-
Journal of BehavioralMedicine, Vol. 10, No. 2, 1987
Sex Differences in Weight Loss Among
Adults with Type II Diabetes Mellitus
Carma A. Heitzmann, _ Robert M. Kaplan, 2'3Dawn K. Wilson, 2 and
Jeffery Sandier 4
Accepted for publication."November 30, 1985
The treatment of choice for Type II or non-insulin-dependent
diabetes mellitusis a behavioral program for the management of
weight. However, compliancewith this lifelong dietary regimen is
often poor. In the current investigationmale and female adults wRh
diagnosed Type II diabetes were randomly assign-ed to either a
behavior modification, a cognitive modification, a
cognitive-behavior modification, or a control group. Patients were
evaluated in termsof weight, percentage of body fat, and
glycosylated hemoglobin measures.Men lost signficantly more weight
than women and subjects in the behaviormodification group lost more
weight and demonstrated greater decreases indiabetes control than
subjects in the cognitive-behavior modification,cogn#ive, and
control groups. A significant interaction indicated that
diabeticmen may benefit more from behavioral weight reduction
programs thandiabetic women. Several explanations for these
findings are considered.
KEYWORDS:behavior modification; diabetes; sex
differences;weightloss.
This work was supported by Grants K04 HL 00809 and R01 AM 27901
from the National In-stitutes of Health to Robert M. Kaplan.
_Arizona State University, Tempe, Arizona 85287.2Centerfor
BehavioralMedicine,San DiegoState University,San Diego, California
92182.3To whom correspondence should be addressed.4San Diego,
California.
197
0160-7715/87/0400-0197505,00/0 © 1987 Plenum Publishing
Corporation
-
198 Heitzmann,Kaplan,Wilson,andSandier
INTRODUCTION
Adjustment to a chronic disease often requires difficult changes
in life-style. The vast literature on patient compliance indicates
that the morenumerous the changes and the more complex and enduring
the regimen, themore difficulty patients experience in adhering to
their physicians' recom-mendations (e.g., Cohen, 1979; Haynes et
al., 1979; DiMatteo and DiNicola,
1982). Diabetes mellitus requires patients, among other things,
to adopt com-plex nutritional practices, manage their weight, and
monitor their bloodand/or urine glucose levels (Fisher et al.,
1982). Consequently, numerousreports have documented that meeting
self-care requirements is often dif-ficult for these patients
(Watts, 1980).
The specific requirements of diabetes management are dependent
uponthe form of the disease. The two most common types of diabetes
are TypeI or insulin-dependent diabetes mellitus (IDDM) and Type II
or non-insulin-dependent diabetes mellitus (NIDDM) (National
Diabetes Data Group, 1979).
Type I diabetes is usually considered to be the more severe form
of the disease.It typically begins in childhood and is caused by
the cessation of the produc-tion of an adequate supply of insulin
by the pancreas. Patients with TypeI diabetes are dependent upon
exogeneous insulin to preserve life (Fisher etal., 1982; Kaplan and
Atkins, 1985).
Although patients with Type II diabetes are typically not
insulin requir-ing, insulin injections or oral medications are
sometimes prescribed to cor-rect hyperglycemia (abnormally high
blood sugar) (American DiabetesAssociation, 1984). Instead,
patients with non-insulin-dependent diabetes areoften encouraged to
control their disease through a program of decreasedcaloric intake
leading to weight management (Skyler, 1979). In fact,behavioral
management is considered the treatment of choice. The
ClinicalEducational Program of the American Diabetes Association
now advises
physicians to withhold prescriptions of medications until a
serious trial ofdiet and exercise has failed (American Diabetes
Association, 1984; Rifkin,1984). This follows from the
understanding that Type II diabetes is very oftenaggravated by
obesity (Fisher et al., 1982) and that 60-90% of all
non-insulin-dependent diabetic patients are obese. Weight reduction
has been shown toimprove glucose intolerance (Wing et al., 1985).
Diabetes, however, as wellas obesity may differentially appear in
males and females.
Sex Differences
A number of authors have reported sex differences in illness and
mor-
tality rates (Marcus and Seeman, 1981; Wingard, 1984). Although
currentvital statistics do reflect a higher incidence of diabetes
among men, this sex
-
SexDifferencesinWeightLoss 199
ratio has evolved only during the last three to four decades.
Earlier studies(e.g., Wilkerson and Krall, 1947) showed an excess
of female to malediabetics, yet more recent data suggest that this
pattern has reversed. Barrett-
Connor and Wingard (1983) offered three hypothese s to account
for thechanging sex ratio: changes in diabetic survival, parity,
and obesity.
A change in the sex distribution of obesity was the explanation
prefer-red by these researchers. Since Type II diabetes is closely
linked to obesity,it follows that "diabetes is more prevalent in
women in societies where menare leaner than women and more
prevalent in men in cultures where the
reverse is true" (Barrett-Connor and Wingard, 1983, p. 267).
