Article

Postpartum diet quality in Australian women following a gestational diabetes pregnancy

Education and Prevention Division, Australian Diabetes Council, Glebe, New South Wales, Australia.
European journal of clinical nutrition (Impact Factor: 2.71). 07/2012; 66(10):1160-5. DOI: 10.1038/ejcn.2012.84
Source: PubMed

ABSTRACT

Background/objectives:
To describe the diet quality of a national sample of Australian women with a recent history of gestational diabetes mellitus (GDM) and determine factors associated with adherence to national dietary recommendations.

Subjects/methods:
A postpartum lifestyle survey with 1499 Australian women diagnosed with GDM ≤3 years previously. Diet quality was measured using the Australian recommended food score (ARFS) and weighted by demographic and diabetes management characteristics. Multinominal logistic regression analysis was used to determine the association between diet quality and demographic characteristics, health seeking behaviours and diabetes-related risk factors.

Results:
Mean (±s.d.) ARFS was 30.9±8.1 from a possible maximum score of 74. Subscale component scores demonstrated that the nuts/legumes, grains and fruits were the most poorly scored. Factors associated with being in the highest compared with the lowest ARFS quintile included age (odds ratio (OR) 5-year increase=1.40; 95% (confidence interval) CI:1.16-1.68), tertiary education (OR=2.19; 95% CI:1.52-3.17), speaking only English (OR=1.92; 95% CI:1.19-3.08), being sufficiently physically active (OR=2.11; 95% CI:1.46-3.05), returning for postpartum blood glucose testing (OR=1.75; 95% CI:1.23-2.50) and receiving risk reduction advice from a health professional (OR=1.80; 95% CI:1.24-2.60).

Conclusions:
Despite an increased risk of type 2 diabetes, women in this study had an overall poor diet quality as measured by the ARFS. Women with GDM should be targeted for interventions aimed at achieving a postpartum diet consistent with the guidelines for chronic disease prevention. Encouraging women to return for follow-up and providing risk reduction advice may be positive initial steps to improve diet quality, but additional strategies need to be identified.

Full-text

Available from: Denise Koh, Oct 26, 2015
ORIGINAL ARTICLE
Postpartum diet quality in Australian women following a
gestational diabetes pregnancy
MK Morrison
1
, D Koh
2
, JM Lowe
3
, YD Miller
4
, AL Marshall
5
, K Colyvas
6
and CE Collins
7
BACKGROUND/OBJECTIVES: To describe the diet quality of a national sample of Australian women with a recent history of
gestational diabetes mellitus (GDM) and determine factors associated with adherence to national dietary recommendations.
SUBJECTS/METHODS: A postpartum lifestyle survey with 1499 Australian women diagnosed with GDM p3 years previously.
Diet quality was measured using the Australian recommended food score (ARFS) and weighted by demographic and diabetes
management characteristics. Multinominal logistic regression analysis was used to determine the association between diet quality
and demographic characteristics, health seeking behaviours and diabetes-related risk factors.
RESULTS: Mean (
±
s.d.) ARFS was 30.9
±
8.1 from a possible maximum score of 74. Subscale component scores demonstrated that
the nuts/legumes, grains and fruits were the most poorly scored. Factors associated with being in the highest compared with
the lowest ARFS quintile included age (odds ratio (OR) 5-year increase ¼ 1.40; 95% (confidence interval) CI:1.16–1.68), tertiary
education (OR ¼ 2.19; 95% CI:1.52–3.17), speaking only English (OR ¼ 1.92; 95% CI:1.19–3.08), being sufficiently physically active
(OR ¼ 2.11; 95% CI:1.46–3.05), returning for postpartum blood glucose testing (OR ¼ 1.75; 95% CI:1.23–2.50) and receiving risk
reduction advice from a health professional (OR ¼ 1.80; 95% CI:1.24–2.60).
CONCLUSIONS: Despite an increased risk of type 2 diabetes, women in this study had an overall poor diet quality as measured
by the ARFS. Women with GDM should be targeted for interventions aimed at achieving a postpartum diet consistent with the
guidelines for chronic disease prevention. Encouraging women to return for follow-up and providing risk reduction advice may be
positive initial steps to improve diet quality, but additional strategies need to be identified.
European Journal of Clinical Nutrition (2012) 66, 1160–1165; doi:10.1038/ejcn.2012.84; published online 11 July 2012
Keywords: gestational diabetes; diet quality; women
INTRODUCTION
Gestational diabetes mellitus (GDM) is a form of glucose
intolerance diagnosed during pregnancy.
1
It affects an estimated
5% of Australian women, increasing up to 14% in some high-risk
groups.
2
GDM is associated with increased perinatal risks, while
long-term consequences include development of type 2 diabetes
and increased cardiovascular risk.
3
Although research to date has
varied in estimates of future type 2 diabetes risk, one recent
Australian study reported a 9.6 times greater risk of type 2
diabetes in women with previous GDM and a cumulative risk of
25% after 15 years.
4
Research demonstrates that intensive lifestyle interventions
are effective in the prevention of type 2 diabetes,
5
hence, the
diagnosis of GDM provides an opportunity for early intervention in
an at-risk group. Despite this, there is some evidence to suggest
that women diagnosed with GDM have postpartum lifestyle beha-
viours that are not consistent with guidelines for the prevention of
type 2 diabetes, including suboptimal physical activity levels,
6,7
poor intake of fruit and vegetables and high-fat diets.
8–10
However,
to date there has been little published data on the postpartum
dietary intakes of Australian women with prior GDM.
Recent studies examining whole diets, as opposed to single
nutrients or dietary components, have highlighted the important
role of dietary patterns and overall diet quality in the prevention
of type 2 diabetes.
11–14
Healthful dietary patterns characterised by
high consumption of fruit and vegetables, whole grains, fish and
poultry may delay the progression to type 2 diabetes,
13,15
whereas
Western dietary patterns have been demonstrated to increase
risk.
