
maintained stable weight between all surveys than for women who
experienced weight change.
There was a trend for Younger women in the ‘none’ physical activity
group to have more GP visits than women in the ‘high’ physical
activity group.
Mid-aged women in the ‘none’ physical activity group had
significantly more GP visits at all surveys compared to all other
physical activity groups.
Number of Medicare claims
Total number of Medicare claims increased over time but did not differ
according to BMI categories for the Younger cohort.
50
In the Mid-aged cohort, total number of Medicare claims increased
over time in the healthy, overweight and obese BMI categories and
was highest for women in the obese group, compared to women in the
healthy and overweight group. Mid-aged women in the obese group,
had a mean of 19.4 [18.2, 20.6] Medicare claims for 2004, compared
with 14.4 [14.0, 15.0] for women in the healthy weight group.
Mid-aged women in the ‘low’ physical activity group made more
Medicare claims if they were obese compared to women with healthy
weight. Women in the ‘low’ physical activity group who were also
obese had a mean of 17.8. [16.3, 19.3] Medicare claims for 2004,
whereas women in the ‘low’ physical activity group who were of
healthy weight had a mean of 13.8 [12.8, 14.9] Medicare claims for
that year.
There was a trend for Younger women in the ‘none’ physical activity
group to have more total Medicare claims then women in the ‘high’
physical activity group.
Mid-aged women in the ‘none’ physical activity group had more total
Medicare claims at all surveys compared to all other physical activity
groups.
5.2. Introduction
5.2.1. Health care usage and BMI
In the following section we compare health care usage according to women’s weight
(BMI) and change in weight. Health care usage was assessed from Medicare data for
consenting women. For the Mid-aged cohort Medicare data were available for 1998,
2001 and 2004 corresponding to Surveys 2, 3 and 4. For the Young cohort, data were
available for 2000 and 2003 corresponding to Surveys 2 and 3. From these data, we
calculated:
Total Charge - the aggregate total cost in dollars incurred by each participant
expressed as the mean and 95% confidence interval (95%CI);
Total Claims - the aggregate number of Medicare claims for each participant (mean
and 95% CI);
GP visits - the aggregate number of GP visits (claims) for each participant (mean and
95% CI).
5.2.2. Total charges and BMI
Total health care charges for Survey 2, Survey 3 and Survey 4 for women in each
BMI category (as reported at Survey 1) are shown in Figure 5-1 and Figure 5-2. For
the Younger women there were no differences in total charges according to BMI
category, but charges were higher at Survey 3 compared with Survey 2. Among Mid-
aged women, charges increased at each survey for women in the healthy, overweight
and obese groups. Women in the obese BMI category had higher total charges at each
51
survey compared with women in the healthy weight group and higher charges at
Survey 4 compared with the overweight women (Mean total charges in 2004 for
obese women: $760 [$710, $810]; and women with healthy weight: $636 [$610,
$662]). Total charges for women in the underweight group were also higher than for
the healthy weight group, but these estimates are imprecise (with wide confidence
intervals).
Figure 5-1: Total Charges at Surveys 2 and 3 by BMI category at Survey 1 for Younger women.
Figure 5-2: Total Charges at Surveys 2, 3 and 4 by BMI category at Survey 1 for the Mid-aged
cohort.
5.2.3. Number of GP visits by BMI
The numbers of GP visits according to BMI category are shown in Figure 5-3 and
Figure 5-4. Younger women in the obese group had more GP visits at both surveys
52
(around 5 visits for the year) than did women in the healthy weight group (around 3.9
visits for 2003). Mid-aged women in the obese category had more visits than women
in the healthy weight and overweight groups at all surveys. However, within each
BMI category, there was not a consistent trend for more visits at each survey.
Women in the healthy weight and overweight groups made slightly more visits at
Survey 3 than at Survey 2 (an increase of around 0.5 visits), but there was no
significant increase in visits at Survey 4.
Figure 5-3: Number of GP visits at Surveys 2 and 3 by BMI at Survey 1 for the Younger cohort.
Figure 5-4: Number of GP visits at Surveys 2, 3 and 4 by BMI at Survey 1 for the Mid-aged cohort.
53
5.2.4. Total Medicare claims by BMI category
For Younger women, total number of Medicare claims did not differ according to
BMI category (Figure 5-5). For Mid-aged women, the total number of Medicare
claims was highest at each survey for women in the obese group compared with
claims for women in the healthy weight and overweight groups. There was a steep
increase in Medicare claims for women in the obese group at Survey 4 (Figure 5-6).
Mid-aged women in the obese group, had a mean of 19.4 [18.2, 20.6] Medicare
claims for 2004. This compares with 14.5 [14.0, 15.0] for women in the healthy
weight group.
Figure 5-5: Total Medicare claims by at Surveys 2 and 3 BMI at Survey 1 for the Younger cohort.
Figure 5-6: Total Medicare claims at Surveys 2, 3 and 4 by BMI at Survey 1 for the Mid-aged
cohort.
54
5.2.5. Health care costs and weight change
The previous results were based on BMI category reported at Survey 1. However,
many women’s weight changed between surveys and so their patterns of health care
use may reflect their new weight or the change in weight. To address this possibility,
we also looked at health care costs according to patterns of weight change. Weight
change between surveys was measured as average percent change in weight in
kilograms per year, compared to the weight reported at the previous survey. This
change was categorised as:
Stable (S) (gained or lost less than 2.5% per year)
Gain (G) (gained greater than or equal to 2.5% per year)
Lost (L) (lost greater than or equal to 2.5% per year).
Over the course of four surveys women had various trajectories or patterns of weight
gain or loss. For example, a woman who gained weight between Survey 1 and Survey
2, maintained a stable weight between Survey 2 and Survey 3, and who gained more
weight between Survey 3 and Survey 4 would be designated “GSG”. Health care costs
associated with the major patterns of change were examined (patterns that included
fewer than 200 women were not used for this analysis).
5.2.6. Total charges, total Medicare claims and weight change
Among Younger women, total charges tended to increase for all weight change
groups, but there were few differences between groups on total charges, or total
Medicare claims (Figure A15 and Figure A16). Mid-aged women in the stable weight
group had lower charges (at Survey 2 and Survey 3) and fewer Medicare claims (at all
Surveys) than women in the SGS group. These women also had lower charges and
fewer Medicare claims than the LGS group (at Survey 2 and Survey 4) (Figure 5-7
and Figure 5-8).
55
Figure 5-7: Total charges at Surveys 2, 3 and 4 by weight change trajectory for the Mid-aged cohort:
S=no weight change G=weight gain L=weight loss between two consecutive surveys.
Figure 5-8: Total Medicare claims at Surveys 2, 3 and 4 by weight change trajectory for the Mid-
aged cohort: S=no weight change G=weight gain L=weight loss between two consecutive surveys.
