Effect of Diet and Exercise, Alone or Combined, on Weight and Body Composition in Overweight-to-Obese Postmenopausal Women




Lifestyle interventions for weight loss are the cornerstone of obesity therapy, yet their optimal design is debated. This is particularly true for postmenopausal women; a population with a high prevalence of obesity yet toward whom fewer studies are targeted. We conducted a year-long, 4-arm randomized trial among 439 overweight-to-obese postmenopausal sedentary women to determine the effects of a calorie-reduced, low-fat diet (D), a moderate-intensity, facility-based aerobic exercise program (E), or the combination of both interventions (D+E), vs. a no-lifestyle-change control (C) on change in body weight and composition. The group-based dietary intervention had a weight-reduction goal of ≥10%, and the exercise intervention consisted of a gradual escalation to 45-min aerobic exercise 5 day/week. Participants were predominantly non-Hispanic whites (85%) with a mean age of 58.0 ± 5.0 years, a mean BMI of 30.9 ± 4.0 kg/m2 and an average of 47.8 ± 4.4% body fat. Baseline and 12-month weight and adiposity measures were obtained by staff blinded to participants' intervention assignment. Three hundred and ninety nine women completed the trial (91% retention). Using an intention-to-treat analysis, average weight loss at 12 months was −8.5% for the D group (P < 0.0001 vs. C), −2.4% for the E group (P = 0.03 vs. C), and −10.8% for the D+E group (P < 0.0001 vs. C), whereas the C group experienced a nonsignificant −0.8% decrease. BMI, waist circumference, and % body fat were also similarly reduced. Among postmenopausal women, lifestyle-change involving diet, exercise, or both combined over 1 year improves body weight and adiposity, with the greatest change arising from the combined intervention.


The prevalence of obesity has reached epidemic levels over the past few decades, and concurrent with this rise are increases in numerous obesity-associated diseases including heart disease, certain types of cancer, and diabetes (1). The cornerstone of therapeutic interventions to treat or prevent these diseases is weight loss via lifestyle modification, including hypocaloric diet and/or increased physical activity along with behavioral techniques to support these changes (2). Typical weight loss resulting from lifestyle change is between 5 and 10% of baseline weight, so such approaches rarely bring an obese individual to a normal body weight (3). However, losing even this modest amount of weight brings health benefits (4). Multiple nonrandomized interventions have demonstrated improvements in biomarkers relating to diabetes, cardiovascular disease, and cancer risk (3,5). Smaller-scale randomized studies of lifestyle change to induce weight loss have shown improvements in hypertension and metabolic syndrome (6,7,8). The Diabetes Prevention Program (DPP) randomized over 3,000 individuals at high risk for developing diabetes to an intensive, individual-focused combined diet-and-exercise lifestyle intervention vs. pharmaceutical therapies or placebo and demonstrated significant reductions in diabetes incidence. Particularly, those in the lifestyle intervention arm experienced the greatest (58%) decrease in disease incidence over 3 years, providing strong evidence for the efficacy of the lifestyle program (9).

Given the health impacts of the obesity epidemic and the research suggesting that weight loss can ameliorate these problems, there have been numerous calls for optimal obesity treatment strategies. The NIH Obesity Education Initiative Expert Panel suggested a caloric deficit of 500–1,000 kcal/day using an individualized dietary strategy, along with 45 min of moderate-intensity physical activity 5 days/week (5). The Institute of Medicine recommended at least 1 h/day of moderately intense physical activity coupled with a caloric deficit whereas the US Department of Agriculture similarly suggests individuals engage in close to 1 h of moderate-to-vigorous intensity exercise on most days of the week, without exceeding caloric intake requirements (3,10). The US Center for Disease Control instead suggests at least 30 min/day of moderate-intensity exercise most days of the week while maintaining sensible portion sizes (11). The recommendations from these agencies are all based on expert opinion supported by the available data. Although numerous studies have examined the effect of lifestyle interventions on body weight, few have focused on postmenopausal women, a group experiencing particularly high rates of overweight and obesity using a tested intervention such as that undertaken by the DPP, a program known to reduce disease risk in other populations, and measured effects over a 12-month period (12). Hence, questions still remain regarding the best approach for weight loss from lifestyle change for this group in particular, and randomized, controlled studies are the best way to demonstrate effectiveness of interventions that could influence public health recommendations.

We conducted a year-long, group-based lifestyle intervention to examine the independent and combined effects of physical activity and dietary weight loss on body weight as well as body composition among 439 sedentary, overweight/obese postmenopausal women. We hypothesized that the women randomized to the intervention groups would experience greater weight loss and improvements in body composition parameters than those randomized to the control group (who maintained their usual lifestyle), and furthermore that weight loss and changes in body composition would vary among the intervention groups based on the separate and combined impacts of diet and exercise on overall energy balance.