16
Likewise, a variety of diet quality tools that measure adher-
ence to dietary guidelines have demonstrated that a high-diet
quality, representing alignment with national dietary guidelines, is
inversely associated with obesity, blood lipids, hyperglycaemia and
hyperinsulinaemia, as well as all-cause mortality and indices of self-
rated health.
17,18
In prospective studies, overall diet quality has also
been inversely associated with type 2 diabetes risk in women,
independent of body mass index (BMI).
19
Diet quality may therefore
have an important role in mediating the development of chronic
disease in a group known to be at high risk of type 2 diabetes.
The aim of this study was to describe the diet quality of a
national sample of Australian women with a recent history of GDM
and determine factors associated with adherence to national
dietary recommendations.
1
Education and Prevention Division, Australian Diabetes Council, Glebe, New South Wales, Australia;
2
Universiti Kebangsaan Malaysia, Faculty of Education, Bangi, Malaysia;
3
University of Toronto, Sunnybrook Health Sciences Ce ntre, Toronto, Ontario, Canada;
4
University of Queensland, School of Psychology, St Lucia, Queensland, Australia;
5
Queensland University of Technology, School of Public Health, Brisbane, Australia;
6
School of Mathematical and Physical Sciences, Faculty of Science and IT, University of
Newcastle, Callaghan, New South Wales, Australia and
7
Priority Research Centre in Physical Activity and Nutrition, School of Health Sciences, Faculty of Health, University of
Newcastle, Callaghan, New South Wales, Australia. Correspondence: Professor CE Collins, Nutrition and Dietetics, NHMRC CDA Research Fellow, Priority Research Centre in
Physical Activity and Nutrition, School of Health Sciences, Faculty of Health, University of Newcastle, HA12, Hunter Building, University Drive, Callaghan, New South Wales 2308,
Australia.
E-mail: Clare.Collins@newcastle.edu.au
Received 23 March 2012; revised 22 May 2012; accepted 1 June 2012; published online 11 July 2012
European Journal of Clinical Nutrition (2012) 66, 11601165
&
2012 Macmillan Publishers Limited All rights reserved 0954-3007/12
www.nature.com/ejcn
Page 1
MATERIALS AND METHODS
This was a cross-sectional study of Australian women with a recent history
of GDM. Participants were recruited from the National Diabetes Service
Scheme (NDSS) database. The NDSS is an initiative of the Australian
Government that provides subsidised blood glucose testing strips and free
syringes to residents diagnosed with diabetes. Registrants also have the
option of nominating whether or not they consent to being contacted for
research purposes. Study inclusion criteria were: diagnosed with GDM p3
years previously, registered with the NDSS and consented to be contacted
for research purposes. Women were excluded if they were aged o18 years
at the time of registration. Eligible women were invited to participate by
mail. Additional women were recruited from two major maternity clinics in
Brisbane, Australia. Women from the clinics were pregnant at the time of
recruitment, but surveyed 6 months postpartum. This additional sampling
was to recruit women with very recent GDM, who may be missed in the
NDSS database due to status update delay. The University of Newcastle
Human Research Ethics Committee, the University of Queensland, the
Royal Brisbane Women’s Hospital and the Mater Health Services approved
the study, and Diabetes Australia Ltd approved the NDSS database search.
Survey design
The survey was administered by two methods. First, a self-administered
written questionnaire and second, a telephone interview conducted in parallel
by trained interviewers using Computer-Assisted-Telephone-Interviewing for
Windows (WinCati, Version 4.2; Sawtooth Technologies, Northbrook, IL,
USA); full details of which have been described elsewhere.
6,20
Briefly, the
survey questions addressed demographics, educational attainment,
language spoken at home and occupation using standard items from
the 2001 Australian census.
21
Information regarding GDM management,
lifestyle-related risk factors, family and medical history, and postpartum
follow-up were collected by self-report. Data on respondents height and pre
and postpartum weight were self-reported and used to calculate BMI as
weight (kg)/height (m)
2
. Physical activity was assessed using the validated
Active Australia Questionnaire (AAQ), which involves recall of frequency
and duration of physical activity in the past week. The AAQ is a widely used
reliable and valid measure of physical activity.
22,23
Physical activity levels
were defined according to the AAQ criteria,
24
whereby ‘sufficient’ physical
activity was defined as the accumulation of at least 150 min of moderate or
equivalent weighted vigorous activity over at least five sessions in the past
week. Physical activity overreporters were recoded according to the AAQ
guidelines.
24
The self-administered questionnaire was pilot tested with a
convenience sample of women (n ¼ 23) from the Diabetes Australia-NSW
membership database. The telephone questionnaire was pilot tested with
six women who had a recent GDM (o3 years) pregnancy using a snowball
sampling method.
Australian recommended food score (ARFS)
Diet quality was assessed using the ARFS. The ARFS is a diet quality score
modelled on the Recommended Food Score developed by Kant and
Thompson
25
and derived from the Victorian Cancer Council’s Dietary
Questionnaire for Epidemiological Studies (DQES) food frequency
questionnaire (FFQ).
26
The DQES was originally developed for use in an
ethnically diverse cohort,
26
and has been validated against 7-day weighed
food records in young Australian women and found to an accurate
estimate of usual dietary intake.
27
The ARFS is an index of dietary variety
and nutritional quality with higher scores reflecting greater adherence to
the Dietary Guidelines for Australians
28
and food variety within core food
groups of the Australian Guide to Healthy Eating.
29
It has been validated in
a nationally representative sample of Australian women,
18
with a higher
ARFS associated with a lower percentage of energy from total and
saturated fat, a higher percentage of energy from carbohydrates and
protein, and higher intake of micronutrients.