Interestingly,while women have a higher incidence of obesity than
men (Abraham, 1983),
contemporary men are more obese and contemporary women less
obese thantheir predecessors (Barrett-Connor and Wingard, 1983).
Population studieshave shown that the greatest increase in obesity
for men occurs for the agerange 40-59 years. This is the period
when Type II diabetes typically hasits onset (Jeffery et al.,
1984). Because of the role of obesity in thepathogenesis Type II
diabetes, weight loss is a major concern for both malesand females
with this condition.
Behavioral Approaches
The most common educational strategy for the management of
TypeII diabetes is to provide information. However, there is
substantial evidencethat patient knowledge is uncorrelated with
metabolic control (Kaplan et al,,1985; Kaplan and Atkins, 1985,
Marquis and Ware, 1979). In addition, there
are few experimental evaluations showing the efficacy of
education sessionsthat focus on learning factual information.
Instead, diabetic patients mustlearn how to control their
behavior.
Weight Loss
Behavioral and cognitive-behavioral techniques for weight
control, onthe other hand, have produced more impressive results.
Although "the resultsof treatment for obesity are remarkably
similar and remarkably poor"
(Stunkard, 1959, p. 79), behavioral techniques have repeatedly
fared superiorto numerous alternatives (Abramson, 1977; Jeffery et
al., 1978). Examplesof behavioral methods for weight loss that may
be appropriate for obesediabetic adults are described by Mahoney
and Mahoney (1976), Ferguson
(1975b), and Stuart (1971). Typically, behavior modification for
weight lossincludes (1) describing the behavior to be controlled,
(2) identifying the stimuliwhich typically occur prior to eating,
(3) employing behavioral strategies for
-
200 Heitzmann,Kaplan,Wilson,andSandier
controlling eating behavior, and (4) changing the consequences
of eating(Stunkard, 1979). Studies have demonstrated that obese
diabetic adults whosuccessfully lose weight may experience improved
binding of insulin to recep-tors (Archer et al., 1975) and
correcting of hyperinsulinemia (a conditionunique to Type II
diabetes in which there is an excess of insulin in blood)(Farrant
et al., 1969; Pfieffer, 1974). With weight loss, insulin
productionappears to return to normal and there is less insulin
resistance (AmericanDiabetes Association, 1984). With adequate
weight loss, many patients cancease taking insulin or oral
hypoglycemic drugs. For the obese diabetic thereis more than
cosmetic benefit from losing weight. Significant weight loss
maylead to better control over the disease and, consequently, to an
improvedquality of life.
While diet in and of itself may exert an important influence on
weightand diabetes management, the combination of a dietary program
with exer-cise may produce even more promising results (Dahlkoetter
et al., 1979).Despite this, only a few studies have attempted to
modify exercise patternsamong sedentary or chronically ill persons
(e.g., Martin et al., 1984; Atkinset al., 1984). For the adult with
diabetes, consistent moderate exercise maylead to improved
regulation of glucose, reductions in weight, and betteroverall
regulation of the metabolic condition (Kaplan and Atkins, 1985).
Un-fortunately we know relatively little about the benefits of
behavioral weightloss programs for men, since the majority of
studies employing behaviormodification techniques has utilized
either primarily or exclusively femalesubjects [Straw and Terre
(1983), Black and Friesen (1983), and Hagen etal., (1976) provide
recent examples using exclusively women].
Weight control programs which have included men typically have
hadvery low male-to-female ratios, e.g., 11 men to 122 women
(Abrams andFollick, 1983) and 14 men to 46 women (Graham et al.,
1983). Inaddition, and possibly because of the small number of male
participants,data are typically combined for the two groups, a
practice that does not deterresearchers and clinicians from
determining the efficacy of their programs
but does prevent the comparison of men and women in terms of
relativeweight loss. The few studies that have included larger
samples of men haveproduced mixed results. Foreyt et al. (1982)
accumulated data from 154 menand 494 women in a behavioral weight
loss program. They found that men
lost more weight and maintained a greater weight loss than
women. In con-trast, Forster and Jeffery (1984) found no difference
between the sexes forshort-term weight loss. However, in their
study of 55 men and 58 women,women achieved better maintenance of
their weight reduction.
Also scarce are reports examining the weight loss of diabetic
men andwomen. In fact, it is not unusual for researchers to screen
and subsequently
drop from their programs obese individuals who have other health
condi-
-
SexDifferencesin WeightLoss 201
tions such as diabetes, arthritis, and cardiovascular disease
(e.g., James andHampton, 1982). Despite this, it has been asserted
that obese diabetic pa-tients have eating patterns similar to those
of other obese persons.
Implicitly, then, overweight patients with diabetes should
respond tobehavioral self-management techniques (Leon, 1976), and a
recent
study with 23 diabetic patients has provided data to support
this assertion(Rainwater et al., 1982).