5.3. Number of GP visits and weight change
Among Younger women, there were no significant differences in the number of GP
visits according to weight change groups (Figure A17). Mid-aged women who
maintained stable weight between all surveys also tended to have a lower number of
GP visits when compared with GSG, GSS, LGS, and SGS groups (on all Surveys).
The number of visits for these women increased from Survey 2 to Survey 3 but then
remained stable, at around 4.5 visits per year [4.4, 4.7]. Trends in GP visits for the
other weight change groups are hard to interpret due to the large confidence intervals.
56
Figure 5-9: Number of GP visits at Surveys 2, 3 and 4 by weight change trajectory for the Mid-aged
cohort: S=no weight change G=weight gain L=weight loss between two consecutive surveys.
5.4. Health care usage and physical activity
Physical activity was classified at each survey into four groups: ‘none’, ‘low’,
‘moderate’, and ‘high’ physical activity (see Appendix 3, Section A3.4 for more
details).
5.4.1. Total charges, total claims and physical activity
Among Younger women, there was a trend towards lower charges for those with
higher levels of physical activity, however the confidence intervals are wide and the
differences between the ‘low’, ‘moderate’ and ‘high’ physical activity groups are not
significant. The main difference is between the ‘none’ group (at Survey 2) and
particularly between this group and the ‘high’ physical activity group (at Survey 3).
Total charges were higher in the ‘none’ group than in all other groups at Survey 2,
and higher than the ‘moderate’ and ‘high’ physical activity groups at Survey 3. In
2003, mean charges for the ‘none’ group were $621 [$524, $718] which is around
$200 higher than the total charges for the ‘high’ physical activity group ($417 for the
year [$387, $446]. The ‘low’ physical activity group had an intermediate value of
$516 [$479, $552] (Figure 5-10). A similar pattern was seen for total Medicare
claims (Figure A18). Separate analyses were undertaken to explore the interaction
between BMI and physical activity. While the confidence intervals overlap, women
with lower levels of physical activity tended to have more total Medicare claims,
regardless of their BMI category (Figure A19 to Figure A22).
57
Figure 5-10: Total Charges at Surveys 2 and 3 according to physical activity group at Survey 3 for
the Younger cohort.
Total charges for Mid-aged women were higher at all surveys for women in the
‘none’ physical activity group compared with women in the ‘high’ physical activity
group. Women in this group also had a greater increase in charges by Survey 4, and
at this time the charges for women in this category were higher than the charges for
women in any other physical activity group (at around $818 [$752, $885] for the year)
and were almost $200 higher on average than the charges for women in the ‘high’
physical activity group. Charges did not differ between ‘low’, ‘moderate’ and ‘high’
physical activity groups, although they did increase at each survey for all groups
(Figure 5-11). Total Medicare claims followed a similar pattern (Figure A23).
Figure 5-11: Total Charges at Surveys 2, 3 and 4 according to physical activity group at Survey 4
for the Mid-aged cohort.
58
Among Mid-aged women in the ‘none’ physical activity group, there was no
difference in total claims according to BMI category (Figure A23). Women in the
‘low’ physical activity group, made more claims if they were in the obese BMI
category compared to women with healthy weight (Figure 5-12). Women in the ‘low’
physical activity group who were also obese had a mean of 17.8 [16.3, 19.3] claims
for 2004, whereas women in the ‘low’ physical activity group who were of healthy
weight had a mean of 13.8 [12.8, 14.9] claims for that year.
There were no differences between BMI categories in the other physical activity
groups.
Figure 5-12: Total Medicare claims at Surveys 2, 3 and 4 in the ‘low’ physical activity group by BMI
category at Survey 3 for the Mid-aged cohort.
5.4.2. Number of GP visits and physical activity
There was a trend for Younger women in the ‘none’ physical activity group to have
more GP visits than women in the ‘higher’ physical activity group (Figure 5-13).
However, the confidence intervals overlap and so the differences between groups
were not significant except at Survey 2 when the ‘none’ group had more GP visits
than the ‘moderate’ and ‘high’ physical activity groups.
Mid-aged women in the ‘none’ physical activity group had significantly higher
numbers of GP visits at all surveys compared to all other physical activity groups
(Figure 5-14). Women in this group had a mean of 6.4 [6.0, 6.8] visits in 2004.
59
Figure 5-13: Number of GP visits at Surveys 2 and 3 according to physical activity group at
Survey 3 for the Younger cohort.
Figure 5-14: Number of GP visits at Surveys 2, 3 and 4 according to physical activity group at Survey
4 for the Mid-aged cohort.
5.5. Case study
The quantitative data indicate that overweight and obesity are related to increased use
of health services, and by extrapolation to the costs associated with health service use.
Mid-aged participant Jenni commented on these issues. At Survey 1, Jenni had a BMI
of 36 and commented that she would like to lose and maintain a lower weight for
health reasons. Jenni wrote that a holistic approach to weight loss would be
advantageous:
60
I do not feel that the medical profession treats this (weight loss) as a serious
problem with needs for many areas of assistance, eg, psychological, dietician,
physical trainer. This problem needs to be helped by a combined treatment as
the ad hoc manner now employed leaves the patient floundering.
By Survey 2 Jenni had experienced some stressful life events and was ‘on a slow
weight loss diet’ as a result of ‘having high cholesterol and mild diabetes.’ Perhaps as
a result of the diet, her BMI fell to 35 at this time. At Survey 3, with a BMI still at 35,
Jenni had stopped most physical activities as a result of a ‘severe joint inflammation.’
Jenni’s health problems at Survey 4 caused both physical and social health problems.
Despite these difficulties, her BMI had fallen to 32. However, this still places her in
the obese category. The costs of medication, tests, and medical appointments were
impacting the family finances:
Continuing bad health leads to excessive costs for medication, tests and
medical appointments which takes almost all my small pension and means my
husband cannot retire due to having to pay normal household bills without my
income.
Jenni had ‘severe arthritis’ which meant she could no longer undertake physical
activities, which in turn made ‘managing weight a problem and a worry.’
This case study highlights a number of important issues: a lack of integration between
allied, complimentary and medical services can lead to frustration and confusion; an
increasing need for services can adversely impact financial and retirement planning;
in turn, a lack of money can limit service access; which in turn can lead to difficulty
in seeking help with weight loss. Furthermore, unresolved health issues can impede
weight loss. The inter-related nature of these issues points to the complexity that
surrounds achieving weight loss for some women.