Methods and Procedures

The Nutrition and Exercise in Women (NEW) study, conducted from 2005 to 2009, was a 12-month randomized, controlled trial using a 4-arm design to compare the effect of three lifestyle-change interventions (moderate-to-vigorous intensity aerobic exercise, dietary weight loss, or both interventions combined) vs. control (no lifestyle change) on body weight and composition. All study procedures were reviewed and approved by the Fred Hutchinson Cancer Research Center (FHCRC) institutional review board in Seattle, WA, and all participants provided signed informed consent.

Participant recruitment, inclusion, and exclusion criteria

The target population for the NEW study included postmenopausal women from the greater Seattle, WA area, aged 50–75 years, who were overweight or obese (BMI ≥25 kg/m2, or ≥23 kg/m2 for Asian-American women), and exercising <100 min/week at moderate intensity or greater. Specific exclusion criteria included: diagnosed diabetes, fasting blood glucose ≥126 mg/dl, or use of diabetes medications; use of postmenopausal hormone therapy within the prior 3 months; history of breast cancer or other serious medical condition(s); alcohol intake in excess of 2 drinks/day or current smoker; contraindication to participating in the diet or exercise intervention for any reason, including an abnormal exercise tolerance test, current or planned participation in another structured weight loss program, use of weight loss medications, or additional factors that might interfere with measurement of outcomes or with the success of the intervention (e.g., inability to attend facility-based sessions).

Women were recruited through targeted mass mailing campaigns and media publicity or community outreach (Figure 1). Invitation letters were sent to 126,802 age-eligible women and 5,621 responded; in addition we received 2,048 media and community outreach-prompted calls. 929 women were telephone screen-eligible, a total of 703 women attended the information session, 684 women were screened in clinic, and 439 were randomized into the study (∼80% from mass mailings, 20% from media and community outreach).

Figure 1.

Flow diagram of the progress through the NEW trial. DXA, dual X-ray absorptiometry; VO2max, cardiorespiratory fitness test.

Study design and randomization

The study design for the NEW trial is shown in Figure 1. Following baseline data collection, eligible women were randomized into one of four study arms: (i) dietary weight loss (D, N = 118); (ii) moderate-to-vigorous intensity aerobic exercise (E, N = 117); (iii) both interventions combined (D+E, N = 117); or (iv) no-lifestyle-change control (C, N = 87). The random assignment was generated by a computerized program, stratified according to BMI (<30 kg/m2 or ≥30 kg/m2) and participants' self-reported race/ethnicity (non-Hispanic white, black, or other). In addition, to achieve a proportionally smaller number of women assigned to the control group, a permuted blocks randomization with blocks of four was used, wherein the control assignment was randomly eliminated from each block with a probability of ∼1 in 4.

Lifestyle-change interventions

The NEW dietary weight-loss intervention comprised our modification of the dietary component of the DPP (9) and the Look Action for Health in Diabetes (AHEAD) (13) lifestyle intervention programs, with the following goals: total daily energy intake of 1,200–2,000 kcal/day based on baseline weight, <30% daily energy intake from fat, and a 10% reduction in body weight by 6 months with maintenance thereafter to 12 months. Content of the dietary counseling sessions was modified to better fit our study population (less focus on diabetes or diabetes risk), and the frequency and type of sessions (individual vs. group) also varied from DPP and Look AHEAD. Women met individually with a study dietitian for personalized goal-setting on at least two occasions, followed by weekly meetings in groups of ∼5–10 women, through the first 6 months. Thereafter (months 7–12), dietitians had contact with participants twice a month, including one face-to-face contact (individual or group session) and one additional contact via phone or email. Participants were permitted additional in-person sessions, phone, or email contacts beyond the 32 expected, if they or the dietician felt these would help to achieve intervention goals. This combination of individual and group-based approaches was used to maximize the benefits of targeted, personalized recommendations along with the social support and greater cost-effectiveness of a group setting. Women were asked to record all food eaten daily for at least 6 months, or until they reached their individual weight loss goal (10%). Food journals were collected by the dietitian and returned with feedback. Journaling, weekly weigh-ins, and session attendance were tracked to promote adherence to the diet intervention.

