Outcomes of a 12-month web-based intervention for overweight and obese men.

Abstract

This study assessed the effect of a 1-year internet-based weight loss intervention for men.
Four hundred forty-one overweight and obese men were randomized to intervention or delayed treatment. Participants completed a Web-based assessment of diet and physical activity behaviors and weekly tailored Web modules addressing weight-related behaviors.
At 12 months compared to controls, intervention men decreased percent of energy from saturated fat and increased grams of fiber and fruit/vegetable servings per 1.000 kcal (p values < 0.001) and walked 16 min more per day (p < 0.05). No between-group differences in body mass index (BMI), weight, or waist circumference were seen, but among completers, men in the highest tertile of intervention participation had lower weight (98.74 vs. 102.37 kg), BMI (32.38 vs. 33.46), and waist circumference (42.17 vs. 43.47 cm) compared to men who participated less often.
The intervention improved diet and activity behaviors, but weight loss occurred only for those with the highest adherence.

Full text

Page 1

ORIGINAL ARTICLE
Outcomes of a 12-Month Web-Based Intervention
for Overweight and Obese Men
Kevin Patrick, M.D., MS & Karen J. Calfas, Ph.D. & Gregory J. Norman, Ph.D. &
Dori Rosenberg, MPH, MS & Marion F. Zabinski, Ph.D. & James F. Sallis, Ph.D. &
Cheryl L. Rock, Ph.D., RD & Lindsay W. Dillon, MPH
Published online: 6 August 2011
# The Society of Behavioral Medicine 2011
Abstract
Background This study assessed the effect of a 1-year
internet-based weight loss intervention for men.
Methods Four hundred forty-one overweight and obese
men were randomized to intervention or delayed treatment.
Participants completed a Web-based assessment of diet and
physical activity behaviors and weekly tailored Web
modules addressing weight-related behaviors.
Results At 12 months compared to controls, intervention
men decreased percent of energy from saturated fat and
increased grams of fiber and fruit/vegetable servings per
1.000 kcal (p values < 0.001) and walked 16 min more per
day (p < 0.05). No between-group differences in body mass
index (BMI), weight, or waist circumference were seen, but
among completers, men in the highest tertile of intervention
participation had lower weight (98.74 vs. 102.37 kg), BMI
(32.38 vs. 33.46), and waist circumference (42.17 vs. 43.47
cm) compared to men who participated less often.
Conclusions The intervention improved diet and activity
behaviors, but weight loss occurred only for those with the
highest adherence.
Keywords Weight control.Diet.Physical activity.Men’s
health
Introduction
The surgeon general has highlighted the critical problem of
obesity in the USA [1]. Among adults aged at least
20 years, 66.3% are overweight or obese (body mass index
[BMI]≥25 kg/m2) with 32.2% obese (BMI=30–39.9
kg/m2) and 4.8% extremely obese (BMI≥40 kg/m2) [2].
Overweight is associated with increased risk for several
types of cancer, cardiovascular diseases, type 2 diabetes,
and other chronic diseases [3, 4], and the prevalence of
overweight among men continues to increase [5]. While the
overall prevalence of obesity (BMI≥30 kg/m2) is higher
among women (40.1%) than men (33.9%), the prevalence
of obesity in men is increasing [2]. High energy intake,
insufficient physical activity, and sedentary behaviors
contribute to obesity, so multibehavior interventions are
needed for weight loss and weight loss maintenance [6, 7].
Most Americans do not meet national physical activity or
dietary guidelines [8, 9] so intervening on these behaviors
is a cornerstone of any intervention aimed at promoting
weight loss in the overweight.
Women are much more likely than men to participate
in weight loss programs, and for this reason most weight
loss interventions have been evaluated with samples that
are predominantly female [10–13]. An early study of
preferences for weight loss treatment options among men
showed a strong preference not to give up alcohol, engage
in strenuous exercise or focus on psychological aspects of
weight [14]. Other studies show that men are less likely to
endorse traditional settings for weight loss interventions
like medical or commercial settings or psychology clinics
[15, 16]. Instead, men indicate a preference for individu-
alized programs located in a gym that allow individual
tailoring of both nutrition and exercise. Motivators to seek
treatment varied by population studied but include
K. Patrick (*):K. J. Calfas:G. J. Norman:D. Rosenberg:
M. F. Zabinski:C. L. Rock:L. W. Dillon
Department of Family and Preventive Medicine,
University of California,
9500 Gilman Drive 0811, La Jolla,
San Diego, CA 92093-0811, USA
e-mail: kpatrick@ucsd.edu
D. Rosenberg:J. F. Sallis
Department of Psychology, San Diego State University,
San Diego, CA, USA
ann. behav. med. (2011) 42:391–401
DOI 10.1007/s12160-011-9296-7