18
The ARFS requires respondents to report their usual consumption of
foods over the preceding 12 months. It includes nine questions regarding
frequency of consumption of core foods and details of usual food choices
within each group. These questions are closed ended with multiple response
categories. This is followed by a 48-item FFQ with dichotomised response
categories. The FFQ includes only foods from the original DQES FFQ that
make a healthful contribution to dietary intake. The ARFS scoring is mostly
independent of reported quantities of food, rather is based on frequency of
consumption of core food items. Items from the 48-questions FFQ consumed
less than once a week scored zero and those consumed once a week or more
scored one. An additional score of one was allocated for each of the
following: consuming two or more fruit serves per day, four or more vege-
tables per day, the use of reduced fat or skim milk or soy milk, consuming at
least 500 ml of milk per day, using high fibre, wholemeal, rye or multigrain
breads, consuming at least four slices of bread per day, using polyunsaturated
or monounsaturated spreads or no fat spread, having one or two eggs per
week, using ricotta or cottage cheese and using low fat cheese, consuming ice
cream and cheese each less than once a week, yoghurt more than once a
week. Frequency of alcohol consumption between 1–2 days/month and
4 days/week was allocated one point, and one point was allocated for
quantity of one or two standard drinks. Zero points were added for alcohol
consumed outside these ranges. Further details are provided in Table 1.
The maximum ARFS that indicates greater adherence to the recommenda-
tions in both the Dietary Guidelines for Australians and AGHE is 74.
Table 1. ARFS: scoring method, component scores (mean
±
s.d.) and total ARFS for women with previous GDM
Food group Items allocated one point Maximum
score
Mean s.d.
Vegetables XFour vegetables/day; potatoes; tomato sauce/paste/dried; tomatoes fresh/canned; capsicum;
lettuce/endive/salad greens; cucumber; celery; beetroot; carrots; cabbage/brussels sprouts; cauliflower;
broccoli; silverbeet/spinach; peas; green beans; bean/alfalfa sprouts; pumpkin; onions/leeks; garlic;
mushrooms; zucchini
22 11.7 4.4
Fruit XTwo serves fruit/day; Xone/week of each fruit or vegetable juice; canned or frozen fruit; oranges
or other citrus; apples; pears; bananas; melons; pineapple; strawberries; apricots; peach/nectarine;
mango/pawpaw; avocado
14 4.8 3.1
Grains XFour slices bread/day; Xone/week of each bread type—white high fibre; wholemeal; rye; multigrain;
wholemeal; Xone/week allbran; sultana bran/fibre plus/branflakes; weet-bix/vitabrits/weeties; rice;
pasta/noodles; vegemite/marmite/promite; porridge; muesli; cornflakes/nutrigrain/special K;
14 4.3 1.7
Dairy 4500 ml milk/day; reduced fat or skim; pone/week cheese, ice cream;Xone week yoghurt;
ricotta/cottage cheese; low-fat cheese
7 2.7 1.1
Nuts/
legumes
Nuts; peanut butter; Xone/week of each baked beans; soy beans/tofu; soya milk; other beans
(chickpeas, lentils)
7 1.7 1.1
Meat, eggs,
poultry
1–4/week of beef; veal; lamb; pork; chicken; up to two eggs/week 5 2.8 1.2
Fish 1–4/week of fish (steamed, baked, grilled); canned fish (salmon, tuna, sardines) 2 1.0 0.8
Fats Use polyunsaturated/monounsaturated spread or nil margarine 1 0.5 0.5
Alcohol oOne/month up to 4 days/week beer/wine/spirits; maximum/day 1–2 standard drinks 2 1.1 0.8
Total ARFS 74 30.9 8.1
Abbreviations: ARFS, Australian recommended food score; GDM, gestational diabetes mellitus; s.d., standard deviation.
Postpartum diet quality in GDM
MK Morrison et al
1161
& 2012 Macmillan Publishers Limited European Journal of Clinical Nutrition (2012) 1160 1165
Page 2
For analysis, ARFS was divided into quintiles to create a categorical
variable, with quintile one representing the lowest category of dietary
quality and quintile five the highest dietary quality. Those with more than
four missing items were excluded from the analysis and missing values
were recoded as zero for those with up to four items missing.
Statistical analysis
To correct for potential sampling bias, descriptive statistics, ARFS and
component scores were adjusted for age, country of birth, state of
residence and insulin usage using weights from 15 880 women with
complete data in the NDSS data set. Unweighted analyses were used to
examine the predictors of ARFS. Univariate w
2
analyses were performed to
determine variables associated with ARFS quintiles. Statistically significant
variables (Pp0.05), as well as age and BMI, were included in a multiple
variable multinominal logistic regression analysis. Likelihood ratio tests
were used to assess significance of effects in the logistic regression model
and used as the basis for retaining a variable in the model. The Pearson w
2
was used to check the goodness of fit of the model. The multiple variable
model provides odds ratio (OR) estimates adjusted for other variables in
the model. ORs for quintiles 2–5 were referenced to quintile 1, and 95%
confidence intervals were calculated for each of these quintiles. Analyses
were completed using SPSS version 18.0 (IBM Corp., Somers, NY, USA).
RESULTS
Of the 15 893 women registered on the NDSS with gestational
diabetes, invitations were sent to 5147 women who met the
inclusion criteria, with 302 women unable to be contacted. Of
those invited, 1736 women consented to participate (36%
response rate). Ineligible respondents who were currently
pregnant (n ¼ 189), diagnosed with other forms of diabetes
(n ¼ 9) or those with missing demographic data required for
sample weighting (n ¼ 39), were excluded from analyses. Final
data were available for 1499 respondents.
Using weighted data, the mean age
±
s.d. was 34.2
±
5.1.
Approximately two thirds were Australian born (64.5%) or
currently employed (67.4%). Less than half (40.1%) were tertiary
educated, 22.6% spoke a language other than English and 1.7%
were from an Aboriginal or Torres Strait Islander background. A
previous diagnosis of GDM (before the index pregnancy) was
reported by 13.1% of respondents, 25.7% used insulin during the
index pregnancy, 29.0% were overweight and 26.3% were obese
with a mean (
±
s.d.) self-reported BMI of 27.1
±
6.5.