Despite the enthusiasm surrounding weight loss programs for
patientswith Type II diabetes, there have been remarkably few
systematic evalua-tions. Wing (1984) noted that no published study
has evaluated the effectsof diabetes weight control programs with a
follow-up of greater than 6
\
months. Wing herself has now reported some follow-up data and is
cautiously
optimistic. Yet there are many different approaches to these
problems andthere is considerable uncertainly about the most
appropriate clinical interven-tion. The behavior therapist has many
options including traditional behaviormodification,
cognitive-behavior modification, and cognitive therapy.
Diabetic patients may enter programs with different motives than
clients intraditional weight management studies. Thus, the
differential effectivenessof these programs for patients with
diabetes must be established. Further,there are few available data
on the impact of these programs upon men, since
male subjects have rarely been included in weight management
evaluations.In summary, Type II diabetes mellitus is a major public
health prob-
lem. Diabetes is the sixth leading cause of death in the United
States (Mor-bidity and Mortality Weekly Report, 1984) and is
associated with severecomplications including blindness, heart
disease, and kidney failure. Weightloss is the treatment of choice
for Type II diabetes (American DiabetesAssociation, 1984; Rifkin,
1984), yet we have few data on the likelihood thatpatients will
comply with a weight loss program.
In order to investigate these questions, we assigned male and
femaleadults with Type II diabetes to different behavioral
strategies for weight loss.These strategies were compared to a
control group receiving only experimenterattention. The patients
were followed over an 18-month period.
METHOD
Subjects
The subjects were 46 Type II, non-insulin-dependent diabetic
patients(22 men and 24 women). The subjects ranged in age from 29
to 79 years (M-- 52.94, SD = 12.08). Two female subjects were black
and all others werewhite Caucasian. There was considerable range in
income and education;
-
202 Heitzmann, Kaplan, Wilson, and Sandier
however, all but three were high-school graduates and the
majority had some
college. The mean body weight of the subjects was 179.79 lb,
with a stan-dard deviation of 35.14 lb (men M = 198.71 women M =
159.69). Althoughnone of these individuals is insulin dependent,
60% use insulin or oralhypoglycemic drugs to control persistent
hyperglycemia. The remaining sub-jects control their disease solely
through diet and exercise. Criteria for beingadmitted to the study
included either fasting blood glucose levels higher than140 mg/dl
or normal fasting blood sugar but oral glucose tolerance tests
thatindicated blood glucose levels exceeding 200 mg/dl at 2 hr
after the ad-ministration of a 75-g carbohydrate dose. These values
were obtained froma referring physician or determined by our
collaborating endocrinologist(J.S.). Each patient was examined by a
board-certified specialist in en-docrinology and metabolism who
classified the case as Type II diabetesmellitus.
Recruitment and Screening
The subjects were recruited in a variety of ways. Many of the
patientswere referred by a local physician specializing in diabetes
care. Patients werealso solicited through public service
announcements made periodically bylocal radio stations.
The diabetic patients taking part in this study were screened
for otherpossible health problems before being allowed to
participate. The first ofthese screening procedures was an exercise
stress test used to diagnose heartdisease. The treadmill test also
provided data on exercise tolerance, restingand exercise heart
rate, and resting and exercise blood pressure. Each testwas given
by a certified cardiovascular technician under the supervision ofa
physician board-certified in internal medicine. The results of
these tests pro-vided a baseline measure of exercise tolerance to
ensure that it was medical-
ly safe for each individual to participate in exercise
regimens.In addition, a doctor's release form was required from
each patient's
personal physician. The physician's release gave consent for the
subject totake part in the study and listed any restrictions upon
the patient's activity.This release form specifically requested
information pertaining to possibleheart problems or other diseases
that might interfere with the patient's fullparticipation in the
program. Those with significant heart or vascular diseasewere
excluded from the study.
Intervention Methods
The interventions employed in this study were designed to
improve com-
pliance to structured exercise and dietary programs. The study
evaluates
-
SexDifferencesinWeightLoss 203
behavior modification, cognitive modification, and
cognitive-behaviormodification programs.
Following pretreatment assessment, subjects were randomly
assignedto one of these experimental groups or to a control group.
Originally, 55subjects were randomized and 44 completed the
18-month follow-up.Although they did report for the 18-month
follow-up, two additional pa-tients provided data for analysis.
Attrition was not significantly differentacross groups. In
addition, equal proportions of men and women finishedthe study.
Computer-generated permutations of the numbers 1 through 4
wereobtained and sealed in envelopes. The envelopes were then
numbered sequen-
tially within groups of four, and the experimenter opened an
envelope cor-responding to when the subject entered the study, for
purposes of treatmentassignment. The numbers 1 through 4 identified
the four groups listed below.The behavior modification group (entry
N = 13, 18-month N = 10) con-
centrated on self-control procedures and participants were asked
to keep dailyrecords of weight, type and amount of food eaten, and
events surroundingeating. They were also asked to monitor their
daily exercise by recording thetime allocated to exercise, the
total time actually spent exercising, and the
place where they exercised. The core of the behavior
modification programwas taken from Ferguson's Habits Not Diets
(1975a).