Page 2

occurrence of a negative health event or general health
concerns and improved fitness or job performance.
Improvements in physical appearance was endorsed but
less often [15–17].
Therefore, the individualized nature of eHealth inter-
ventions, delivered via the Web and other electronic
media, may be particularly useful to improve diet,
physical activity, and sedentary behaviors among men,
for the purpose of weight loss. A recent review found
mixed results of eHealth interventions on these behav-
iors, with effect sizes generally in the small to medium
range [18]. Another review found it difficult to draw
conclusions about the overall effectiveness of Web-based
weight interventions because of the heterogeneity of
designs and approaches [19]. Other internet interventions
produced short-term increases in physical activity [20] and
other health behaviors [21]. Studies, to date, of physical
activity and diet behavior change in the context of weight
loss trials using eHealth technologies, are limited because
they usually assess only short-term outcomes [22] and have
been conducted primarily with women [20, 23]. What
motivates men to change weight-related behaviors may be
different from that of women [13]. One recent randomized
controlled trial among a group of overweight/obese male
staff and students at a university in Australia evaluated an
internet intervention for weight loss and found no differences
between groups at 12 months but did demonstrate that
participants who self-monitored as instructed maintained
greater weight loss than noncompliers [24].
To address the gaps in knowledge noted above and to
advance our understanding of how to improve weight status
in overweight men, we developed and tested the efficacy of
a 12-month Web-based intervention designed for over-
weight and obese men. We hypothesized that, compared to
control participants, men enrolled in the intervention would
show improvements in diet and physical activity behaviors
and weight loss at 1 year. We also hypothesized that weight-
related outcomes would be related to adherence to the
intervention.
Methods
Design
The study was approved by the institutional review boards
of the University of California, San Diego, and San Diego
State University. A total of 441 overweight men signed
consent forms at their baseline visit and were randomly
assigned to either the weight loss intervention or a wait-list
control group. Participants were assessed at baseline, 6, and
12 months. Prior to the 12 month visit, wait-list men were
given the option of crossing over to the intervention for the
full 12 months via an e-mail that was sent out to them
asking if they wanted to cross over to the intervention.
Participants were asked to respond to accept or decline the
option to cross over. If a wait-list participant had not
responded to the initial invitation e-mail, they were asked
over the phone when scheduling their 12-month appoint-
ment. Figure 1 presents the study design.
Participants and Setting
Participating men were 25 to 55 years old (M=43.9, SD=
8.0) with BMI of at least 25 kg/m2(overweight or obese).
Participants were 71% white non-Hispanic, 70% were
married or living with a partner, and two thirds were
college graduates. See Table 1 for a summary of demo-
graphic characteristics. Participants were recruited from the
community from February 2004 through March 2005.
Perhaps reflecting the difficulty in having men address
health issues, this recruitment period was longer than
anticipated and, of the 1,008 men who contacted our
research office, fewer than half (44%) enrolled in the study.
Recruitment strategies included printed advertisements to
four local newspapers, radio advertisements played on three
radio stations over the course of 1 to 2 weeks per radio
station, a TV news story featuring our study, and flyers
placed in local businesses and organizations such as
libraries and medical office waiting rooms. Figure 1 dis-
plays the flow diagram of participants through the study. Of
522 eligible men, 84% (n=441) signed consent forms,
completed the baseline assessment, and were randomized.
Eligible men who were not randomized were more likely to
be younger, not married, and of an ethnic minority
compared to men who did enroll in the study (p values<
0.05). Participants came to the research office for baseline
assessment, randomization, and orientation to the interven-
tion. Simple fixed allocation randomization by computer
assigned participants to the treatment or control group, and
men entered the study within a week of their baseline
measurement visit. Participants completed assessments
remotely via the Web at 6 months and returned to the
research office for the 12-month assessment. The 12-month
assessments were completed by April 2006. Participants
were paid US $20 and $100 for completing the 6- and 12-
month measurement appointments, respectively.
Intervention
The intervention was based primarily on social cognitive
theory [25] and also informed by the behavioral determi-
nants model, an approach that describes the social cognitive
theory related behavioral correlates of exercise [26, 27].
The intervention was designed to influence factors hypoth-
esized to lead to behavior change [28], such as goal setting,
392ann. behav. med. (2011) 42:391–401

Page 3

use of behavioral skills, and increasing social support and
self-efficacy.
The intervention built on a body of previous work by our
group that included print-based physical activity interven-
tions for adult patients [29], primary care-based physical
activity and diet interventions for adults and adolescents
[30–32], and Web-based physical activity and diet inter-
ventions for overweight women [33]. To tailor the current
intervention to the target population of overweight adult
males, interviews with two experts in weight loss inter-
ventions for men and focus groups with men were
conducted to provide feedback on draft intervention content
and the web site. In focus groups, men reported that they
wanted an individualized program that focused on “the
facts” and not their feelings. Focus group participants
indicated that they wanted a to-do list (i.e., “tell me what to
do and I’ll do it”) but also flexibility in being able to choose
which weight-related behaviors to work on instead of being
forced to work on a certain behavior. Participants wanted
professional, quick, and accurate content on the web site
with a link to more detailed information. Participants did
not want a web site with poor navigation, clutter,
unnecessary cuteness, pop-ups, complex menus, or flashing
graphics. Based on this feedback, Web content was created
in short, weekly sessions using “business-like” language
and graphics and little extraneous information. For exam-
ple, weekly goals were set and reported on their “to do” list
online. Focus group participants also endorsed the use of
pedometers to keep track of their physical activity because
they thought it was a novel and interesting device.
Energy density has emerged as a dietary characteristic
that can be manipulated to maintain volume and satiation,
while minimizing the maladaptive response to restriction of
energy intake. Individuals who successfully lose weight and
maintain weight loss, report eating a diet that is lower in
energy density with more vegetables, fruit, and fiber than
diets of the general population [34–36]. Thus, rather than
directly targeting caloric restriction, the dietary goals were
based on improving diet quality and healthfulness. Dietary
goals were also chosen for their contribution to decreasing
risk of chronic diseases, as described in the Dietary
Guidelines for Americans in place at the time of the
intervention [37].
To promote weight loss, the intervention was designed to
improve diet and physical activity behaviors in five areas:
(a) increased fruit and vegetable intake to five to nine or
more servings per day; (b) increased consumption of whole
grain products to more than or equal to three servings per
day; (c) decreased saturated fat intake to ≤20 g per day
through the use of strategies such as substitution, reducing
Fig. 1 Participant flow diagram
ann. behav. med. (2011) 42:391–401 393