The ARFS was calculated for 1447 women (52 women had more
than four missing items, so were excluded from the analyses).
Mean (
±
s.d.) diet quality score was 30.9
±
8.1 from a possible
maximum score of 74. Subscale component scores are reported in
Table 1 and demonstrate that the meat, alcohol and vegetable
components were the most highly scored groups relative to the
other components with nuts/legumes, grains and fruits the most
poorly scored.
Table 2 reports the demographic characteristics, health seeking
behaviours and diabetes-related risk factors of women with GDM
by ARFS quintile. Independent variables found to be significant
(Pp0.05) in univariate analyses included region of birth, speaking
only English, being tertiary educated, returning for postpartum
follow-up blood glucose testing, being sufficiently physically
active and receiving risk reduction advice from a health
professional. When these variables (as well as age and BMI) were
included in multinominal logistic regression analyses, they
remained significant, with the exception of region of birth that
was excluded from the final model, see Table 3. The Pearson w
2
was not significant (ChiSq (5116) ¼ 5116, P ¼ 0.499), indicating a
satisfactory fit of the model to the data.
Table 3 contains all the significant effects in the multiple
variable multinomial logistic regression model expressed as OR
and 95% CIs for ARFS quintiles 2–5, using the lowest quintile as
the reference group for each OR. The reference groups for the
categorical explanatory variables are indicated by OR ¼ 1.
Interpretation of the effects is similar for all variables in the model
as they have a positive relationship with dietary score. The relative
impact of the six significant factors can be assessed by comparing
the OR’s for ARFS quintile 5. Factors associated with being in the
highest compared with the lowest ARFS quintile included age
(OR 5-year increase ¼ 1.40; 95% CI:1.16–1.68), tertiary education
(OR ¼ 2.19; 95% CI:1.52–3.17), speaking only English (OR ¼ 1.92;
95% CI:1.19–3.08), being sufficiently physically active (OR ¼ 2.11;
95% CI:1.46–3.05), returning for postpartum blood glucose testing
(OR ¼ 1.75; 95% CI:1.23–2.50) and receiving risk reduction advice
from a health professional (OR ¼ 1.80; 95% CI:1.24–2.60). There
was a trend such that as BMI increased women were less likely to
be in the highest compared with the lowest ARFS quintile
(reference group). However, this failed to reach significance in the
likelihood ratio test (P ¼ 0.078) and was excluded from the final
model. Table 3 also provides OR estimates for the other three
quintiles of diet quality to show the overall pattern across
quintiles.
DISCUSSION
This is the first Australian study to date investigating diet quality in
a national sample of women with a history of GDM. Despite their
increased risk of developing type 2 diabetes, women in this study
had an overall poor diet quality as measured by the ARFS,
indicating suboptimal intake of key food groups and eating
patterns not aligned with national guidelines.
28
These findings are
consistent with research done with representative samples of
young and mid-aged Australian women whereby poor diet quality
and disparities between national food group recommendations
and dietary intakes have been reported.
18,30,31
Analysis by component subscores indicated that nuts/legumes,
fruit and grains were the food groups most poorly scored by
women with previous GDM. To achieve a higher score in these
food categories, women would need to consume a variety of high
fibre and wholegrain breads and cereals, legumes and increase
the amount and variety of fruit consumed each week. Despite an
already elevated risk of type 2 diabetes in this group, it is plausible
that poor diet quality, as found in this study, may further increase
their risk for long-term chronic disease risk including both type 2
diabetes
14,32,33
and cardiovascular disease.
34
This highlights a
need to target specific dietary changes for women with previous
GDM to prevent subsequent chronic disease.
Consistent with other studies, we found that tertiary educated
35
and older women had better diet quality. These results are
consistent with the findings of Collins et al.,
18
who found the same
relationship in a nationally representative sample of mid-aged
Australian women.
18
In the current study, we also found that
those who spoke only English were almost twice as likely to
have an ARFS in the upper quintile after adjustment for educa-
tion and other significant variables, indicating that language or
cultural barriers influence an individual’s ability to achieve a high
quality diet. Considering that the risk of developing GDM in
Australia is greater among women from non-English speaking
backgrounds,
2,36
this is an important finding and indicates that
this group may require additional support and/or targeted
interventions.
As may be expected, the current study confirms that women
who practise other preventative health behaviours are more likely
to report better quality dietary intakes. In the present study,
women who met the guidelines for physical activity were more
than twice as likely to be in the upper compared with the lower
quintile for diet quality. Women who sought postpartum testing
for diabetes also reported better diet quality. Although previous
studies have shown low rates of postpartum testing for diabetes
following a GDM pregnancy,
37–39
this finding suggests that either
they are the more motivated group to improve their lifestyle
Postpartum diet quality in GDM
MK Morrison et al
1162
European Journal of Clinical Nutrition (2012) 1160 1165 & 2012 Macmillan Publishers Limited
Page 3
following GDM or that being advised to return for follow-up acts
as a motivating factor for improved diet quality.
The finding that women who received risk reduction advice
from a health professional were more likely to have better diet
quality highlights the importance of providing lifestyle interven-
tions targeting postpartum risk reduction. Despite this, we have
previously demonstrated poor follow-up and limited provision of
postpartum dietary advice for this high-risk group.
40
With diabetes
prevention studies providing evidence of the benefit of intensive
lifestyle interventions for reducing the incidence of type 2
diabetes in those at highest risk,
41–43
these results support the
need for additional resources to address postpartum lifestyle
management.
The association between BMI and diet quality has been reported in
previous studies.