The cognitive modification group (entry N = 13, 18-month N =
10)discussed the important role that cognitions play in developing
control overbehaviors and positive and negative self-statements.
Subjects were instructedto set reasonable goals for themselves and
to keep a diary of their self-statements .during eating and
exercise. These diaries were then reviewed atlater sessions. Most
of the exercises for the cognitive group were derived from
the chapter on Cognitive Ecology by Mahoney and Mahoney
(1976).The cognitive-behavior modification group (entry N = 15,
18-month
N = 12) was asked to monitor their diet and exercise behavior in
addition
to their self-statements as described above. Subjects in all
three experimen-tal groups, as well as the control group, were
given dietary advice by aregistered nutritionist and a prescribed
exercise regimen based on their exer-cise tolerance tests. Each
group met for seven weekly group sessions. Theamount of time spent
in these sessions was constant across conditions in orderto control
for subject-experimenter exposure time. The average length of
each
group meeting was 1.5 hr. Subjects also received home visits
during whichthe various diet and exercise self-report forms were
presented, as was adescription of the group in which the individual
would be participating. In-cluding the office visit and the home
visit there were nine sessions with eachpatient.
The relaxation control group (entry N = 14, 18-month N = 12)
was
exposed to progressive muscle relaxation training to help them
cope withstress. Psychological stress has been identified as a
factor which may affect
-
204 Heitzmann,Kaplan,Wilson,and Sandier
the status of Type II diabetes. Stress is believed to cause
hormonal changeswhich result in fluctuations in blood glucose
control. Progressive muscle relax-ation is a frequently prescribed
self-management technique to combat stress.A study with obese
female diabetics employing relaxation trainingdemonstrated that
improved glucose tolerance followed such treatment(Surwit and
Feinglos, 1982). Hence, there appears to be both empirical
andtheoretical justification for using a relaxation procedure with
an attentioncontrol group. These individuals were also provided
with factual informa-tion about diabetes but no other interventions
were used. Three relaxation
control subjects participated in the follow-ups but did not
attend the meetings.It is important to emphasize that the
relaxation control group was designedto control for attention and
did not receive the in-depth personal exercisescharacterized by
Surwit and Feinglos (1982).
Follow-ups
Follow-ups were conducted at 3, 6, 12, and 18 months following
theonset of the project. Patients were evaluated on several
physical measuresin addition to structured interviews. These same
measures were employed atthe initial interview.
Physical Measures
Weight. Because men typically weigh more than women and
are,therefore, likely to experience greater weight loss simply
because they areheavier, the Quetelet Index was used as a measure
of relative weight (fora review see Crique et al., 1982). The
Quetetet Index was deemed the mostsuitable tool for epidemiological
studies because of its correlation with ma-jor cardiovascular risk
factors such as age, cholesterol, triglycerides, bloodpressure,
etc. This index is the ratio of weight (lb)/height (in.) 2 x
100.
Glycosylated Hemoglobin. With the subject in a seated position,
5 mlof venous blood was drawn from the anticubital vein. The San
Diego In-stitute of Pathology received the whole blood to determine
glycosylatedhemoglobin according to the Isolab Quick-Sep Kit
QS-9100 method.Glycosylated hemoglobin is believed to be an
indication of average blood
glucose control over the preceding 90 to 120 days (Nathan et
al., 1984). Thus,glycosylated hemoglobin or HbA1 has become a
common indicator ofdiabetes control. The normal range extends to
6.5 °70of the total hemoglobin,
with higher levels indicating progressively poorer control of
diabetes.Percentage of Body Fat. Percentage of body fat is
calculated by
pinching skin folds at the bicep, tricep, subscapular, and
suprailiac regions
-
Sex Differencesin WeightLoss 205
of the body and measuring the skin-fold thickness using Lange
calipers. Thevalues are then summed and a standardized table is
used to convert this figure
into a percentage.
RESULTS
Comparisons between the subjects in terms of general physical
fitnesswere made prior to the interventions. The initial screening
test data wereused for this purpose. Analysis of variance revealed
that the four experimentgroups were not significantly different for
weight, health status, medicineusage, age, percentage fat, or any
other variable prior to treatment (see TableI). Men and women did
not differ significantly on any of the per-formance or biochemical
measures (e.g., predicted maximum heart rate, bloodpressure, etc.).
Although the men weighed more than the women [t(44) =
3.81, p < .01], the Quetelet Index was nearly identical for
the men (M =3.88) and the women (M = 3.7) [F(1,53) = .001, p >
.99]. However, thewomen had a higher initial percentage body fat (M
= 49.60, SD = 13.64)than the men (M = 29.67, SD = 11.56) [t(44) =
4.32, p < .01].