Page 4

portion size, decreasing frequency, or changing cooking
methods; (d) increasing steps per day to at least 10,000 on
at least 5 days/week; and (e) strength training at least two
times per week targeting at least two body areas (upper
body, core, lower body). Strength training increases fat-free
mass, and associated increases in energy expenditure may
contribute to reductions in weight over time [38]. Weight
loss interventions are by definition multiple risk behavior
interventions. While it is challenging to intervene on
multiple behaviors, there is evidence that dietary counseling
can be effective for at-risk patients [39]. Through the
intervention, participants were encouraged to start with a
realistic goal in each of the five areas and gradually work
toward meeting the program goals.
The intervention consisted of three components, an
initial computerized assessment to tailor recommenda-
tions for behavioral targets, weekly Web-based learning
activities, and individualized feedback on their progress.
The intervention included theory-based tailoring of
content and was informed by frequent reassessment of
health behaviors, and it offered personalized feedback
[40]. Intervention participants met at the study office with
a “case manager” to orient them to the web site. Case
managers did not provide intervention content but were
available to address technical questions. Participants
completed a computerized assessment of five dietary and
physical activity behaviors and set goals for each. Based
on these results, participants set an initial goal in each area
representing a small improvement over the baseline
assessment. The intervention focused on making small,
incremental improvements over time. Based on our
formative work, participants were allowed to choose what
behaviors to work on each week. Participants were
encouraged, but not required, to take a printed copy of
their goals to their health care provider and to discuss their
goals and importance of weight loss. Over the next
12 months, participants completed weekly Web-based
activities, including learning about and applying theoret-
ically derived behavior change skills and reading about
diet and physical activity topics. The web site included
skill-building tools and physical activity and nutrition
information and tips; a goal setting and reporting page
where goals could be set on the target behaviors; progress
graphs for each of the five behaviors; and relevant news
stories that rotated every few weeks. Personalized graph-
ical feedback was provided weekly and displayed
improvements and instances where behaviors fell below
previously attained levels. Finally, participants had an
opportunity to e-mail a question to our study experts
(dietitian, physical activity expert, clinical psychologist),
Intervention
224
Control
217
Overall
441
Between-group test
Participants (n)
Age in years (mean, SD)
25 to 49 (n, %)
50 to 56 (n, %)
Ethnicity (n, %)
White non-Hispanic
Black non-Hispanic
Hispanic
Asian/Pacific Islander
Native American
Multiethnic
Did not state
Marital status (n, %)
Married or living with partner
Single, separated, divorced
Education level (n, %)
Some HS, HS graduate
Some college
College graduate
Postgraduate training
Body mass index (n, %)
Overweight (25–29.9)
Obesity I (30–34.9)
Obesity II (35–39.9)
Obesity III (>40)
44.9 (7.8)
148 (66.1)
76 (33.9)
42.8 (8.0)
161 (74.2)
56 (25.8)
43.9 (8.0)
309 (70.1)
132 (29.9)
χ(1)2=3.47
p=0.063
χ(2)2=3.47a
p=0.225163 (72.8)
14 (6.3)
34 (15.2)
3 (1.3)
0 (0.0)
5 (2.2)
5 (2.2)
150 (69.1)
9 (4.1)
46 (21.2)
4 (1.8)
2 (0.9)
4 (1.8)
2 (0.9)
313 (71.0)
23 (5.2)
80 (18.1)
7 (1.6)
2 (0.5)
9 (2.0)
7 (1.6)
χ(1)2=1.33
p=0.248163 (72.8)
61 (27.2)
147 (67.7)
70 (32.3)
310 (70.3)
131 (29.7)
χ(3)2=6.50
p=0.090 23 (10.3)
69 (30.8)
56 (25.0)
76 (33.9)
14 (6.5)
57 (26.3)
76 (35.0)
70 (32.3)
37 (8.4)
126 (28.6)
132 (29.9)
146 (33.1)
χ(3)2=0.65
p=0.88438 (17.0)
91 (40.6)
75 (33.5)
20 (8.9)
31 (14.4)
93 (42.9)
74 (34.1)
19 (8.8)
69 (15.6)
184 (41.7)
149 (33.8)
39 (8.8)
Table 1 Demographic and
weight status characteristics
of men randomized
into the study
aEthnicity categorized in three
levels (white non-Hispanic,
Hispanic, other) for the chi-
square test
394 ann. behav. med. (2011) 42:391–401

Page 5

and selected questions and answers would be posted on
the web site for all to see.
Intervention men were given pedometers (Yamax
Digiwalker) to assist in self-monitoring daily steps and
were encouraged to input the data on the web site to
assist with goal setting. Men also reported minutes spent
in physical activities not measurable by a pedometer (e.
g., swimming, cycling, and activities in settings such as
gyms) enabling manual entry of activities unlinked to
actual step counts. The web site converted activity
reports to an equivalent number of steps using an energy
expenditure algorithm developed for this intervention,
based upon standard energy expenditure values [41].
Finally, the case manager had occasional e-mail, and, if
necessary, telephone contact with participants to facilitate
interaction with the web site and troubleshoot technical
difficulties. Because failure to use Web-based interven-
tions has been a limitation of previous studies [18],
prompting participants to use the web site was considered
essential to improve intervention adherence. These calls
did not involve weight-related counseling as the intent was
to keep costs low and test an eHealth intervention.
Men randomized to the wait-list control condition
were given access to an alternate web site containing
general health information of interest to men but not
likely to lead to changes in diet or physical activity
behaviors (e.g., information on stress, hair loss, worksite
injury prevention). At the end of the 12 months, wait-
listed men were given the option to cross over to the
weight loss intervention.
Measures
Assessments were taken by trained assessors who were
blinded to the treatment group.
Body height was measured with an Accu-Hite® wall
stadiometer model 216. Weight was measured using
standard procedures with the digital Body Comp Scale™
from American Weights and Measures. Waist circumference
was measured at the navel with a steel tape measure. Each
measure was taken twice by trained assessors, and the
average of the two readings was calculated. BMI was
calculated as kilograms per square meters. For measure-
ment visits, participants were instructed to wear loose,
comfortable clothing, and additional clothing like jackets,
sweaters, or sweatshirts were removed.
Diet was measured using the 122-item Fred Hutchison
Cancer Research Center Food Frequency Questionnaire at
baseline, 6, and 12 months. This instrument, originally used
in the Women’s Health Initiative, has acceptable measure-
ment characteristics [42], and we used a version of the Food
Frequency Questionnaire appropriate for men previously
used in the Prostate Cancer Prevention Trial [43]. In a
randomly selected subset of 150 men in the PCPT study,
correlations between the Food Frequency Questionnaire,
and six 24-h recalls were between 0.51 and 0.87 for micro
and macro nutrient estimates after adjustment for energy
and de-attenuated for measurement error [44]. Correlations
between Food Frequency Questionnaires were above 0.60
for nutrient estimates. The Food Frequency Questionnaire
has been validated for estimating total vegetable and fruit
intake [45] and includes a measure of alcohol intake. The
four dietary estimates used in the current study included:
percent energy from total dietary fat, percent energy from
saturated fat, fiber grams per 1,000 kcals/day, and servings
for fruits and vegetables per 1,000 kcals/day.
Physical activity was measured using the International
Physical Activity Questionnaire long version, a compre-
hensive assessment of health-related physical activity and
sedentary behavior in adults [46]. The International
Physical Activity Questionnaire assesses daily physical
activity in work, chores, transportation, and leisure using
a reference period of the “last 7 days” to assess frequency
and duration of moderate and vigorous intensity activities
and walking. Additional items inquire about hours of sitting
on weekdays and weekends. In a 12-country study,
International Physical Activity Questionnaires produced
repeatable data (Spearman’s rho clustered around 0.8), and
correlations with accelerometer data had a median rho of
about 0.30, comparable to most other self-reports [46]. The
two estimates of physical activity derived from the
International Physical Activity Questionnaire for the current
study were total walking minutes per day and moderate to
vigorous MET minutes per week.
Intervention dose was estimated by summing the number
of times men logged onto the intervention web site to set a
weekly behavior goal. Since frequency of goal setting was
highly related among the target behaviors (rs ranging from
0.95 to 0.99), one behavior, goal setting for increasing
number of daily steps, was used as an estimate of
intervention dose.
Statistical Analyses
For BMI, body weight, and waist circumference outcomes,
which were measured at baseline and 12-months, analyses
were conducted as intent-to-treat by replacing missing
values at the 12-month end-point with the baseline value.
Tests of between-group differences on these study outcomes
used the 12-month values as dependent variables in the
analysis of covariance (ANCOVA) models adjusting for the
baseline value. For the behavioral outcomes, maximum
likelihood repeated measures models tested between group
differences over time. Analyses were conducted using all
available data at baseline (n=441), 6 months (n=291,
66%), and 12 months (n=309, 70%) assuming data were
ann. behav. med. (2011) 42:391–401395