44,45
Although we found a trend towards women
with a lower BMI having better diet quality, these results did not
reach statistical significance in logistic regression analysis. Postpartum
weight retention may have confounded this relationship between
Table 2. Percentage of women in each quintile of the ARFS by demographic characteristics, health seeking behaviours and diabetes-related risk
factors
Quintiles of ARFS; 1 ¼ lowest, 5 ¼ highest (ARFS score, %)
Unweighted
mean
±
s.d.
ARFS
1
(p24)
n ¼ 312
2
(2529)
n ¼ 304
3
(3033)
n ¼ 256
4
(3438)
n ¼ 321
5
(39 þ )
n ¼ 254
w
2
Region of birth
Australia 31.1
±
8.0 20.6 22.0 18.9 21.5 17.0
Asia 29.7
±
9.0 31.3 17.4 16.7 16.7 18.1
Pacific Islands 33.0
±
6.9 10.2 22.0 15.3 35.6 16.9
w
2
(20) ¼ 30.32, P ¼ 0.07
Europe 31.4
±
8.5 24.1 18.0 12.8 24.8 20.3
Middle East 29.0
±
9.3 35.0 20.0 5.0 20.0 20.0
Other 33.1
±
7.8 14.6 17.1 17.1 31.7 19.5
Language
English only 31.3
±
8.1 19.7 21.7 18.4 22.4 17.9
w
2
(4) ¼ 14.38, P ¼ 0.006
a
Other 30.1
±
8.5 30.6 17.5 14.4 21.4 16.2
Tertiary educated
Yes 32.3
±
7.8 16.0 21.0 17.2 24.6 21.2
w
2
(4) ¼ 22.22, Po0.001
a
No 30.3
±
8.3 25.0 21.0 18.0 20.6 15.3
Employed
Yes 31.3
±
8.1 20.6 21.0 18.2 22.7 17.5
w
2
(4) ¼ 1.91, P ¼ 0.75
No 30.8
±
8.3 23.5 21.0 16.8 21.2 17.5
Insulin requiring
Yes 31.1
±
7.9 22.3 21.6 18.0 19.2 18.9
w
2
(4) ¼ 3.29, P ¼ 0.51
No 31.0
±
8.7 21.3 20.8 17.6 23.4 17.0
Previous GDM
Yes 31.0
±
8.1 19.2 17.7 19.2 25.1 18.7
w
2
(4) ¼ 3.29, P ¼ 0.51
No 31.7
±
8.2 22.0 21.6 17.5 21.7 17.3
Return for follow-up BGL
Yes 31.9
±
8.4 18.5 20.5 17.5 23.3 20.2
w
2
(4) ¼ 13.52, P ¼ 0.009
a
No 30.3
±
7.8 24.7 21.5 17.9 21.1 14.8
Sufficiently active
Yes 32.6
±
8.1 16.6 18.4 18.2 24.2 22.8
w
2
(4) ¼ 23.64, Po0.001
a
No 30.4
±
8.1 23.7 22.9 17.6 20.9 14.9
Risk reduction advice
Yes 31.7
±
8.0 19.4 21.0 16.9 23.2 19.4
w
2
(4) ¼ 12.41, P ¼ 0.02
a
No 30.0
±
8.3 25.3 20.9 19.1 20.4 14.3
Hyperlipidaemia
Yes 31.5
±
8.0 17.9 21.4 17.9 27.9 15.0
w
2
(4) ¼ 3.83, P ¼ 0.43
No 31.1
±
8.2 22.0 21.0 17.7 21.6 17.8
Hypertension
Yes 30.5
±
8.8 25.4 21.8 16.2 18.8 17.8
w
2
(4) ¼ 3.08, P ¼ 0.54
No 31.2
±
8.1 21.0 20.9 17.9 22.7 17.5
Type 2 diabetes
Yes 31.1
±
7.2 12.1 30.3 21.2 18.2 18.2
w
2
(4) ¼ 3.72, P ¼ 0.51
No 30.1
±
8.2 21.8 20.8 17.6 22.3 17.5
Abbreviations: ARFS, Australian recommended food score; BGL, blood glucose level; GDM, gestational diabetes mellitus; s.d., standard deviation.
a
Statistically
significant at Pp0.05.
Postpartum diet quality in GDM
MK Morrison et al
1163
& 2012 Macmillan Publishers Limited European Journal of Clinical Nutrition (2012) 1160 1165
Page 4
weight and diet quality. The use of self-reported weight may
also have biased BMI calculations. Studies using postal survey
methodology have demonstrated that self-report underestimates
weight in women by an average of 0.95 kg, with those in
overweight and obese categories underestimating by up to
2.5 kg.
46
With both body weight and dietary patterns being
important determinants of type 2 diabetes risk,
19
this trend
warrants further investigation in particular with women with a
longer postpartum duration.
This study has several limitations; most notable is the low (36%)
response rate. It is also possible that a response bias towards
potentially more health conscious women may present an
optimistic assessment of postpartum diet quality. As with any
tool used to measure dietary intake, the ARFS has a number of
limitations. Respondents are asked to report their usual consump-
tion of foods over the preceding 12 months, therefore results may
be influenced by the season in which the questionnaire was
administered or be more likely to emphasise recently consumed
foods. It is possible that our findings are also influenced by under
or overreporting. However, as the ARFS focuses on frequency of
consumption of core foods and the variety of food choices within
those groups, the scoring is independent of reported amounts of
food items that would have limited the associated measurement
error. Further, we did not collect longitudinal data to determine
associations between diet quality and long-term chronic disease
risk. Despite these limitations, our study did have a large sample
size drawn from a population-based registry as opposed to a
hospital or insurance-based data set, strengthening the applic-
ability of the study to a larger population of women with prior
GDM.