5.6. Discussion
There are clear associations between weight and health care usage. Among Younger
women, these associations are less strong than among the other cohorts, and there is
no difference in total charges according to BMI group. Younger women in the obese
group did make more Medicare claims for GP visits than did Younger women in the
healthy weight group. Among Mid-aged women, women in the obese group had
higher total charges, higher total claims, and more GP visits at each survey when
compared to women in the healthy weight range. Women in the obese group also had
more claims and more GP visits than women in the overweight group at all surveys,
and had higher total charges than overweight women at Survey 4. In 2004, total
charges for women in the obese group were around $130 higher per woman (on
average) than charges for women in the overweight group. Mid-aged women who
maintained a stable weight across all surveys tended to have lower charges, fewer GP
visits and fewer total Medicare claims.
These findings are consistent with earlier analyses that have identified a clear
association between diabetes and health care costs. Young et al. analysed survey data
1996-1999 for the Mid-aged women (1.9% with diabetes) and the Older women
(8.1% with diabetes). The survey data were linked with Medicare claims data to
61
identify the number of general practice and specialist visits and use of glycosylated
haemoglobin (HbA1c), lipids and microalbuminuria tests. Women with diabetes
were more likely to have hypertension, heart disease and eyesight problems, have
high rates of polypharmacy (four or more medications: Mid-aged 32%, Older 64%)
and more consultations with general practitioners and specialists than women without
diabetes. Having more frequent consultations with a general practitioner was
significantly associated with having tests that are recommended for routine
monitoring of diabetes. Thus, while there was an increasing use of services by women
with diabetes, in part this was due to an increase in compliance with guidelines for the
management of diabetes.
5.7. References
Young A, Lowe J, Byles J & Patterson A. Trends in health service use for women in
Australia with diabetes. Australian and New Zealand Journal of Public Health, 2005;
29: 422-428.
62
6. Appendices
APPENDIX 1: PAPERS...............................................................................................................63
APPENDIX 2: THE AUSTRALIAN LONGITUDINAL STUDY ON WOMEN’S HEALTH..
..............................................................................................................................64
A 2.1: PARTICIPATION AND RETENTION .....................................................................................65
A 2.2: CASE STUDIES ................................................................................................................. 70
APPENDIX 3: ABBREVIATIONS AND DEFINITIONS ........................................................71
A 3.1: AREA OF RESIDENCE (RRMA).........................................................................................71
A 3.2: BODY MASS INDEX (BMI) ..............................................................................................71
A 3.3: MENTAL HEALTH COMPONENT SCORE (MCS) AND PHYSICAL HEALTH COMPONENT
SCORE (PCS)................................................................................................................... 71
A 3.4: PHYSICAL ACTIVITY........................................................................................................72
A 3.5: REFERENCES....................................................................................................................72
APPENDIX 4: TABLES...............................................................................................................74
APPENDIX 5: GRAPHS..............................................................................................................77
63
Appendix 1: Papers
A collection of published and unpublished papers on findings on women’s weight can
be found in the supplementary document.
64
Appendix 2: The Australian Longitudinal
Study on Women’s Health
The Australian Longitudinal Study on Women’s Health (ALSWH) – widely known as
Women’s Health Australia - is a longitudinal population-based survey, funded by the
Australian Government Department of Health and Ageing. The Project began in 1996
and examines the health of over 40,000 Australian women (Brown et al., 1998).
The ALSWH involves three large, nationally representative, cohorts of Australian
women representing three generations:
The Younger women, aged 18-23 when first recruited in 1996
(n=14247), are in their late 20s – early 30s, the peak years for
relationship formation, childbearing, and establishing adult health
habits (e.g. physical activity, diet) and paid and unpaid work patterns.
The Mid-aged women, initially aged 45-50 (n=13716), are
experiencing menopause, as well as changes in household structure,
family care giving, and impending retirement, which are common at
this life stage. Some are showing early signs of age-related physical
decline, while some are adopting new health behaviours in preparation
for a healthy old age.
The Older women, aged 70-75 when first recruited (n=12432), are in
their 80s and facing the physical, emotional and social challenges of
old age.
Features of the Study design include:
Women were randomly selected from the Medicare Australia database
and invited to participate in the longitudinal Study.
Women in rural and remote areas of Australia were intentionally over-
sampled to ensure adequate numbers for statistical analysis.
After Survey 1 in 1996, the three age cohorts have been surveyed
sequentially, one cohort per year, on a rolling basis starting in 1998.
The Study was designed to explore factors that influence health among women who
are broadly representative of the entire Australian population. The Study assesses:
Physical and emotional health (including well-being, major diagnoses,
symptoms)
Use of health services (GP, specialist and other visits, access,
satisfaction)
Health behaviours and risk factors (diet, exercise, smoking, alcohol,
other drugs)
Time use (including paid and unpaid work, family roles, and leisure)
65
Socio-demographic factors (location, education, employment, family
composition)
Life stages and key events (such as childbirth, divorce, widowhood).
The Project provides a valuable opportunity to examine associations over time
between aspects of women’s lives and their physical and emotional health. It provides
an evidence base to the Australian Government Department of Health and Ageing –
as well as other Australian and State/Territory Departments – for the development and
evaluation of policy and practice in many areas of service delivery that affect women.
An overview of the Study and investigators, copies of the questionnaires, and
abstracts of publications and presentations can be located on the study’s website
www.alswh.org.au
A 2.1: Participation and retention
Participation response rates to Survey 1 (1996) cannot be exactly specified as some
women selected for the sample may not have received the invitation. For example,
deaths or changes of address may not have been notified to the Health Insurance
Commission (now Medicare Australia). It is estimated that 41-42% of the Younger
women, 53-56% of the Mid-aged women and 37-40% of the Older women agreed to
participate in the longitudinal Study. Comparison with the 1996 Census showed that
the respondents were broadly representative of the general population of women of
the same age, with some over-representation of women with tertiary education and
under-representation of immigrant women of non-English speaking background.
The Project has been able to retain a very high proportion of the original participants,
particularly among the Mid-aged and Older women.
Table A1: Participation and retention of Younger women.
Survey 1 Survey 2 Survey 3 Survey 4
Age in years 18-23 22-27 25-30 28-33
Eligible at previous survey 14247 14116 13886
Ineligible
deceased between surveys 22 10 15
frailty (e.g. dementia, stroke) 3 6 4
withdrawn before mailout survey date 106 214 311
Total ineligible 131 230 330
Eligible at current survey 14116 13886 13556
Non-respondents
withdrawn from the study 124 200 157
contacted but did not return 1332 653 1390
unable to contact participant 2972 3952 3029
Total non-respondents 4428 4805 4576
Respondents
completed survey 14247 9688 9081 8980
Retention rate as % eligible 68.6% 65.4% 66.2%
Among the Younger women, 69% responded to Survey 2 in 2000, 65% to Survey 3 in
2003 and 66% have responded to Survey 4 in 2006 (Table A1). This retention
compares well with other surveys of this highly mobile age group. The major reason
for non-response among the Younger women was that the research team was unable
66
to contact the women (21% of eligible women at Survey 2, 28% at Survey 3 and 22%
at Survey 4), despite using all possible methods of maintaining contact (Lee et al.,
2000; Lee et al., 2005). Younger women in their twenties are characterised by high
levels of mobility, change of surnames on marriage, often not having telephone
listings and not being registered to vote, and making extended trips outside Australia
for work, education or recreation.