Page 6

missing at random (MAR). Models were specified with a
between-subject factor of treatment group (1=intervention;
0=control), a within-subject factor of time (0=baseline, 1=
6 months, 2=12 months), and the treatment×time interac-
tion. The significance of the treat×time interaction effect
was tested at 6 and 12 months. Data were also analyzed by
imputing missing data with the baseline observation carried
forward (BOCF) method. The BOCF approach assumes
that weight and behavior have returned to prestudy levels,
which is consistent with the missing not at random
assumption, where discontinuing participation in the study
or not completing assessments may have been due to
regaining weight or not making progress with target
behavior changes. Some have argued that BOCF reflects
the usual recidivism after weight loss. [47].
An exploratory evaluation was conducted of the dose–
response relationship of the intervention on anthropo-
metric and behavioral outcomes in the subsample of
study completers randomized to the intervention group.
ANCOVA models tested if outcomes at 12 months
differed by tertiles of weekly goal setting in the
intervention group. These models adjusted for the
baseline value of the outcome variable. All analyses
were conducted using SPSS version 15.0 (SPSS, Inc.,
Chicago, IL). P values were not adjusted for multiple
tests. All reported p values are for two-sided tests with
effects considered statistically significant at p<0.05. An
anticipated sample size of 215 participants per group
allowed for 20% attrition over the 12-month period and
provided 80% power to detect a standardized effect size of
0.27 or greater.
Results
Treatment groups did not differ statistically by age category,
ethnicity, education level, marital status, or BMI at baseline
(Table 1). However, men in the intervention group were
slightly older than men in the control group (p=0.063).
Forty-five (10%) men withdrew their participation from the
study by 12 months. An additional 87 (20%) men did not
actively withdraw from the study, but either could not be
reached or were not willing to be assessed at 12 months.
Completion of the 12-month assessment (70%) did not vary
by treatment group. However, men who did not complete
the 12-month assessment were younger (mean age=42.5,
SD=8.2) and had a higher BMI (35.2, SD=4.1) compared
to men who completed the 12-month assessment (mean
age=44.5, SD=7.6; BMI=33.7, SD=4.0) (p<0.05). More
white non-Hispanic men (73%) completed the 12-month
assessment compared to Hispanic and other race/ethnicity
men (63%) (p<.05). Fifty-eight percent of men had
complete data at all time points. As a result, age, BMI at
baseline, and white non-Hispanic ethnicity were included as
covariates in the analyses.
Outcome Analyses
Baseline adjusted differences at 12 months were not found
between the intervention and wait-list control groups in the
intent-to-treat analysis for BMI, weight, or waist circum-
ference (see Table 2).
Table 3 displays model-estimated means and standard
errors for the behavioral outcomes at baseline, 6, and
12 months using all available data assuming data were
MAR. All three dietary behaviors had significant group×
time interactions indicating that intervention participants
reduced intake of saturated fat and increased dietary fiber,
fruit and vegetable intakes at 6 and 12 months. BOCF
analyses showed the same results with slightly attenuated
between-group differences (data not shown).
Significant group×time interactions were found for
estimated total walking minutes per day for 6 and
12 months, with men in the intervention group walking
about 16 min/day more than men in the control group at
12 months. In the BOCF analysis, the difference between
groups was still significant at 6 months (p=0.047) but not
at 12 months (p=0.118) (data not shown).
Exploratory Intervention Dose Analysis
The exploratory dose analysis was conducted as a com-
pleter analysis and included 149 of the 154 men who were
assessed at 12 months in the intervention group and had
height and weight measures. These men logged on to the
web site an average of 23.4 weeks (SD=16.7) to set weekly
step goals, and those in the highest tertile set goals an
average of 43 weeks (SD=7.2). Table 4 presents differences
in outcomes for intervention men by tertiles of participation
in weekly goal setting. All three anthropometric measures
showed a statistically significant difference by goal setting
tertile. Men in the highest tertile completed at least 60% of
the goal setting sessions and lost more weight and had
greater reduction in BMI and waist circumference com-
pared to the men in the other goal setting tertiles. For the
dietary variables, men in the third tertile differed from men
in the first two tertiles. Men who used the intervention the
most decreased total fat and saturated fat and increased
dietary fiber, fruit, and vegetable consumption by 12 months
compared to men who used the intervention less frequently.
Additionally, 35% (17) of men in the highest tertile of
goal setting lost at least 5% of their initial body weight by
12-months. This proportion was significantly higher than
proportions in the other goal setting tertiles where 12% (six
and six) of men lost at least 5% of their initial body weight
(χ(2)2=10.8, p<0.001).
396ann. behav. med. (2011) 42:391–401