CONCLUSION
Women with previous GDM should be targeted for dietary
interventions aimed at improving overall diet quality in the
postpartum period. In particular, barriers to healthy eating may
need to be addressed in those at highest risk of poor diet quality
including younger women, those with a lower level of education,
women who speak a language other than English and those who
do not seek postpartum follow-up. Our study suggests that health
professionals could have an important role in providing post-
partum risk reduction advice that may improve overall diet quality,
and further research is needed to assess the impact of health
professional advice on preventative behaviours and subsequent
chronic disease risk among women with GDM. A systematic
approach to follow-up is urgently needed to ensure that all
women diagnosed with GDM receive adequate information and
support to achieve a diet consistent with the guidelines for
chronic disease prevention.
CONFLICT OF INTEREST
The authors declare no conflict of interest.
ACKNOWLEDGEMENTS
We are very grateful to the women with GDM whose participation made this study
possible. This study was funded by the Diabetes Australia Research Trust, the
University of Queensland’s Enabling Grants Scheme, the National University of
Malaysia PhD Scholarship, the Dietitians Association of Australia Unilever Post-
Graduate Research Scholarship, the Lions District 201N3 Diabetes Foundation and
the Neville Samson Diabetes Grants-In-Aid. CE Collins if funded by a National Health
and Medical Research Council Career Development Fellowship. We acknowledge
David McIntyre and Wendy Brown for input into the sampling strategies and
reviewing the measures used, and the National Diabetes Services Scheme and
Diabetes Australia-NSW for their support.
REFERENCES
1 Hoffman L, Nolan C, Wilson JD, Oats JN, Simmons D. Gestational diabetes
mellitus—management guidelines. Med J Aust 1998; 169: 93–97.
2 Templeton M, Pieris-Caldwell I. Gestational Diabetes Mellitus in Australia, 2005–06.
Australian Institute of Health and Welfare: Canberra, 2008.
Table 3. Effect sizes for the multinomial logistic regression model of variables associated with diet quality
a
Quintiles of ARFS; 1 ¼ lowest
b
,5¼ highest (ARFS score)
Quintile 2 (2529) Quintile 3 (3033) Quintile 4 (34-38) Quintile 5 (39 þ )
Adjusted OR 95% CI Adjusted OR 95% CI Adjusted OR 95% CI Adjusted OR 95% CI P
c
Age (5 year increase) 0.91 (0.77–1.08) 1.15 (0.96–1.37) 1.29 (1.09–1.53) 1.40 (1.16–1.68) o0.001
Tertiary educated
Yes 1.74 (1.22–2.47) 1.55 (1.07–2.24) 1.93 (1.36–2.74) 2.19 (1.52–3.17) o0.001
No 1111
Sufficiently active
Yes 1.12 (0.78–1.61) 1.43 (0.99–2.07) 1.60 (1.12–2.27) 2.11 (1.46–3.05) o0.001
No 1111
Follow-up blood glucose testing
Yes 1.24 (0.89–1.72) 1.31 (0.93–1.85) 1.44 (1.03–1.99) 1.75 (1.23–2.50) 0.03
No 1111
Language
English only 2.11 (1.35–3.32) 2.10 (1.30–3.38) 1.67 (1.09–2.57) 1.92 (1.19–3.08) 0.005
Other 1111
Risk reduction advice
Yes 1.31 (0.93–1.83) 1.19 (0.84–1.70) 1.55 (1.10–2.18) 1.80 (1.24–2.60) 0.02
No 1111
Abbreviation: ARFS, Australian recommended food score; CI, confidence interval; OR, odds ratio.
a
Diet quality was the response variable in the model and was
measured using ARFS quintiles; the significant effects related to diet quality are the six variables listed in the table.
b
Quintile 1 is the reference group
(scoresp24).
c
Significance of the effect of each variable by the likelihood ratio test.
Postpartum diet quality in GDM
MK Morrison et al
1164
European Journal of Clinical Nutrition (2012) 1160 1165 & 2012 Macmillan Publishers Limited
Page 5
3 Kitzmiller J, Dang-Kilduff L, Taslimi MM. Gestational diabetes after delivery: short-
term management and long-term risks. Diabetes Care 2007; 30: S225–S235.
4 Lee A, Hiscock RJ, Wein P, Walker S, Permezel M. Gestational diabetes mellitus:
clinical predictors and long-term risk of developing type 2 diabetes. Diabetes Care
2007; 30: 878–883.
5 The Diabetes Prevention Program Research Group. Reduction in the incidence of
type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002; 346:
393–403.
6 Koh D, Miller YD, Marshall AL, Brown WJ, McIntyre D. Health-enhancing physical
activity behavior and related factors in postpartum women with recent gesta-
tional diabetes mellitus. J Sci Med Sport 2010; 13: 42–45.
7 Smith BJ, Cheung NW, Bauman AE, Zehle K, McLean M. Postpartum physical
activity and related psychosoci al factors among women with recent gestational
diabetes mellitus. Diabetes Care 2005; 28: 2650–2654.
8 Zehle K, Smith B.J, Chey T, McLean M, Bauman A.E, Cheung N.W. Psychosocial
factors related to diet among women with recent gestational diabetes: oppor-
tunities for intervention. Diabetes Educator 2008; 34: 807–814.
9 Kieffer EC, Sinco B, Kim C. Health behaviors among women of reproductive age
with and without a history of gestational diabetes mellitus. Diabetes Care 2006;
29: 1788–1793.
10 Stage E, Ronneby H, Damm P. Lifestyle change after gestational diabetes. Diabetes
Res Clin Pract 2004; 63: 67–72.
11 Fung TT, McCullough M, van Dam RM, Hu FB. A prospective study of overall
diet qual ity and risk of type 2 diabetes in women. Diabetes Care 2007; 30:
1753–1757.
12 Hodge AM, English DR, O’Dea K, Giles GG. Dietary patterns and diabetes inci-
dence in the Melbourne Collaborative Cohort Study. Am J Epidemiol 2007; 165:
603–610.
13 Montonen J, Ja
¨
rvinen R, Helio
¨
vaara M, Reunanen A, Aromaa A, Knekt P. Food
consumption and the incidence of type II diabetes mellitus. Eur J Clin Nutr 2005;
59: 441–448.