A 2 × 4 (sex by experimental group) analysis of variance using
theQuetelet Index at the initial visit as a covariate (to control
for initial differencesin relative weight among subjects) was
performed to examine changes inweight. Weight (in pounds) at the
18-month follow-up served as the depen-dent variable. There was a
significant main effect for sex of subject[F1,37) = 4.23, p <
.048]. Men lost an average of 8.00 lb, while womengained an average
of .083 lb. There was also a significant main effect forgroup
[F(3,37) = 3.16, p < .038]. A post hoc analysis using the Tukey
Atest indicated that subjects in the behavior modification group
lost significant-ly more weight than subjects in the
cognitive-behavior modification and con-trol groups (p < .05). A
similar analysis of percentage body fat, controlling
TableI. Comparison of Groups at Entrya
WeightQWB HBA_(%) (lb)
Cognitive-behavioral .70 (.06) 11.52 (2.38) 182.44
(34.39)Behavioral .74 (.08) 9.99 (3.04) 184.08 (43.33)Cognitive .74
(.08) 10.17 (2.30) 171.79 (43.11)Control .71 (.08) 10.99 (2.24)
170.00 (29.73)
p (based on ANOVA) .24 .14 .67
aQWB, Quality of Well-BeingScale;HBA_percentageof
glycosylatedhemoglobincells, Entries are means and numbers in
parentheses arestandard deviations.
-
206 Heitzmann,Kaplan,Wilson,andSandier
WEIGHTLOSSFORWOMENBROKEN DOWN BY GROUP
0
-5I-.-
O -10_J
CO -15aZ
-20 ---- BEHAVIORMOD.O _ COGNITIVED.. -25 ..... COG.
BEHAVIOR
_--- CONTROL
-30
-35 i i i i, /,,/ IOmo 3mo 6mo 12mo"-- 18mo
FOLLOW-UP PERIOD
WEIGHT LOSS FOR MENBROKEN DOWN BY GROUP
5 ....
-5
O3
O -10 \ ,,,_. _ .,...,,.o
CO -15 \ /r" "'_",-, /=" \ / ----BEHAV,OR OD.
_/ _ COGNITIVE
O ..... COG. BEHAVIORD.. -25
------ CONTROL
-30
-35 I I I I // I0mo 3mo 6mo 12mo// 18mo
FOLLOW-UP PERIOD
Fig. 1. Interaction of sex and group assignment.
for initial body fat, showed the same effect for group [F(3,37)
= 2.64, p< .05]. However, changes using the Quetelet measure as
a dependent variablewere not statistically significant.
There is also evidence that diabetic men may benefit more
frombehavioral weight reduction programs than women, as suggested
by aborderline significant interaction [F(3,31) = 2.79, p <
.057]. Although thisanalysis does not reach the conventional
significance level, it indicates atrend in this direction and
accordingly another Tukey A test was performedto determine the
specifics of the relationship (p < . 10). The interaction ofsex
and group assignment is shown in Fig. 1. The top portion gives the
datafor women. As the figure demonstrates, all treatment groups
lost more weight
than the control group. However, the differences were not
statistically signifi-
-
Sex Differences in Weigh! Loss 207
cant for women and the rank order of the groups was not
consistent overtime. The bottom portion of Fig. 1 presents data for
men. Men in the behaviormodification group lost a significant
amount of weight while in the program(29 lb) and maintained a
weight loss of 9 lb over 18 months. Weight lossfor the cognitive
group also remained significant through 18 months.However,
differences between the cognitive-behavior modification and
thecontrol groups were not statistically significant for men. When
evaluated forchanges in percentage fat, the interaction was not
statistically significant,however, there was a strong trend in the
same direction as for the data onweight loss.
Differences between groups for HbA1 were not statistically
significant.We have previously reported that changes in
glycosylated hemoglobins lagbehind weight loss (Wilson et al.,
1985). At the 18-month follow-up menhad experienced significantly
greater reductions in HbA1 than women[F(1,37) = 4.31, p < .05].
Although the experimental groups did not differin HbA1 reduction,
there was a significant correlation between weight lossat 18 months
and reduction in HbA1 (r = .32, p < .05).
DISCUSSION
Reviews of the compliance literature often suggest that there is
no rela-tionship between sex of subject and program adherence
(e.g., Marston, 1970;Haynes et al., 1979). The present study, in
contrast, suggests that there aresex differences for weight loss in
patients with Type II diabetes. Men lostsignificantly more weight
than women. These findings are consistent withother studies in
which men lost more weight than women (Stunkard andMcLaren-Hume,
1959). Stuart and Jacobson (1979) suggested that
observeddifferences in weight loss between men and women are often
negated by ad-justments for initial body weight. However, the
differences we observed re-mained statistically significant after
adjustment for obesity (Quetelet Index).These results are also
consistent with other evidence suggestive of male/femaledifferences
in dietary control and weight loss. In one study in which menand
women were asked to make changes in life-style, i.e., alcohol,
smoking,obesity, and drug abuse, men did twice as well as women
overall and, mostimportantly, were more successful than the women
in terms of weight con-trol (Burnum, 1974). Further, a program
designed to improve the complianceof patients on hemodialysis
resulted in a significantly greater number of menthan women who
complied with certain food restrictions (Blackburn, 1977).