Demographic characteristics (country of birth, marital status, education, employment
and living arrangements) of the Younger respondents at Survey 1 (1996) and Survey 2
(2000) were compared with those of women of the same age in the Australian
population, using data from the 1996 and 2001 Censuses respectively. The
comparisons revealed few differences; however, there was some under-representation
of women from non-English language countries at both surveys. The disparity in
education increased between 1996 and 2001. Whereas at the 1996 Census almost 70%
of women in the Younger cohort had no post school qualifications (ALSWH and the
general population), 31% and 49% had no post school qualifications in the ALSWH
sample and the 2001 Census respectively. Some of these differences will be due to
overseas graduates returning home and Australian graduates working overseas.
ALSWH women were less likely to be employed compared with women of the same
age in the 1996 Census (52% versus 60%). When many were still students, they were
more likely to be employed than women of the same age in the 2001 Census (85%
versus 67%).
Table A2: Participation and retention of Mid-aged women.
Survey 1 Survey 2 Survey 3 Survey 4
Age in years 45-50 47-52 50-55 53-58
Eligible at previous survey 13716 13606 13309
Ineligible
deceased between surveys 50 66 88
frailty (e.g. dementia, stroke) 7 14 14
withdrawn before mailout survey date 53 217 229
Total ineligible 110 297 331
Eligible at current survey 13606 13309 12978
Non-respondents
withdrawn from the study 156 155 136
contacted but did not return 254 999 887
unable to contact participant 858 929 1052
Total non-respondents 1268 2083 2075
Respondents
completed survey 13716 12338 11226 10903
Retention rate as % eligible 90.7% 84.3% 84.0%
Retention has been much higher among the Mid-aged women; 91% responded to
Survey 2 in 1998 and 84% responded to Survey 3 in 2001 and Survey 4 in 2004
(Table A2). The major reasons for non-response among Mid-aged women were that
the research team was unable to contact the women (6%, 7% and 8% of eligible
women at Survey 2, Survey 3 and Survey 4 respectively) and non-return of
questionnaires by women who could be contacted (2%, 8% and 7% of eligible women
at the second, third and fourth surveys). Mid-aged women typically lead busy lives,
often working as well as caring for their parents and children. Our data revealed that
the women who could not be contacted were more likely to be separated, divorced or
widowed.
67
Data from the 1996 and 2001 Censuses were used to compare demographic
characteristics (country of birth, marital status, education, employment and living
arrangements) of women of the same age in the Australian population with Mid-aged
respondents at Survey 1 (1996) and Survey 3 (2001). There were few differences,
however there was some under-representation of women from non-English speaking
countries and women who were separated or divorced at both surveys.
Table A3: Participation and retention of Older women.
Survey 1 Survey 2 Survey 3 Survey 4
Age in years 70-75 73-78 76-81 79-84
Eligible at previous survey 12432 11535 10187
Ineligible
deceased between surveys 529 569 769
frailty (e.g. dementia, stroke) 106 264 381
withdrawn before mailout survey date 262 515 507
Total ineligible 897 1348 1657
Eligible at current survey 11535 10187 8530
Non-respondents
withdrawn from the study 311 385 267
contacted but did not return 481 860 592
unable to contact participant 309 295 513
Total non-respondents 1101 1540 1372
Respondents
completed survey 12432 10434 8647 7158
Retention rate as % eligible 90.5% 84.9% 83.9%
Of the Older women, 91% responded to Survey 2 in 1999, 85% to Survey 3 in 2002
and 84% to Survey 4 in 2005 (Table A3). Among Older women the major reason for
non-response was non-return of the questionnaire (4% of eligible women at Survey 2,
8% at Survey 3 and 7% at Survey 4). These and other non-respondents tended to
report poorer self-rated health at Survey 1 than respondents.
Demographic characteristics (country of birth, marital status, education and living
arrangements) of the Older respondents at Survey 1 (1996) and Survey 3 (2002) were
compared with those of women of the same age in the Australian population, using
data from the 1996 and 2001 Censuses respectively. Comparisons showed few
differences. There was some under-representation of women from non-English
speaking countries in the ALSHW sample at both surveys. The high level of missing
data in the Census made comparisons difficult for marital status and educational
qualifications.
Data are available for the Younger cohort at Survey 1 (1996), Survey 2 (2000),
Survey 3 (2003) and Survey 4 (2006), for the Mid-aged cohort at Survey 1 (1996),
Survey 2 (1998), Survey 3 (2001) and Survey 4 (2004) and for the Older cohort at
Survey 1 (1996), Survey 2 (1999), Survey 3 (2002) and Survey 4 (2005). The
following tables provide information on completion of surveys.
68
Table A4: Completion of surveys by Younger women (n=14247).
Completion of Surveys n
Respondent at Surveys 2, 3 and 4 6755
Respondent at Surveys 2 and 3, non-respondent at Survey 4 1017
Respondent at Surveys 2 and 3, deceased/ withdrawn due to frailty at Survey 4 9
Respondent at Surveys 2 and 3, other ineligible at Survey 4 9
Respondent at Survey 2, non-respondent at Survey 3, respondent at Survey 4 795
Respondent at Survey 2, non-respondent at Surveys 3 and 4 954
Respondent at Survey 2, non-respondent at Survey 3, deceased/ withdrawn due to
frailty at Survey 4
3
Respondent at Survey 2, non-respondent at Survey 3, other ineligible at Survey 4 108
Respondent at Survey 2, deceased/ withdrawn due to frailty by Survey 3 9
Respondent at Survey 2, other ineligible by Survey 3 29
Non-respondent at Survey 2, respondent at Surveys 3 and 4 886
Non-respondent at Survey 2, respondent at Survey 3, non-respondent at Survey 4 400
Non-respondent at Survey 2, respondent at Survey 3, deceased/ withdrawn due to
frailty at Survey 4
1
Non-respondent at Survey 2, respondent at Survey 3, other ineligible at Survey 4 4
Non-respondent at Surveys 2 and 3, respondent at Survey 4 544
Non-respondent at Surveys 2, 3 and 4 2205
Non-respondent at Surveys 2 and 3, deceased/ withdrawn due to frailty at Survey 4 6
Non-respondent at Surveys 2 and 3, other ineligible at Survey 4 190
Non-respondent at Survey 2, deceased/ withdrawn due to frailty at Surveys 3 and 4 7
Non-respondent at Survey 2, other ineligible at Surveys 3 and 4 185
Deceased/ withdrawn due to frailty by Survey 2 25
Other ineligible by Survey 2 106
The numbers of Younger women who completed different surveys are shown in Table
A4. Forty-seven percent completed Surveys 1, 2, 3 and 4, 19% completed three of the
four surveys and a further 14% completed two of the four surveys. Few Younger
women were deceased (n=47) and 13 were too ill to complete further surveys.