14 Nettleton JA, Steffen LM, Ni H, Liu K, Jacobs Jr DR. Dietary patterns and risk of
incident type 2 diabetes in the multi-ethnic study of atherosclerosis (MESA).
Diabetes Care 2008; 31: 1777–1782.
15 Esposito K, Kastorini CM, Panagiotakos DB, Giugliano D. Prevention of type 2
diabetes by dietary patterns: a systematic review of prospective studies and
meta-analysis. Metab Syndr Relat Disord 2010; 8: 471–476.
16 Fung TT, Schulze M, Manson JE, Willett WC, Hu FB. Dietary patterns, meat
intake, and the risk of type 2 diabetes in women. Arch Intern Med 2004; 164:
2235–2240.
17 Wirt A, Collins CE. Diet Quality - what is it and does it matter? Public Health Nutr
2009; 12: 2473–2492.
18 Collins CE, Young AF, Hodge A. Diet quality is associated with higher nutrient
intake and self-rated health in mid-aged women.
J Am Coll Nutr 2008; 27:
146–157.
19 Hu FB, Manson JE, Stampfer MJ, Colditz GA, Liu S, Solomon CG, Willett WC. Diet
and lifestyle and risk of type 2 diabetes mellitus in women. New Engl J Med 2001;
345: 790–797.
20 Morrison MK, Lowe JM, Collins CE. Perceived risk of type 2 diabetes in Australian
women with a recent history of gestational diabetes mellitus. Diabet Med 2010;
27: 882–886.
21 Trewin D. Census Dictionary. Australian Bureau of Statistics, Commonwealth of
Australia. Australian Bureau of Statistics: Canberra, 2001.
22 Armstrong T, Bauman A, Davies J. Physical Activity Patterns of Australian Adults.
Results of the 1999 National Physical Activity Survey. Australian Institute of Health
and Welfare: Canberra, 2000.
23 Brown WJ, Trost SG, Bauman A, Mummery K, Owen N. Test-retest reliability of four
physical activity measures used in population surveys. J Sci Med Sport 2004; 7:
205–215.
24 Australian Institute of Health and Welfare. The Active Australia Survey: a Guide and
Manual for Implementation, Analysis and Reporting. Australian Institute of Health
and Welfare: Canberra, 2003.
25 Kant AK, Thompson FE. Measures of overall diet quality from a food frequency
Questionnaire: National health interview survey 1992. Nutr Res 1997; 17:
1443–1456.
26 Ireland P, Jolley D, Giles G, O’Dea K, Powles J, Rutishauser I, Wahlqvist ML,
Williams J. Development of the Melbourne FFQ: a food frequency questionnaire
for use in an Australian prospective study involving an ethnically diverse cohort.
Asia Pac Clin Nutr 1994; 3: 19–31.
27 Hodge A, Patterson A, Brown W, Ireland P, Giles G. The anti cancer council of
Victoria FFQ. Relative validity of nutrient intakes compared with diet diaries in
young to middle-aged women in a study of iron supplementation. Aust NZ
J Public Health 2000; 24: 576–583.
28 Commonwealth of Australia, National Health and Medical Research Council.
Dietary Guidelines for Australian Adults: A Guide to Healthy Eating. Australian
Government Publishing Service: Canberra, Australia, 2003.
29 Smith A, Kellett E, Schmerlaib Y. The Australian Guide to Healthy Eating. Australian
Government Publishing Service: Canberra, 1998.
30 Hure A, Young A, Smith R, Collins C. Diet and pregnancy status in Australian
women. Public Health Nutr 2009; 12: 853–861.
31 Blumfield ML, Hure AJ, Macdonald-Wicks LK, Patterson AJ, Smith R, Collins CE.
Disparities exist between National food group recommendations and the dietary
intakes of women. BMC Women’s Health 2011; 8: 11–37.
32 de Munter JS, Hu FB, Spiegelman D, Franz M, van Dam RM. Whole grain, bran,
and germ intake and risk of type 2 diabetes: a prospective cohort study and
systematic review. PLoS Med 2007; 4: e261.
33 Venn BJ, Mann JI. Cereal grains, legumes and diabetes. Eur J Clin Nutr 2004; 58:
1443–1461.
34 Belin RJ, Greenland P, Allison M, Martin L, Shikany JM, Larson J, Tinker L, Howard
BV, Lloyd-Jones D, Van Horn L. Diet quality and the risk of cardiovascular disease:
the Women’s Health Initiative (WHI). Am J Clin Nutr 2011; 94: 49–57.
35 Darmon N, Drewnowski A. Does social class predict diet quality? Am J Clin Nutr
2008;
87: 1107–1117.
36 Holdenson Z, Catanzariti L, Phillips G, Waters AM. A picture of diabetes in over-
seas-born Australians. Australian Institute of Health and Welfare 2003.
37 Blatt AJ, Nakamoto JM, Kaufman HW. Gaps in diabetes screening during preg-
nancy and postpartum. Obstet Gynecol 2011; 117: 61–68.
38 Shah BR, Lipscombe LL, Feig DS, Lowe JM. Missed opportunities for type 2 dia-
betes testing following gestational diabetes: a population-based cohort study.
BJOG 2011; 118: 1484–1490.
39 Morrison MK, Collins CE, Lowe JM. Postnatal testing for diabetes in Australian
women following gestational diabetes mellitus. Aust NZ J Obstet Gynaecol 2009;
49: 494–498.
40 Morrison MK, Collins CE, Lowe JM. Dietetic practice in the management of
gestational diabetes mellitus: A survey of Australian dietitians. Nutr Dietetics 2011;
68: 189–194.
41 Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA,
Nathan DM. Reduction in the incidence of type 2 diabetes with lifestyle inter-
vention or metformin. N Engl J Med 2002; 346: 393–403.