Page 7

Discussion
In contrast to findings in other studies of internet-based
interventions [22, 23, 34], we did not find a significant
effect of an internet-based weight loss intervention on BMI,
weight, or waist circumference over 1 year. However,
several weight-related diet and physical activity behaviors
were improved through the intervention. In an attempt to
simplify the intervention, dietary recommendations focused
on behaviors (e.g., substitution of low for high-caloric
Intervention Control
OutcomeMean(SE) Mean(SE)P value
Percent energy from fat
Baseline
6 months
12 months
Percent energy from saturated fat
Baseline
6 months
12 months
Fiber grams/1,000 kcal/day
Baseline
6 months
12 months
Servings of fruits and vegetables/1,000 kcal/day
Baseline
6 months
12 months
IPAQ total walking (min/day)
Baseline
6 months
12 months
IPAQ square root MVPA MET (min/week)
Baseline
6 months
12 months
37.43
32.73
33.14
(0.43)
(0.53)
(0.53)
38.00
35.81
36.32
(0.44)
(0.54)
(0.53)
<0.001a
<0.001b
12.03
10.14
10.35
(0.16)
(0.19)
(0.19)
12.31
11.48
11.59
(0.16)
(0.19)
(0.19)
<0.001a
<0.001b
9.05
11.86
11.34
(0.18)
(0.27)
(0.29)
9.01
9.90
9.99
(0.18)
(0.28)
(0.29)
<0.001a
<0.001b
1.31
2.11
2.11
(0.05)
(0.09)
(0.10)
1.31
1.64
1.73
(0.05)
(0.09)
(0.10)
<0.001a
0.002b
61.54
84.75
85.62
(4.38)
(5.16)
(5.38)
61.21
65.31
69.93
(4.46)
(5.36)
(5.39)
0.014a
0.049b
52.51
56.91
57.95
(1.99)
(2.22)
(2.31)
52.24
53.96
53.28
(2.02)
(2.29)
(2.32)
0.375a
0.148b
Table 3 Estimated means and
standard errors from fitted
maximum likelihood repeated
measures mixed models under
the MAR assumption
for missing data
Models include age, baseline
BMI (kilograms per square me-
ter), and white non-Hispanic
ethnicity as covariates. Number
of participants for all models for
Intervention was Baseline=224,
6 months=152, 12 months=
154; and for Control was Base-
line=217, 6 months=139,
12 months=155. All other
group×time interaction tests
were similar between BOCF and
MAR analysis.
IPAQ International Physical
Activity Questionnaire, MVPA
moderate to vigorous physical
activity, MET metabolic equiva-
lent of task
aP value is for group×time
interaction effect at 6 months
bP value is for group×time
interaction effect at 12 months
Table 2 Intervention effects on anthropometric measures at 12 months (intent to treat analyses)
Baseline 12 monthsBetween-group difference at 12 months
adjusted for baseline value (95% CI)
P value
Body mass index (kg/m2)
Intervention
Control
Body weight (kg)
Intervention
Control
Waist circumference (cm)
Intervention
Control
34.2 (4.2)
34.3 (4.0)
33.8 (4.5)
34.2 (4.2)
−0.266
(−0.535, 0.003)
0.053
104.7 (15.3)
104.6 (15.3)
103.8 (16.1)
104.4 (15.4)
−0.694
(−1.52, 0.135)
0.101
113.7 (11.0)
112.9 (11.1)
112.1 (11.8)
111.6 (11.6)
−0.289
(−1.16, 0.585)
0.516
All values expressed as mean (SD). Intent-to-treat analysis (n=441) with missing data imputed by baseline value carried forward. Adjusted
difference and 95% confidence interval from ANCOVA model with 12-month posttest value as the dependent variable adjusting for baseline value,
age, and white non-Hispanic ethnicity
ann. behav. med. (2011) 42:391–401397