Table A5: Completion of Surveys by Mid-aged women (n=13716).
Completion of Surveys n
Respondent at Surveys 2, 3 and 4 9874
Respondent at Surveys 2 and 3, non-respondent at Survey 4 823
Respondent at Surveys 2 and 3, deceased/ withdrawn due to frailty at Survey 4 74
Respondent at Surveys 2 and 3, other ineligible at Survey 4 38
Respondent at Survey 2, non-respondent at Survey 3, respondent at Survey 4 577
Respondent at Survey 2, non-respondent at Surveys 3 and 4 701
Respondent at Survey 2, non-respondent at Survey 3, deceased/ withdrawn due to
frailty at Survey 4
15
Respondent at Survey 2, non-respondent at Survey 3, other ineligible at Survey 4 122
Respondent at Survey 2, deceased/ withdrawn due to frailty by Survey 3 63
Respondent at Survey 2, other ineligible by Survey 3 51
Non-respondent at Survey 2, respondent at Surveys 3 and 4 299
Non-respondent at Survey 2, respondent at Survey 3, non-respondent at Survey 4 107
Non-respondent at Survey 2, respondent at Survey 3, deceased/ withdrawn due to
frailty at Survey 4
5
Non-respondent at Survey 2, respondent at Survey 3, other ineligible at Survey 4 6
Non-respondent at Surveys 2 and 3, respondent at Survey 4 153
69
Non-respondent at Surveys 2, 3 and 4 444
Non-respondent at Surveys 2 and 3, deceased/ withdrawn due to frailty at Survey 4 8
Non-respondent at Surveys 2 and 3, other ineligible at Survey 4 63
Non-respondent at Survey 2, deceased/ withdrawn due to frailty at Surveys 3 and 4 17
Non-respondent at Survey 2, other ineligible at Surveys 3 and 4 166
Deceased/ withdrawn due to frailty by Survey 2 57
Other ineligible by Survey 2 53
Table A5 shows the numbers of Mid-aged women who were enrolled in Survey 1
according to their history of completing Surveys 2, 3 and 4. Seventy-two percent of
the Mid-aged women completed all four surveys. A further 12% completed three of
the four surveys and 9% completed two of the four surveys. Women were mainly
ineligible to continue in the Study due to withdrawal.
Table A6: Completion of Surveys by Older women (n=12432).
Completion of Surveys n
Respondent at Surveys 2, 3 and 4 6721
Respondent at Surveys 2 and 3, non-respondent at Survey 4 727
Respondent at Surveys 2 and 3, deceased at Survey 4 579
Respondent at Surveys 2 and 3, withdrawn due to frailty at Survey 4 274
Respondent at Surveys 2 and 3, other ineligible at Survey 4 97
Respondent at Survey 2, non-respondent at Survey 3, respondent at Survey 4 234
Respondent at Survey 2, non-respondent at Surveys 3 and 4 383
Respondent at Survey 2, non-respondent at Survey 3, deceased at Survey 4 128
Respondent at Survey 2, non-respondent at Survey 3, withdrawn due to frailty at
Survey 4
70
Respondent at Survey 2, non-respondent at Survey 3, other ineligible at Survey 4 331
Respondent at Survey 2, deceased by Survey 3 497
Respondent at Survey 2, withdrawn due to frailty by Survey 3 220
Respondent at Survey 2, other ineligible by Survey 3 173
Non-respondent at Survey 2, respondent at Surveys 3 and 4 144
Non-respondent at Survey 2, respondent at Survey 3, non-respondent at Survey 4 67
Non-respondent at Survey 2, respondent at Survey 3, deceased at Survey 4 18
Non-respondent at Survey 2, respondent at Survey 3, withdrawn due to frailty at
Survey 4
15
Non-respondent at Survey 2, respondent at Survey 3, other ineligible at Survey 4 5
Non-respondent at Surveys 2 and 3, respondent at Survey 4 59
Non-respondent at Surveys 2, 3 and 4 195
Non-respondent at Surveys 2 and 3, deceased at Survey 4 44
Non-respondent at Surveys 2 and 3, withdrawn due to frailty at Survey 4 22
Non-respondent at Surveys 2 and 3, other ineligible at Survey 4 74
Non-respondent at Survey 2, deceased at Surveys 3 and 4 72
Non-respondent at Survey 2, withdrawn due to frailty at Surveys 3 and 4 44
Non-respondent at Survey 2, other ineligible at Surveys 3 and 4 342
Deceased by Survey 2 529
Withdrawn due to frailty by Survey 2 106
Other ineligible by Survey 2 262
The numbers of Older women who completed different surveys are shown in Table
A6. Fifty-four percent of Older women completed Surveys 2, 3 and 4, 9% completed
70
three of the four surveys and a further 15% completed two of the four surveys. Not
unexpectedly in this age group, discontinuation was commonly due to death or frailty.
A 2.2: Case studies
The case studies provided throughout the current report were developed by analysing
free text comments made by participants in response the survey question asking: Have
we missed anything? If you have anything else you would like to tell us, please write
on the lines below. At each survey time point participants in the Australian
Longitudinal Study on Women’s Health are provided with this opportunity and many
choose to share specific aspects of their health and well-being. While acknowledging
that there are limitations in the use of these data, they provide the researchers with a
unique opportunity to gain further insight and understanding into specific issues.
The case studies included in the current report were chosen to illustrate the
quantitative findings. The qualitative data were searched using the terms, ‘weight’,
‘exercise’, and ‘diet’. All comments were then reviewed to determine those that
provided the greatest level of detail concerning the issues at hand. BMI data included
with the case studies were extracted from the quantitative databases.
At each Survey, women were asked about diagnosed medical conditions. For the
earlier surveys this question was framed as “Have you ever been told by a doctor that
you have (list of conditions)?” However this question was revised for later surveys to
read: “In the past three years have you been diagnosed or treated for (list of
conditions)?” The list of conditions differed for each cohort, and changed slightly
with each survey as the women aged.
The prevalence of these conditions has been calculated for all women in the Study.
The prevalence of conditions at Survey 1 is based on the number of women who
indicated they had “ever been told by a doctor” that they had had the condition (the
numerator), divided by the total number of women who provided any information
about that condition (the denominator). Women who did not provide information at
Survey 1, but who subsequently reported not having the condition could be
considered as not having the condition at Survey 1 and so were included in the
denominator for this Survey.