Although women in the Stanford Heart Disease Prevention Program
hadmore desirable initial eating behaviors, men responded better to
instructionand were able to make more changes (Stern et al., 1976).
As discussed
-
208 Heitzmann,Kaplan,Wilson,andSandier
previously, authors often do not report results according to sex
of the sub- 'ject. Our data and those cited above suggest that
there may be "hidden dif-ferences" between males and females that
go unreported in the literature.However, in the few available
studies, men seem to be able to achieve greaterweight or dietary
changes than women.
Although it is difficult to account for these male/female
differences
given the present data, several explanations appear plausible.
First, the highermorbidity rate among women (Wingard, 1982)
indicates a greater sensitivityof women to physical symptoms. In
turn, there are a number of physicalsymptoms which are often
related to dieting and weight loss (e.g., fatigue,dizziness,
general weakness, and headache). Therefore, women may be morelikely
than men to discontinue or not comply with a weight control
programsince they are more acutely aware of somatic discomfort.
Some evidence forthis argument is provided by the differences found
between the heart ratesof men and those of women at the conclusion
of the treadmill test. It ap-pears that the men were more likely to
exert themselves to a point where theirheart rates were
substantially increased (mean heart rate = 160), a level ofexertion
typically accompanied by fatigue, weakness, perspiration, etc.
Thewomen, on the other hand, were more likely to discontinue the
test beforereaching this point (M = 152, t = 1.74, p < .10),
thereby minimizing oreliminating any physical discomfort.
The brevity of this treatment program may be partially
responsible forthe absence of weight reduction among female
subjects. Recent weightmanagement studies have suggested that
treatment must extend over a periodof several months, even up to a
year, in order for significant effects to be
elicited and maintained (e.g., Wing et al., 1985). The finding
that the menin this study were able to benefit from only 7 weeks of
intervention may reflectthe specific nature of their weight
problem, namely, that weight loss was en-couraged to manage their
Type II diabetes. The subjects in this program maybe a distinct
group of obese persons whose excess weight poses a greater threatto
their personal well-being than it does to their nondiabetic yet
obese counter-parts. In short, the factors that contribute to
successful weight control maybe somewhat different for the diabetic
individual, and treatment (at least
for men) may not have to be as long since initial motivation is
likely to beheightened.
Length of treatment may also account for the lowered
effectiveness of
the cognitive-behavior therapy program as compared to behavior
modifica-tion alone. The seven behavior modification session were
devoted to a delinea-
tion of self-control principles and an application of these
principles to dietand exercise behaviors. On the other hand, the
combined intervention
necessitated a division of time between behavioral and cognitive
strategies.A significant portion of the sessions was devoted to
instruction in monitor-
-
SexDifferencesin WeightLoss 209
ing and modifying self-statements. Therefore subjects in the
combined groupwere asked to process a greater amount of information
in the same limitedamount of time. It is possible that changes in
overt behavior may be morereadily achieved after short-term
interventions than can modifications inthought processes,
particularly with a highly motivated group for whom theconsequences
associated with not altering life-style patterns are formidable.The
behavior modification group may have provided a sufficient but not
over-whelming amount of information leading to changes in dietary
and exercisehabits.
In summary, male and female diabetic adults appear to differ in
theease with which they lose weight following intervention.
Although reviewsof the compliance literature often conclude that
there are few sex differencesin medicine-taking compliance, our
findings are consistent with several studiesin which males and
females are asked to make life-style changes. While themodest
number of participants in this study may somewhat limit the
generalizability of the results, it is important to note that
the percentage ofmen in this sample is substantially larger than is
found in the majority ofpublished weight control studies. In
addition, very few studies of weightmanagement with chronically ill
patients have been published. A physicalillness, such as diabetes,
may provide a greater stimulus for participation.Whether or not
these results generalize to men who do not have diabetesremains to
be demonstrated.
REFERENCES
Abraham, S. (1983). Obese and overweight adult in the United
States. In Vital and HealthStatistics, Set. 11, No. 230. Natl.
Center Health Stat., DHHS Publ. (PHS) 83-1680.
Abrams, D. B., and Follick, M. J. (1983). Behavioral weight-loss
intervention at the worksite:Feasibility and maintenance. J.
Consult. Clin. Psychol. 51(2): 226-233.
Abramson, E. E. (1977). Behavioral approaches to weight control.
Behav. Res. Ther. 15: 355-363.American Diabetes Association (1984).
Physicians Guide to Type H Diabetes (NIDDM) Diagnosis
and Treatment, ADA, New York.Archer, J. A., Gorden, P., and
Roth, J. (1975). Defect in insulin binding to receptors in
obese
man: Amehoration with calorie restriction. J. Clin. Invest. 55:
166-174.
Atkins, C. J., Kaplan, R. M., Timms, R. M., Reinsch, S., and
Lofback, K. L. (1984). Behavioralexercise programs in the
management of chronic obstructive pulmonary disease.J. Consult.