Page 8

density foods or reducing portion size) rather than specific
caloric restriction goals. As a result, there were positive
changes in dietary behavior but not enough of a negative
energy balance to result in weight loss or improved BMI.
Nevertheless, these outcomes are encouraging, given the
potential population level reach of eHealth interventions
and their relatively low cost per user compared to more
intensive interventions. This is one of the first randomized
controlled trials of a theory-based eHealth weight loss
intervention developed exclusively for men.
One distinguishing feature of the other studies that did
result in weight loss is they offered more intense interven-
tion, including periodic individualized weight counseling
by trained staff as opposed to intervention staff in the
current study who interacted exclusively to encourage use
of the web site and troubleshoot technical problems.
Personal communication may enhance participant motiva-
tion, skills to change behavior, or engagement with the
intervention. The rationale for relying exclusively on the
Web for intervention in the present study was to keep
implementation costs low and enhance feasibility. Thus, a
promising—albeit exploratory—finding was that those who
completed at least 60% of the recommended weekly goal
setting sessions lost more weight and reduced their BMI
and waist circumference, and were more likely to lose 5%
of their initial body weight compared to men who logged
on less frequently. Weight loss of the more adherent men
(−3.8 kg, SD 6.1) was less than that observed in the
primary outcomes of either of the Tate et al. studies
(−4.1 kg [17] and −4.4 kg [18] or in the Diabetes
Prevention Program (−6.4 kg [48]). However, the observed
difference is still clinically meaningful and was achieved
with a lower intensity intervention, a significantly less
intense intervention compared to the Diabetes Prevention
Program. In the Diabetes Prevention Program, every 1 kg
of weight loss was associated with a 13% reduction in risk
for type II diabetes [49]. Given the low-cost and feasibility
for wide implementation of internet-based interventions, the
effects observed in the present study, while small, could
have benefits on a population level.
Understanding the role of behavior changes in weight
loss is challenging. Reported walking, but not overall
physical activity, was increased by the intervention. This
is not surprising given that the differential increase (16 min/
day), while meaningful, was probably diluted when looking
at physical activity over the course of a whole day. In line
with expectations, the intervention increased fiber, fruit, and
vegetable intakes. However, the overall intervention did not
Table 4 Intervention dose analysis for anthropometric and behavioral measures at 12 months by tertiles of weekly goal setting for intervention
study completers (n=149)
Behavior outcomeWeekly goal setting tertilesF (2, 143) (p value)Partial Eta2
1
(n=50)
2
(n=50)
3
(n=49)
Body mass index (kg/m2)a
33.53
(0.24)
102.78
(0.74)
43.42
(0.33)
33.62
(0.91)
10.53
(0.33)
10.56
(0.58)
1.87
(0.19)
73.03
(8.25)
53.20
(3.59)
33.38
(0.24)
102.00
(0.74)
43.52
(0.32)
34.03
(0.93)
10.91
(0.33)
10.80
(0.57)
1.90
(0.19)
80.23
(8.29)
61.00
(3.60)
32.39
(0.24)
98.76
(0.75)
42.18
(0.32)
30.78
(0.93)
9.26
(0.33)
13.36
(0.58)
2.73
(0.19)
87.92
(8.41)
61.90
(3.64)
6.50
(0.002)
8.01
(<0.001)
5.35
(0.006)
3.62
(0.029)
6.69
(0.002)
7.16
(0.001)
6.46
(0.002)
0.79
(0.455)
1.54
(0.218)
0.08
Body weight (kg)a
0.10
0.08
Waist circumference (cm)a
% energy from fata
0.05
% energy from saturated fata
0.09
Fiber grams/1,000 kcal/daya
0.09
Servings of fruits and vegetables/1,000 kcal/daya
0.08
IPAQ total walking (min/day)0.01
IPAQ square root MVPA MET (min/week) 0.03
All values expressed as mean (SE) at 12 months. Tertiles are of weeks of completed intervention goal setting (1=0 to 12; 2=13 to 30; 3=31 to 53).
ANCOVA models include baseline value, age, baseline BMI, and white non-Hispanic ethnicity as covariates
IPAQ International Physical Activity Questionnaire, MVPA moderate to vigorous physical activity, MET metabolic equivalent of task
aTukey HSD post hoc tests indicated the tertile 3 was significantly different from tertiles 1 and 2
398ann. behav. med. (2011) 42:391–401

Page 9

have a significant effect on weight suggesting that address-
ing diet quality alone is not a promising approach to
promoting major weight loss for most overweight or obese
men because it did not result in enough of an energy deficit.
There is better support for dietary interventions that restrict
energy intake [50, 51], at least in the short-term. However,
the dietary targets utilized in this intervention are charac-
teristic of the dietary pattern associated with higher
likelihood of weight loss in overweight or obese men and
women that is sustained [36, 52]. Although the healthful-
ness of weight loss diets could be enhanced by incorpora-
tion of diet quality goals, there is the possibility that adding
more behavioral goals could be too big a burden for
participants. More research is needed to identify the most
efficient behavioral targets for weight reduction vs. weight
loss maintenance, particularly among men.
An additional contribution of this study was the
finding that it is possible to engage at least some men
in an intervention for weight-related behaviors for an
extended period of time. Many eHealth interventions,
especially low-intensity interventions [53, 54] have
difficulty with adherence over time, yet it is increasingly
understood that successful weight loss and weight loss
maintenance require a long-term commitment to behavior
change [36]. Thus, identifying ways to increase participa-
tion in behaviors that promote weight loss, such as regular
self-monitoring of weight and dietary changes, and weight
loss maintenance, such as habitual physical activity, is a
priority for future research and may well take the form of
the use of a mixture of technologies to increase reach,
novelty, and engagement. An intervention such as this
could be offered as a “first line” of obesity treatment to
men who are unable or unwilling to participate in more
intensive programs. While participation varied in this
study, ongoing participation in the intervention was
essentially the same as other recent studies of Web-based
interventions [18] suggesting there is some promise in
e-mail and telephone prompting by case managers. Men
reported high satisfaction with the pedometers suggesting
that additional technologies may make the intervention
more appealing to them. For example, technologies such
as mobile phones may prove valuable in providing
alternative approaches to behavioral interventions [55]
and their increasing use across a wide range of individuals
suggests that the generalizability of these approaches will
increase in the future.
Study Strengths and Limitations/Future Directions
Limitations of the study include low overall retention and
differential dropout, with Hispanic and more severely obese
men less likely to complete 12-month assessments. It is
possible that men with higher baseline BMIs were more
likely to drop out of a “health-related” study due to
embarrassment about being weighed or the self-perception
of not making adequate progress. Although the proportion
of dropouts did not vary by condition, attrition results limit
generalizability of the findings. However, the two outcome
analysis strategies for handling missing data generated
comparable results, suggesting both efficacy and effective-
ness of the intervention on dietary behaviors. Because of
relatively limited funds to develop the web site, we could
not test whether a state-of-the-art eHealth interface with a
more engaging user experience would have more of an
effect on adherence and, perhaps, behavioral and weight
outcomes. Also, our measure of physical activity was based
upon self-report. Strengths of the study include the
randomized design, a large and diverse sample, validated
behavioral outcome measures, and 12-month outcomes.
Future research should focus on ways to engage users to
participate in the interventions either by use of more user-
friendly web site/technologies, inclusion of incentives to
log on and/or more deliberate caloric restriction resulting
in a negative energy balance. An additional strategy that
would increase the cost of the intervention but could be
weighed against increased effect size is finding the
optimal amount of accompanying individual counseling
(phone or e-mail).
Conclusions
In summary, this study demonstrates that it is possible to
enroll and retain overweight and obese men in a Web-based
weight loss intervention for 1 year, and if they participate,
they can make positive improvements in diet and physical
activity behaviors and weight status. Addressing the public
health crisis of obesity and its consequences will likely
require multilevel interventions and include policy,
environmental, clinical, and family-based approaches
[1]. This study suggests that Web-based approaches may
play a role as well.
Acknowledgments
R01 CA098861). The authors wish to thank Drs. John Foreyt and
Garry Eggers for their consultation on the intervention and design of
this study. Also, the authors wish to acknowledge Kara Mareci for her
contributions to the execution of this study. All human studies have
been approved by the UCSD and SDSU IRBs and have therefore been
performed in accordance with the ethical standards laid down in the
1964 Declaration of Helsinki. All persons gave their informed consent
prior to their inclusion in the study.
This study was supported by the NIH/NCI (RO1
Conflict of Interest
Santech, Inc. (San Diego, CA) which is developing products related to
the research described in this paper. The terms of this arrangement
have been reviewed and approved by San Diego State University and
the University of California, San Diego, in accordance with their
respective conflict of interest policies.
Drs. Patrick, Calfas, and Sallis are co-owners of
ann. behav. med. (2011) 42:391–401 399