71
Appendix 3: Abbreviations and
Definitions
A 3.1: Area of residence (RRMA)
Throughout this report area of residence has been classified according to the Rural,
Remote and Metropolitan Areas classification scheme (RRMA, 1994). The
classification uses postcodes to produce seven categories (2 metropolitan, 3 rural and
2 remote areas) that are based primarily on population numbers and an index of
remoteness. For all area of residence analyses these categories were collapsed into
four groups: ‘urban’ including capital cities and other metropolitan centres, ‘large
rural centres’, ‘small rural centres’ and ‘other rural/remote areas’.When area of
residence is not shown, estimates (such as prevalence and incidence) have been
weighted to correct for the purposeful over-sampling of women from rural and remote
areas, so that the Study populations are representative of the Australian population of
women in these age groups.
A 3.2: Body Mass Index (BMI)
At each survey women are asked to report their height and weight. These self-
reported data are used to calculate body mass index (BMI). BMI is calculated as
weight (kg) divided by the square of height (m). It is then categorised as:
underweight < 18.5; healthy weight [18.5, 25); overweight [25, 30); obese 30).
A 3.3: Mental Health Component Score (MCS)
and Physical Health Component Score (PCS)
MCS and PCS reduce the SF-36 from an eight-scale profile to two summary measures
without substantial loss of information, measuring respectively mental and physical
health. Scoring MCS and PCS measures involves three steps. First, the eight SF-36
scales are standardized using means and standard deviations from each cohort.
Deviation Standard Population n Mean) Populatioed Scale -(Transform
ed ScaleStandardiz =
Second, they are aggregated using cohort specific weights (factor score coefficients).
Third, these scores have been normalized to have a mean of 50 and a standard
deviation of 10 using the Survey 1 data separately for each cohort (Mishra G,
Schofiled MJ. 1998).
50( 10) 36 scaleszed SF Standardiof SumWeighted ScoreComponent +
×
=
This means that the results are not comparable across cohorts so they are shown in
different figures which depict trends over time.
72
A 3.4: Physical Activity
Women in all three cohorts have answered questions about physical activity (PA) in
all surveys. At Survey 1 in 1996, there were two questions about physical activity.
They asked how many times in a normal week women engaged in vigorous exercise
(e.g., aerobics, jogging) or less vigorous exercise (e.g., walking, swimming) lasting
for 20 minutes or more (Brown et al., 2004). Responses were used to derive a PA
score based on frequency of participation in 'vigorous' (7.5 METs) and 'less vigorous'
(4 METs) physical activity lasting at least 20 minutes. [PA score = {frequency *
20mins * 4 (less vigorous) + frequency * 20mins * 7.5 (vigorous)}] MET.mins.
MET.mins are units of energy expenditure – 600 MET.mins is equivalent to 150
minutes of moderate intensity (4 METs) physical activity per week.
At Survey 2, Survey 3 and Survey 4, physical activity was assessed using questions
based on those developed for the evaluation of the National Active Australia
campaign in 1997, and for national monitoring of physical activity in Australia
(National physical activity guidelines for Australians, 1999). The questions asked
about the frequency and total duration of walking (for recreation or transport), and of
vigorous (e.g., aerobics, jogging) and moderate intensity activity (e.g., swimming,
golf) in the last week. In Survey 2 for Mid-aged women gardening was also included
as an example of moderate intensity physical activity. These items have been shown
to have acceptable reliability and validity for population measurement of physical
activity (Sallis et al., 1999; Trost et al., 2002). A PA score was derived from reported
duration of time spent in each form of physical activity during the last week [
{(walking mins * 3.5) + (moderate mins * 4.0) + (vigorous mins * 7.5)} MET.mins]
(Australia’s Health, 2004).
Women in each cohort were grouped into four physical activity categories (none: <
40; low: 40 - <600; moderate 600 - < 1200; high 1200 MET.mins/week). The
National Physical Activity Guidelines suggest that, for health benefit, all Australians
should accumulate at least 30 minutes of at least moderate intensity physical activity
on most, if not all, days of the week (National physical activity guidelines for
Australians.). The ALSWH researchers use a cut-off of 600 MET.mins per week (30
minutes X 5 sessions X 4 METs) to define whether women are accumulating
sufficient physical activity for health benefit, that is, women in the ‘moderate’ and
‘high’ categories meet the guidelines.
A 3.5: References
Australian Institute of Health and Welfare. Australia’s health 2004. Canberra:
Australian Institute of Health and Welfare, 2004
Brown, W. J., Bryson, L., Byles, J. E., Dobson, A. J., Lee, C., Mishra, G., et al.
Women's Health Australia: recruitment for a national longitudinal cohort study.
Women & Health, 1998, 28(1), 23-40.
Brown WJ, Trost SG. Life transitions and changing physical activity patterns in
young women. Am Journal of Preventive Medicine 2003; 25(2): 140-143.
73
Brown WJ, Trost SG, Baumann A, Mummery K, Owen N. Test-retest reliability of
four physical activity measures used in population surveys. J Sci Med Sport 2004;
36:1181-6.
Commonwealth Department of Health and Aged Care. National physical activity
guidelines for Australians. Canberra: Department of Health and Aged Care (1999)
Lee, C., Dobson, A., Brown, W., Adamson, L., & Goldsworthy, J. Tracking
participants: lessons from the Women's Health Australia Project. Australian & New
Zealand Journal of Public Health, 2000, 24(3), 334-336.
Lee, C., Dobson, A. J., Brown, W. J., Bryson, L., Byles, J., Warner-Smith, P., et al.
Cohort Profile: the Australian Longitudinal Study on Women's Health. Int J
Epidemiol, 2005, 34(5), 987-991.
Mishra G, Schofiled MJ. Norms for the physical and mental health component
summary scores of the SF-36 for young, middle-aged and older Australian women.
Quality of Life Research, 1998, Vol 7, pp. 215-220.
Rural, remote and metropolitan areas of classification 1991 Census edition. Canberra:
Australian Government Publishing Service; 1994.
Sallis JF, Owen N. Physical activity and behavioural medicine. Thousand Oaks CA:
Sage (1999).
Trost SG, Owen N, Baumann A, Sallis JF, Brown WB. Correlates of adults’
participation in physical activity: review and update. Med Sci Sports Exerc 2002;
34:1996-2001.