Clin. Psychol. 52: 591-603.
Barrett-Connor, E., and Wingard, D. L. (1983). Sex differences
in diabetes mellitus. In Gold,E. B. (ed.), The Changing Risk of
Disease in Women: An Epidemiological Approach.D. C. Heath,
Lexington, Mass., Chap. 20.
Black, D. R., and Friesen, J. G. (1983). Deposit money: A
component in a self-directed minimalintervention program for weight
control. Behav. Ther. 14: 333-340.
Blackburn, S. L. (1977). Dietary compliance of chronic
hemodialysis patients. J. Am. Diet.Assoc. 70: 31-37.
Bray, G. A., Davidson,M. B., andDrenick,E. J. (1972).Obesity:A
serioussymptom.Ann.Intern. Med. 77: 779-795.
-
210 Heitzmann, Kaplan, Wilson, and Sandier
Burnum, J. F. (1974). Outlook for treating patients with
self-destructive habits. Ann. Intern.Meal. 81: 387-393.
Cohen, S. J. (1979). New Directions in Patient Compliance, D.C.
Health, Lexington, Mass.
Criqui, M. H., Klauber, M. R., Barrett-Connor, E., et al.
(1982). Adjustment for obesity instudies of cardiovascular disease.
Am. J. Epidemiol. 116: 685-691.
Dahlkoetter, J., Callahan, E. J., and Linton, L. (1979). Obesity
and the unbalanced energy
equation: Exercise versus eating habit change. J. Consult. Clin.
Psychol. 47(5): 898-905.DiMatteo, M. R., and DiNicola, D. D.
(1982). Achieving _atient Compliance: The Psychology
of the Medical Practitioner's Role, Pergamon Press, New
York.Farrant, P. C. Meville, R. W. J., and Stewart, G. A. (1969).
Insulin release in response to oral
glucose in obesity: The effect of reduction in body weight.
Diabetolgia 5: 198-200.Ferguson, J. (1975a). Habits Not Diets,
Bull, Palo Alto, Calif.
Ferguson, J. (1975b). Learning to Eat-Behavior Modification for
Weight Control, Bull, PaloAlto, Calif.
Fisher, E. B., Delameter, A. M., Bertelson, A. D., and Kirkley,
B. G. (1982). Psychologicalfactors in diabetes and its treatment.
J. Consult. Clin. Psychol. 50(6): 993-1003.
Foreyt, J. P., Mitchell, R. E., Garner, D. T., Gee, M., Scott,
L. W., and Gotto, A. M. (1982).Behavioral treatment of obesity:
Results and limitations. Behav. Ther. 13: 153-163.
Forster, J. L., and Jeffery (1984). Gender differences related
to weight history, eating patterns,efficacy expectations,
self-esteem, and weight loss among participants in a weight
reduction
program. Unpublished, University of Minnesota,
Minneapolis.Gabbay, K. H., Hasty, K., Breslow, J. L., and Ellison,
R. C. (1977). Glycosylated hemoglobins
and long-term blood glucose control in diabetes mellitus. J.
Clin. Endocrinol. Metabol.44: 859-864.
Graham, L. E., Taylor, B., Hovell, M. F., and Siegel, W. (1983).
Five-year follow-up to abehavioral weight-loss program. J. Consult.
Clin. Psychol. 51(2): 322-323.
Hagen, R. L., Foreyt, J. P., and Durham, T. W. (1976). The
dropout problem: Reducing attri-tion in obesity research. Behav.
Ther. 7: 463-471.
Haynes, R. B., Taylor, D. W., and Sackett, D. L. (1979).
Compliance in Health Care, JohnsHopkins University Press,
Baltimore.
Heitzmann, C. A., and Kaplan, R. M. (1984). Sex by social
support interaction in diabetesmellitus. J. Consult. Clin. Psychol.
52(6): 1087-1089.
James, J. E., and Hampton, B. A. M. (1982). The relative
efficacy of directive and non-directivetreatment in behavioral
weight control. Behav. Ther. 13: 463-475.
Jeffery, R. W., Wing, R. R., and Stunkard, A. J. (1976).
Behavioral treatment of obesity: Thestate of the art. Behav. Ther.
9: 189-199.
Jeffery, R. W., Folsom, A. R., Luekper, R. V., Jacobs, D. R.,
Gillam, R. F., Taylor, H. L.,and Blackburn, (1984). Prevalence of
overweight and weight loss experience. Am. J. Publ.Health 74:
349-352.
Kaplan, R. M., and Atkins, C. J. (1985). The behavioral
management of type II diabetes mellitus.In Kaplan, R. M., Criqm, M.
H. (eds.), Behavioral Epidemiology and Disease Preven-tion, Plenum,
New York.
Kaplan, R. M., Bush, J. W., and Berry, C. C. (1976). Health
status: Types of validity andindex of well-being. Health Serv. Res.
11(4): 478-507.