Page 10

References
1. U.S. Department of Health and Human Services. The surgeon
general's call to action to prevent and decrease overweight and
obesity. Rockville, MD: USDHHS, Public Health Service, Office
of the Surgeon General, 2001.
2. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ,
Flegal KM. Prevalence of overweight and obesity in the United
States, 1999–2004. JAMA. 2006;295:1549–1555.
3. National Institutes of Health (NIH). Clinical guidelines on the
identification, evaluation and treatment of overweight and obesity
in adults: The evidence report. Bethesda, MD: NIH, 1998.
4. Mokdad AH, Fored, ES, Bowman BA, et al. Prevalence of
obesity, diabetes, and obesity-related health risk factors, 2001.
JAMA. 2003; 289:76–79.
5. Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and
trends in obesity among U.S. adults, 1999–2008. JAMA.;
published online January 13, 2010.
6. Fujioka K. Management of obesity as a chronic disease: Non-
pharmacologic, pharmacologic, and surgical options. Obes Res
2002; 10 Suppl 2:S116–S123.
7. McGuire MT, Wing RR, Klem ML, Lang W, Hill JO. What
predicts weight regain in a group of successful weight losers?. J
Consult Clin Psychol. 1999;67(2)177–185.
8. National Center for Health Statistics. Summary health statistics for
U.S. adults: National Health Interview Survey 2002. Hyattsville,
MD: Public Health Service, 2004.
9. U.S. Department of Health and Human Services. Healthy people
2010: Understanding and improving health (2nd ed.). Washington,
DC: Government Printing Office, 2000.
10. Tsai AG, Wadden TA. Systematic review: An evaluation of major
commercial weight loss programs in the United States. Ann Intern
Med. 2005;142:56–66.
11. Finley CE, Barlow CE, Greenway FL, Rock CL, Rolls BJ, Blair
SN. Retention rates and weight loss in a commercial weight loss
program. Int J Obes (Lond). 2007;31(2):292–298.
12. Douketis JD, Macie C, Thabane L, Williamson DF. Systematic
review of long-term weight loss studies in obese adults: Clinical
significance and applicability to clinical practice. Int J Obes
(Lond). 2005;29:1153–1167.
13. Nicklas BJ. Dietary treatments for obesity. J Gend Specif Med.
1999;2(2):65–71.
14. Egger G, Mowbray G. A qualitative analysis of obestity and at-
risk overweight in working men. Aus J Nutrition Dietetics
1993;50:10–4.
15. Wolfe BL, Smith JE. Different strokes for different folks: Why
overweight men do not seek weight loss treatment. Eating
disorders 2002;10:115–24.
16. Sabinsky MS, Toft U, Raben A, Holm L. Overweight men's
motivations and perceived barriers towards weight loss. Eur J Clin
Nutrition 2007;61:526–31.
17. Hankey CR, Leslie WS, Lean MEJ. Why lose weight? Reasons
for seeking weight loss by overweight but otherwise healthy men.
Int J Obesity 2002;26:880–2.
18. Norman GJ, Zabinski MF, Adams MA, Rosenberg DE, Yaroch
AL, Atienza A. A review of eHealth interventions for physical
activity and dietary behavior change. Am J Prev Med. 2007;33
(4):336–345
19. Neve M, Morgan PJ, Jones PR, Collins CE. Effectiveness of web-
based interventions in achieving weight loss and weight loss
maintenance in overweight and obese adults: A systematic review
with meta-analysis. Obesity Rev 2010;11:306–21.
20. Napolitano MA, Fotheringham M, Tate D, et al. Evaluation of an
internet-based physical activity intervention: A preliminary inves-
tigation. Ann Behav Med. 2003;25(2):92–99.
21. Wantland DJ, Portillo CJ, Holzemer WL, Slaughter R and
McGhee EM. The effectiveness of web-based vs. non-web-based
interventions: A meta-analysis of behavioral change outcomes. J
Med Internet Res. 2004;6(4):e40.
22. Tate DF, Wing RR, Winett RA. Using Internet technology to
deliver a behavioral weight loss program. JAMA. 2001; 285:
1172–1177.
23. Tate DF, Jackvony EH, Wing RR. Effects of internet behavioral
counseling on weight loss in adults at risk for type 2 diabetes.
JAMA. 2003;289:1833–1836.
24. Morgan PJ, Lubans DR, Collins CE, Warren JM, Callister R. The
SHED-IT randomized controlled trial: Evaluation of an Internet-
based weight loss program for men. Obesity 2009;17(11):2025–
32.
25. Bandura A. Social foundations of thought and action. Englewood
Cliffs NJ: Prentice-Hall, 1986.
26. Sallis JF, Hovell MF. Determinants of exercise behavior. In:
Pandolf KB, Holloszy (eds.), Exercise and Sports Sciences
Reviews, Vol. 18. Baltimore, MD: Williams and Wilkins, 1990:
307–330.
27. Sallis JF. Physical activity and behavioral medicine. Thousand
Oaks, CA: Sage Publishers, 1999.
28. Baranowski T, Lin LS, Wetter DW, Resnicow K, Hearn MD.
Theory as mediating variables: Why aren’t community inter-
ventions working as desired? Ann Epidemiol. 1997;S7:S89–
S95.
29. Calfas KJ, Long BJ, Sallis JF, Wooten WJ, Pratt M, Patrick K. A
controlled trial of physician counseling to promote the adoption of
physical activity. Prev Med. 1996;25:225–233.
30. Calfas K, Sallis J, Zabinski M, et al. Preliminary evaluation of a
multi-component program for nutrition and physical activity
change in primary care: PACE+for Adults. Prev Med. 2002;
34:153–161.
31. Patrick K, Sallis J, Lydston D, et.al: Preliminary evaluation of a
multi-component program for nutrition and physical activity
change in primary care: PACE+for Adolescents. Arch Pediatr
Adolesc Med. 2001;155(8):940–946.
32. Patrick K, Calfas KJ, Norman GJ, et al. A randomized controlled
trial of a primary care and home based intervention for physical
activity and nutrition behaviors: PACE+for adolescents Arch
Pediatr Adolesc Med. 2006;160:128–136.
33. Calfas, KJ, Patrick, K, Hagler, A, Norman, GJ, Zabinski, MF,
Sallis, JF. Twelve-month dietary and physical activity outcomes in
‘PACEi-Women in Balance’: A primary-care and web-based
intervention. Ann Behav Med. 2006;31(Suppl):S184.
34. Rolls BJ, Bell EA, Dietary approaches to the treatment of obesity,
Med Clin North Am, 2000, 84(2):401–18
35. de Castro JM. Dietary energy density is associated with increased
intake in free-living humans. J Nutr. 2004 134(2):335–41
36. Wing RR, Hill JO. Successful Weight Loss Maintenance. Ann Rev
Nutrition. 2001;21:323–341.
37. US Department of Health and Human Services, U.S. Department of
Agriculture. Dietary guidelines for Americans 2005. www.health.
gov/dietaryguidelines/dga2005/document/pdf/DGA2005.pdf.
38. Schmitz KH, Jensen MD, Kugler KC, Jeffery RW, Leon AS.
Strength training for obesity prevention in midlife women. Int J
Obes. 2003;27:326–333.
39. Goldstein MG, Whitlock EP, DePue J (2004) Multiple behavioral
risk factor interventions in primary care: Summary of research
evidence. Am J Prev Med 27(S1): 61–79
40. Lustria ML, Cortese J, Noar SM, Glueckauf RL. Computer-tailored
health interventions delivered over the Web: Review and analysis of
key components. Patient Educ Couns 2009;74(2):156–73.
41. Ainsworth BE, Haskell WL, Whitt MC, et al. Compendium of
physical activities: An update of activity codes and MET
intensities. Med Sci Sports Exerc. 2000;32(9):S498–516.
400 ann. behav. med. (2011) 42:391–401