74
Appendix 4: Tables
Table A7: Number of women in each BMI category at Surveys 1, 2, 3 and 4 for the Younger, Mid-
aged and Older cohort. BMI category
Under Healthy Over Obese
N Row% N Row% N Row% N Row%
Cohort Survey
1 448 9.3 3364 69.7 726 15.0 291 6.0
2 324 6.7 3140 65.0 918 19.0 447 9.3
3 215 4.5 2958 61.3 1022 21.2 634 13.1
Younger
4 186 3.9 2747 56.9 1134 23.5 762 15.8
1 120 1.5 4111 52.5 2220 28.4 1374 17.6
2 132 1.7 3768 48.2 2449 31.3 1476 18.9
3 108 1.4 3409 43.6 2514 32.1 1794 22.9
Mid-
aged
4 87 1.1 3157 40.3 2636 33.7 1945 24.9
1 115 2.2 2614 51.1 1753 34.3 633 12.4
2 135 2.6 2525 49.4 1801 35.2 654 12.8
3 163 3.2 2461 48.1 1761 34.4 730 14.3
Older
4 222 4.3 2519 49.2 1689 33.0 685 13.4
Table A8: Number of women in each physical activity (PA) category at Surveys 2, 3 and 4 by BMI
category at Survey 2 for the Younger cohort.
Younger cohort
PA group
None Low Moderate High
N Row% N Row% N Row% N Row%
2000 40 12.0 116 34.9 83 25.0 93 28.0
2003 27 8.1 116 34.9 76 22.9 113 34.0
Underweight
2006 46 13.9 122 36.7 75 22.6 89 26.8
2000 244 7.6 1074 33.6 786 24.6 1097 34.3
2003 236 7.4 1108 34.6 740 23.1 1117 34.9
Healthyweight
2006 321 10.0 1213 37.9 766 23.9 901 28.1
2000 78 8.0 333 34.1 242 24.8 323 33.1
2003 79 8.1 316 32.4 227 23.3 354 36.3
Overweight
2006 100 10.2 376 38.5 241 24.7 259 26.5
2000 67 13.2 204 40.2 113 22.3 123 24.3
2003 52 10.3 199 39.3 125 24.7 131 25.8
Obese
2006 73 14.4 191 37.7 112 22.1 131 25.8
75
Table A9: Number of women in each physical activity (PA) group at Surveys 3 and 4 by BMI
category at Survey 3 for the Mid-aged cohort.
Mid-aged cohort
PA group
None Low Moderate High
N Row% N Row% N Row% N Row%
2001 21 17.2 43 35.2 26 21.3 32 26.2 Underweight
2004 23 18.9 33 27.0 16 13.1 50 41.0
2001 442 12.4 1248 35.1 823 23.2 1041 29.3 Healthyweight
2004 416 11.7 1011 28.4 853 24.0 1274 35.8
2001 427 16.0 1024 38.5 541 20.3 671 25.2 Overweight
2004 417 15.7 779 29.3 630 23.7 837 31.4
2001 463 24.8 729 39.0 325 17.4 350 18.7 Obese
2004 398 21.3 619 33.2 406 21.7 444 23.8
Table A10: Number of women in each physical activity (PA) group at Surveys 2, 3 and 4 by BMI
category at Survey 2 for the Older cohort.
Older cohort
PA group
None Low Moderate High
N Row% N Row% N Row% N Row%
1999 29 25.0 31 26.7 28 24.1 28 24.1
2002 31 26.7 38 32.8 17 14.7 30 25.9
Underweight
2005 38 32.8 38 32.8 17 14.7 23 19.8
1999 406 19.6 694 33.5 406 19.6 566 27.3
2002 539 26.0 635 30.6 371 17.9 527 25.4
Healthyweight
2005 726 35.0 560 27.0 353 17.0 433 20.9
1999 436 28.8 489 32.3 244 16.1 347 22.9
2002 558 36.8 422 27.8 245 16.2 291 19.2
Overweight
2005 678 44.7 418 27.6 194 12.8 226 14.9
1999 251 44.0 169 29.6 75 13.1 76 13.3
2002 315 55.2 127 22.2 66 11.6 63 11.0
Obese
2005 369 64.6 99 17.3 52 9.1 51 8.9
76
77
Appendix 5: Graphs
Figure A1: Mean mental health component scores (MCS) and 95% confidence intervals in the
Younger cohort by BMI category for Surveys 1, 2, 3 and 4: U=Underweight H=Healthy weight
V=Overweight B=Obese.
Figure A2: Mean mental health component scores (MCS) and 95% confidence intervals in the Mid-
aged cohort by BMI category for Surveys 1, 2, 3 and 4: U=Underweight H=Healthy weight
V=Overweight B=Obese.
78
Figure A3: Mean mental health component scores (MCS) and 95% confidence intervals in the Older
cohort by BMI category for Surveys 1, 2, 3 and 4: U=Underweight H=Healthy weight V=Overweight
B=Obese.
Figure A4: Mean physical health component scores (PCS) and 95% confidence intervals in the
Younger cohort by BMI category for Surveys 1, 2, 3 and 4: U=Underweight H=Healthy weight
V=Overweight B=Obese.
79
Figure A5: Mean physical health component scores (PCS) and 95% confidence intervals in the Mid-
aged cohort by BMI category for Surveys 1, 2, 3 and 4: U=Underweight H=Healthy weight
V=Overweight B=Obese.
Figure A6: Mean physical health component scores (PCS) and 95% confidence intervals in the Older
cohort by BMI category for Surveys 1, 2, 3 and 4: U=Underweight H=Healthy weight V=Overweight
B=Obese.
80
Figure A7: Mental health component scores (MCS) for the Younger cohort for most commonly
occurring weight change patterns from 1996 to 2006: S=no weight change G=weight gain L=weight
loss between two consecutive surveys.
Figure A8: Mental health component scores (MCS) for the Mid-aged cohort for most commonly
occurring weight change patterns from 1996 to 2004: S=no weight change G=weight gain L=weight
loss between two consecutive surveys.
81
Figure A9: Mental health component scores (MCS) for the Older cohort for most commonly
occurring weight change patterns from 1996 to 2005: S=no weight change G=weight gain L=weight
loss between two consecutive surveys.
82
Figure A10: Physical health component score (PCS) for the Younger cohort for most commonly
occurring weight change patterns from 1996 to 2006: S=no weight change G=weight gain L=weight
loss between two consecutive surveys.
Figure A11: Physical health component scores (PCS) for the Mid-aged cohort for most commonly
occurring weight change patterns from 1996 to 2004: S=no weight change G=weight gain L=weight
loss between two consecutive surveys.
83
Figure A12: Physical health component scores (PCS) for the Older cohort for most commonly
occurring weight change patterns from 1996 to 2005: S=no weight change G=weight gain L=weight
loss between two consecutive surveys.
84
Figure A13: Relationship between BMI (in intervals of 1 kg/m2) and percentage of Younger women
reporting medical problems, surgical procedures, syptoms and GP visits at Survey 1. Data from
Younger women with BMI > 30kg/m2 are included in the BMI category labeled 30.
85
Figure A14: Relationship between BMI (in intervals of 1 kg/m2) and percentage of Mid-aged women
reporting medical problems, surgical procedures, symptoms and health care utilizations at Survey 1.