42 Louheranta J, Mannelin A, Rastas M, Salmin en M, Eriksson V, Uusitupa M, Tuo-
milehto J. The Finnish Diabetes Prevention Study (DPS): lifestyle intervention and
3-year results on diet and physical acti vity. Diabetes Care 2003; 26: 3230–3236.
43 Pan XR, Li GW, Hu YH, Wang JX, Yang WY, An ZX et al. Effects of diet and exercise
in preventing NIDDM in people with impaired glucosetolerance. The Da Qing IGT
and Diabetes Study. Diabetes Care 1997; 20: 537–544.
44 Shah BS, Freeland-Graves JH, Cahill JM, Lu H, Graves GR. Diet quality as measured
by the healthy eating index and the association with lipid profile in low-income
women in early postpartum. JADA 2010; 110: 274–279.
45 Laraia BA, Bodnar LM, Siega-Riz AM. Pregravid body mass index is negatively
associated with diet quality during pregnancy. Public Health Nutr 2007; 10:
920–926.
46 Burton NW, Brown W, Dobson A. Accuracy of body mass index estimated from
self-reported height and weight in mid-aged Australian women. Aust NZ J Public
Health 2010; 34: 620–623.
Postpartum diet quality in GDM
MK Morrison et al
1165
& 2012 Macmillan Publishers Limited European Journal of Clinical Nutrition (2012) 1160 1165
Page 6
  • Source
    • "In addition, very few young women achieved a high diet quality score. The mean diet quality score in the highest tertile of each index was not high, indicating that interventions seeking to optimise diet quality in this age group are warranted as has been suggested previously [26–28]. In addition, a recent systematic review [29] has highlighted that intervention studies specifically targeting body weight are needed to prevent the development of overweight and obesity in this age group. "
    [Show abstract] [Hide abstract] ABSTRACT: This study investigates the relationship between diet quality and weight gain in young women. Young women (n = 4,287, with 1,356 women identified as plausible subsample aged 27.6 ± 1.5 years at baseline) sampled from the Australian Longitudinal Study on Women's Health study completed food frequency questionnaires in 2003, which were used to evaluate diet quality using three indices: Australian Recommended Food Score (ARFS), Australian Diet Quality Index (Aus-DQI), and Fruit and Vegetable Index (FAVI). Weight was self-reported in 2003 and 2009. Multivariate linear regression was used to examine the association between tertiles of each diet quality index and weight change from 2003 to 2009. The ARFS and FAVI were significant predictors of 6-year weight change in this group of young women, while Aus-DQI did not predict weight change (P > 0.05). In the fully adjusted model, those who were in the top tertile of the ARFS significantly gained lower weight gain compared with the lower tertile for the plausible TEI sub-sample (β = -1.6 kg (95% CI: -2.67 to -0.56), P = 0.003). In the fully adjustment model, young women were classified in the highest FAVI tertile and gained significantly less weight than those in the lowest tertile for the plausible TEI (β = -1.6 kg (95% CI: -2.4 to -0.3) P = 0.01). In conclusion, overall diet quality measured by the ARFS and the frequency and variety of fruit and vegetable consumption may predict long-term weight gain in young women. Therefore, health promotion programs encouraging frequent consumption of a wide variety of fruits and vegetables are warranted.
    Full-text · Article · Aug 2013 · Journal of obesity
    • "Moreover, guided by life-course theory regarding critical periods (Lu & Halfon, 2003), we proposed that motherhood (either with a first or an additional child) may bring additional life-stage stresses and burdens that further compromise health of women, if not offset by resources and support. Such chronic and life-stage burdens may be reflected in various risk factors that affect health, such as lower education (Morrison et al., 2012), ethnic minority status (Boardley, Sargent, Coker, Hussey, & Sharpe, 1995), obesity (Gjerdingen et al., 2009), or less optimal infant feeding practices (Hatton et al., 2005). Because both chronic and current stresses associated with motherhood increase women's vulnerability in the psychosocial domains (depressive symptoms and body image dissatisfaction ) and behavioral domains of health (diet and exercise, use of various psychoactive substances, and general self-care; Beeghly et al., 2003; Park et al., 2009; Walker, 1989), we focus on these in this study. "
    [Show abstract] [Hide abstract] ABSTRACT: During the postpartum period, women may experience unfavorable psychosocial and behavioral health in multiple domains with adverse effects on parenting and maternal and infant health. Yet, little is known about the accumulation of poor health across the domains of depressive symptoms; body image; diet and physical activity; substance use including smoking and alcohol; and general self-care at 6 weeks postpartum, the usual end of maternity care. The aims of this study were to evaluate relationships among the domains comprising psychosocial and behavioral health and to examine the distribution and risk factors associated with cumulative poor psychosocial and behavioral health at 6 weeks postpartum. This study was a secondary analysis of cumulative poor health assessed by self-report scales for depressive symptoms, body image dissatisfaction, diet and exercise, substance use, and general self-care among 419 low-income White, African American, and Hispanic women at 6 weeks postpartum. Multivariable Poisson and logistic regression were used in key analyses. The correlation among psychosocial and behavioral domains had a range of r = .50-.00. In this sample of women, 45% had two or more domains in which they had poor health. The model testing risk factors for cumulative poor health was significant (likelihood ratio chi-square = 39.26, df = 11, p < 0.05), with two significant factors: not exclusively breastfeeding (odds ratio [OR] = 1.459, 95% confidence interval [CI] [1.119, 1.901]) and Hispanic ethnicity (OR = 0.707, 95% CI [0.582, 0.858], psuedo-R = .029). Within individual domains, significant risk factors (body mass index, not exclusively breastfeeding, ethnicity, education level, and parity) varied by domain. Many low-income women postpartum have poor psychosocial and behavioral health in multiple domains, which constitute areas for health promotion and early disease prevention.
    No preview · Article · Jul 2013 · Nursing research
  • No preview · Article · Jan 2014 · Women and Birth
Show more