Page 11

42. Patterson RE, Kristal AR, Tinker LF, Carter RA, Bolton MP,
Agurs-Collins T. Measurement characteristics of the Women's
Health Initiative food frequency questionnaire. Ann Epidemiol.
1999;9(3):178–187.
43. Thompson IM Jr., Kouril M, Klien EA, Coltman CA, Ryan A,
Goodman P. The prostate cancer prevention trial: Current
status and lessons learned. Urology. 2001;57(Suppl 1):230–
234.
44. Kristal AR, Arnold KB, Neuhourser ML, Goodman P, Platz EA,
Albanes D, Thompson IM. Diet, supplement use, and prostate
cancer risk: Results from the prostate cancer prevention trial. Am
J Epidemiol 2010;172:566–77.
45. Kristal AR, Vizenor NC, Patterson RE, Neuhouser ML,Shattuck
AL, McLerran D. Precision and bias of food frequency-based
measure of fruit and vegetable intake. Cancer Epidemiol Bio-
markers Prev 2000;9:939–944.
46. Craig CL, Marshall AL, Sjöström M, et al. International physical
activity questionnaire: 12-country reliability and validity. Med Sci
Sports Exerc. 2003;35(8):1381–1395
47. Ware JH. Interpreting incomplete data in studies of diet and
weight loss. N Engl J Med 2003;348(21):2136–7.
48. Diabetes Prevention Program Research Group (DPP) Reduction in
the Incidence of Type 2 Diabetes with Lifestyle Intervention or
Metformin Diabetes Prevention Program Research Group N Engl
J Med 2002; 346:393–403 February 7, 2002
49. Kriska AM, Delahanty LM, Pettee KK. Lifestyle intervention for
the prevention of type 2 diabetes: Translation and future
recommendations. Curr Diab Rep 2004 Apr;4(2):113–118.
doi:10.1007/s11892-004-0066-9
50. Farnsworth E, Luscombe ND, Noakes M, Wittert G, Argyiou E,
Clifton PM. Effect of a high-protein, energy-restricted diet on
body composition, glycemic control, and lipid concentrations in
overweight and obese hyperinsulinemic men and women. Am J
Clin Nutr 2003;78(1):31–39,
51. Dansinger ML, Gleason JA, Griffith JL, Selker HP, Schaefer EJ.
Comparison of the Atkins, Ornish, Weight Watchers, and Zone
diets for weight loss and heart disease risk reduction: A
randomized trial. JAMA 2005;293(1):43–53.
52. Shick SM, Wing RR, Klem ML, McGuire MT, Hill JO, Seagle H.
Persons successful at long-term weight loss and maintenance
continue to consume a low-energy, low-fat diet.J Am Diet Assoc
1998;98(4):408–413.
53. Leslie E, Marshall AL, Owen N, Bauman A. Engagement and
retention of participants in a physical activity website. Prev Med.
2005;40:54–59.
54. Sciamanna CN, Lewis B, Tate D, Napolitano MA, Fotheringham
M, Marcus B. Brief communication: User attitudes toward a
physical activity promotion website. Prev Med. 2002;35:612–615.
55. Patrick K, Griswold W, Raab F, Intille SS. Health and the Mobile
Phone, Am J Prev Med. 2008;35(2) 177–181.
ann. behav. med. (2011) 42:391–401 401