Kaplan, R. M., Chadwick, M. W., and Schimmel, L. E. (1985).
Social learning intervention topromote metabolic control in Type 1
diabetes mellitus: Pilot experiment results.Diabetes Care 8:
152-155.
Leon, G. R. (1976). Current directions in the treatment of
obesity. Psychol. Bull. 83: 557-558.Mahoney, M., and Mahoney, K.
(1976). Permanent Weight Control, W. W. Norton, New York.Marcus, A.
C., and Seeman, T. E. (1981). Sex differences in health statistics:
A re-examination
of the nurturant role hypothesis. Am. Sociol. Rev.
46:119-123.Marquis, K., and Ware, J. E. (1979). New measures of
diabetic patient knowledge, behavior,
and attitudes. Rand Corporation, Santa Monica.Marston, M. V.
(1970). Compliance with medical regimens: A review of the
literature. Nursing
Res. 19: 312-323.
-
Sex Differences in Weight Loss 211
Martin, J. E., Dubbert, P. M., Katell, A. D., Thompson, J. K.,
Raczynski, J. R., Lake, M.,Smith, P. O., Webster, J. S., Sikora,
T., and Cohen, R. E. (1984). Behavioral control ofexercise in
sedentary adults: studies 1 through 6. J. Consult. Clin. Psychol.
52:795-811.
Morbidity and Mortality Weekly Report (1984). Center for Disease
Control, April 24.Nathan, D. M., Singer, D. E., Hurxthal, K., and
Goodson, J. D. (1984). The clinical informa-
tion value of the glycosylated hemoglobin assay. N. Engl. J.
Med. 310(6): 341-346.National Diabetes Data Group (1979).
Classification and diagnosis of diabetes mellitus and other
categories of glucose intolerance. Diabetes 28:
1039-1057.Pfieffer, E. F., (1974). Obesity, islet function and
diabetes mellitus. Norm Metab. Res. Suppl.
4: 143-152.
Rainwater, N., Ayllon, T., Fredriksen, I. W., Moore, E. J., and
Bonar, J. R. (1982). Teachingself-management skills to increase
diet compliance in diabetes. In Stuart, R. B. (ed.),Adherence,
Compliance, and Generalization in Behavioral Medicine,
Brunner/Mazel,New York.
Rifkin, H. (1984). The state of the art: Type II diabetes.
Diabetes Forecast 37(4): 30-32, 67.Skyler, J. S. (1979). Diabetes
and exercise: Clinical implications. Diabetes Care 2:
307-311.Stern, M. P., Farquhar, J. W., Macoby, N., and Russell, S.
H. (1976). Results of a two-year
health education campaign on dietary behavior. Circulation
54(5): 826-833.Straw, M. K., and Terre, C. (1983). An evaluation on
individualized behavioral obesity treat-
ment and maintenance strategies. Behav. Ther. 13:
255-266.Stuart, R. B. (1971). A three dimensional program for
treatment of obesity. Behav. Res. Ther.
9: 177-186.
Stuart, R. B., and Jacobson (1979). Sex differences in obesity.
In Gomberg, E., andFranks, V. (eds.), Gender and Disorder,
Brunner/Mazel, New York.
Stunkard, A. J. (1979). Behavioral medicine and beyond: The
example of obesity. In Pomerleau,O. F., and Brady, J. P. (eds.),
Behavioral Medicine: Theory and Practice, Williams &Wilkins,
Baltimore.
Stunkard, A. J., and McLaren-Hume, M. (1959). The results of
treatment of obesity. Arch.Intern. Med. 103: 79-85.
Surwit, R., and Feinglos, M. (1982). Current behavioral research
in Diabetes Mellitus. Presentedat the meeting of the society of
Behavioral Medicine, Chicago, March.
Verbrugge, L. M. (1976). Females and illness: Recent trends in
sex diffeences in the UnitedStates. J. Health Soc. Behav. 17:
387-403.
Watts, F. N. (1980). Behavioral aspects of the management of
diabetes mellitus: Education,self-care and metabolic control.
Behav. Res. Ther. 18: 171-180.
Wilkerson, H. L. C., and Krall, L. P. (1947). Diabetes in a New
England town: A study of3576 persons in Oxford, Mass. JAMA 135:
209-216.
Wilson, D. K., Kaplan, R. M., Heitzmann, C. A., and Sandier, J.
(1985). Correlates of weightreduction for patients with Type II
diabetes mellitus. Diabetes 34: 206A
Wing, R. (1984). Behavioral approaches to weight control.
Presented at the meeting of theAmerican Diabetes Association, Las
Vegas, June.
Wing, R. R., Epstein, L. H., Nowalk, M. P., Koeske, R., and
Hagg, S. (1985). Behavior change,weight loss, and physiological
improvements in Type II diabetic patients. J. Consult.
Clin. Psychol. 53(1): 111-122.Wingard, D. L. (1982). The sex
differential in mortality rates: Demographic and behavioral
factors. Am. J. Epidemiol. 115: 205-216.