86
Figure A15: Total charges at Surveys 2 and 3 by most commonly occurring weight change patterns
for the Younger cohort from 1996 to 2006: S=no weight change G=weight gain L=weight loss
between two consecutive surveys.
Figure A16: Total number of Medicare claims at Surveys 2 and 3 by most commonly occurring
weight change patterns for the Younger cohort from 1996 to 2006: S=no weight change G=weight
gain L=weight loss between two consecutive surveys.
87
Figure A17: Number of GP visits at Surveys 2 and 3 by most commonly occurring weight change
patterns for the Younger cohort from 1996 to 2006: S=no weight change G=weight gain L=weight
loss between two consecutive surveys.
Figure A18: Total number of Medicare claims at Surveys 2 and 3 by physical activity group at
Survey 3 for the Younger cohort.
88
Figure A19: Total number of Medicare claims at Surveys 2 and 3 by BMI category at Survey 3 for
Younger women in the 'no activity' group at Survey 3.
Figure A20: Total number of Medicare claims at Surveys 2 and 3 by BMI category at Survey 3 for
Younger women in the 'low activity' group at Survey 3.
89
Figure A21: Total number of Medicare claims at Surveys 2 and 3 by BMI category at Survey 3 for
Younger women in the 'moderate activity' group at Survey 3.
Figure A22: Total number of Medicare claims at Surveys 2 and 3 by BMI category at Survey 3 for
Younger women in the 'high activity' group at Survey 3.
90
Figure A23: Total number of Medicare claims at Surveys 2, 3 and 4 by physical activity group at
Survey 4 for the Mid-aged cohort.
www.alswh.org.au
... In any given 5-year period, 20% of women of reproductive age have sufficient weight gain to progress them into a higher body mass index (BMI) category [2,3]. Furthermore, the rate of weight gain is highest (approximately 700 g per year) among women of normal BMI [4,5]. Pregnancy often represents a significant turning point in a woman's cardiovascular and metabolic health trajectory secondary to pregnancy-related changes, including relative insulin resistance, which promotes weight gain [6], and risk of developing obesity subsequently [7,8]. ...
... 2 Termination of pregnancy (TOP) 3 Three hundred and twelve infants with non-missing data included in raw data analysis, one infant with missing data had outcomes imputed and was therefore included in the imputed analysis. 4 Stillbirths excluded from infant outcomes analysis but included for analysis of maternal antenatal outcomes only. ...
... 2 Termination of pregnancy (TOP) 3 Three hundred and twelve infants with non-missing data included in raw data analysis, one infant with missing data had outcomes imputed and was therefore included in the imputed analysis. 4 Stillbirths excluded from infant outcomes analysis but included for analysis of maternal antenatal outcomes only. * = mean and standard deviation. ...
Article
Full-text available
There are well-recognised associations between excessive gestational weight gain (GWG) and adverse pregnancy outcomes, including an increased risk of pre-eclampsia, gestational diabetes and caesarean birth. The aim of the OPTIMISE randomised trial was to evaluate the effect of dietary and exercise advice among pregnant women of normal body mass index (BMI), on pregnancy and birth outcomes. The trial was conducted in Adelaide, South Australia. Pregnant women with a body mass index in the healthy weight range (18.5–24.9 kg/m2) were enrolled in a randomised controlled trial of a dietary and lifestyle intervention versus standard antenatal care. The dietitian-led dietary and lifestyle intervention over the course of pregnancy was based on the Australian Guide to Healthy Eating. Baseline characteristics of women in the two treatment groups were similar. There was no statistically significant difference in the proportion of infants with birth weight above 4.0 kg between the Lifestyle Advice and Standard Care groups (24/316 (7.59%) Lifestyle Advice versus 26/313 (8.31%) Standard Care; adjusted risk ratio (aRR) 0.91; 95% confidence interval (CI) 0.54 to 1.55; p = 0.732). Despite improvements in maternal diet quality, no significant differences between the treatment groups were observed for total GWG, or other pregnancy and birth outcomes.
... Prevalence of women affected by obesity is increasing globally, with prevalence rates increasing from 6.4% in 1975 to 14.9% in 2014 [1]. Women of childbearing age (15 to 44 years) are particularly vulnerable to weight gain, with many large cohort studies demonstrating this life stage is the time of greatest weight gain [2][3][4][5]. For example, the Australian Longitudinal Study of Women's Health has found women in their younger cohort (aged 18-23 years at survey 1) experience an average weight gain of 6.3 kg over 10 years [3]. ...
... Women of childbearing age (15 to 44 years) are particularly vulnerable to weight gain, with many large cohort studies demonstrating this life stage is the time of greatest weight gain [2][3][4][5]. For example, the Australian Longitudinal Study of Women's Health has found women in their younger cohort (aged 18-23 years at survey 1) experience an average weight gain of 6.3 kg over 10 years [3]. Notably, in women of childbearing age, pregnancy has been investigated as a potential trigger for excessive weight gain and the development of overweight and obesity. ...
Article
Full-text available
Background: Women of childbearing age are vulnerable to weight gain. This scoping review examines the extent and range of research undertaken to evaluate behavioral interventions to support women of childbearing age to prevent and treat overweight and obesity. Methods: Eight electronic databases were searched for randomized controlled trials (RCT) or systematic reviews of RCTs until 31st January 2018. Eligible studies included women of childbearing age (aged 15-44 years), evaluated interventions promoting behavior change related to diet or physical activity to achieve weight gain prevention, weight loss or maintenance and reported weight-related outcomes. Results: Ninety studies met the inclusion criteria (87 RCTs, 3 systematic reviews). Included studies were published from 1998 to 2018. The studies primarily focused on preventing excessive gestational weight gain (n = 46 RCTs, n = 2 systematic reviews), preventing postpartum weight retention (n = 18 RCTs) or a combination of the two (n = 14 RCTs, n = 1 systematic review). The RCTs predominantly evaluated interventions that aimed to change both diet and physical activity behaviors (n = 84) and were delivered in-person (n = 85). Conclusions: This scoping review identified an increasing volume of research over time undertaken to support women of childbearing age to prevent and treat overweight and obesity. It highlights, however, that little research is being undertaken to support the young adult female population unrelated to pregnancy or preconception.
... According to a government report published in 2019, about 41% of women and 11% of men aged 15 years and above were obese [12]. Because women of childbearing age between 15 to 49 years old (WCBA) accumulate weight faster than other women [13][14][15][16], the adverse consequences of obesity among this group could be pronounced. Obesity during a woman's childbearing years is associated with an increased risk of infertility, miscarriage, stillbirths and births with congenital disabilities, shoulder dystocia and other adverse obstetric outcomes [17][18][19][20][21][22]. ...