Based on our previous exercise research in a similar population, the goal of the NEW exercise intervention was ≥45 min of moderate-to-vigorous intensity exercise, 5 days/week (225 min/week) for 12 months (14,15). Participants attended at least three sessions/week at our study facility where they were supervised by an exercise physiologist, and exercised for their remaining sessions at home. The exercise training program began with a 15 min session at 60–70% maximal heart rate (determined by baseline exercise treadmill testing) and progressed to the target 70–85% maximal heart rate for 45 min by the 7th week after enrollment where it was maintained for the remainder of the study. Women wore Polar heart rate monitors (Polar Electro, Lake Success, NY) during facility exercise sessions to assist with attaining their target heart rate. In addition, during both facility and home sessions they recorded the mode and duration of exercise, and peak heart rate achieved. Facility-based exercise consisted of treadmill walking, stationary bicycling, and use of other aerobic machines; while a variety of home exercises were encouraged including walking/hiking, aerobics, and bicycling. A small amount of resistance training to strengthen joints and limit injury was recommended, though not required. Activities of at least four metabolic equivalents according to the Compendium of Physical Activities (16) such as brisk walking were counted toward the prescribed aerobic exercise target. Activity logs were reviewed weekly by study staff in order to monitor adherence. Participants who were not meeting exercise targets were contacted by staff to discuss barriers and approaches to increase activity. In addition, the dietitians and exercise physiologists met regularly with a clinical health psychologist experienced in lifestyle behavior change to discuss participant progress and refine behavior modification goals according to each participant's needs.

Women randomized to the diet + exercise group received both the dietary weight loss and aerobic exercise interventions. They participated in separate groups for the dietary weight-loss intervention from women assigned to diet alone. Although the diet + exercise group could use the exercise facility at the same time as participants assigned to the exercise-only group, they were instructed not to discuss the diet intervention. Women randomized to the control group were requested not to change their diet or exercise habits for the duration of the trial. At the end of 12 months, participants in the control group were offered four group nutrition classes and 8 weeks of facility exercise training with individualized guidance from an exercise physiologist, as an incentive to undergo randomization.

Study measures and data collection

All study measures were obtained by trained study personnel who were blinded to the participants' randomization status.

Participants completed a series of questionnaires at their baseline screening visit before randomization, including demographic information, medical history, health habits, reproductive and body weight history, diet intake (via a validated 120-item self-administered Food Frequency Questionnaire (FFQ)) (17), as well as frequency, duration, and intensity of physical activity over the preceding 3-month period (via an interview-administered Minnesota Physical Activity Questionnaire) (18). These same questionnaires were completed at the end of the 12-month study period.

Anthropometric measures were obtained at baseline and again at 12 months, whereas participants wore a hospital gown without shoes. Weight and height were measured to the nearest 0.1 kg and 0.1 cm, respectively, with a balance beam scale and stadiometer. Both measures were made in duplicate and averaged. BMI was calculated as weight in kilograms divided by the square of height in meters. Waist and hip circumferences were measured in duplicate to the nearest 0.5 cm using a fiberglass tape, and averaged. Waist circumference was obtained at the end of normal expiration in the horizontal plane at the minimal waist, whereas the hip measure was obtained at the maximum point also in the horizontal plane. Total and percentage body fat and lean mass were measured using a dual X-ray absorptiometry whole-body scanner (GE Lunar, Madison, WI) with participants in a supine position.

Adherence to the diet and exercise interventions was assessed via multiple approaches. The dietary intervention was evaluated by weight loss, and monitoring frequency of session attendance (19), food journal completion and submission, and self-weighing. These measures were reviewed weekly by study dieticians. Exercise participants tracked their activities in a daily exercise log that was reviewed on a weekly basis by the exercise physiologists. All participants completed certain measures at baseline and 12 months to assess differences and change across study arms. FFQs were completed primarily to assess dietary patterns and change in fat intake. Participants wore pedometers (Accusplit, Silicon Valley, CA) while awake for 7 consecutive days in order to determine an average daily step count. Cardiorespiratory fitness (VO2max) was assessed using a maximal graded treadmill test according to a modified branching protocol starting at 3.0 mph and 0% grade with incremental increases in speed or grade every 2 min (20,21). Heart rate and oxygen uptake were continuously monitored with an automated metabolic cart (MedGraphics, St Paul, MN).

Statistical analyses

Our primary outcome was the comparison of change in body weight by study group from baseline to 12 months. Secondary outcomes included the comparison by study group of the 12-month change in body composition as well as additional anthropometric measures. With 439 women enrolled in the NEW trial, we had over 80% power to detect a 3.16 kg difference between groups, using a Bonferroni adjustment for the multiple comparisons (two-sided α = 0.05/6), and assuming a 10% drop-out rate/5% drop-in rate. We experienced 91% retention at 12 months, with 399 out of the 439 women randomized returning for final assessments (Figure 1). All analyses were based on the assigned intervention at the time of randomization regardless of adherence (i.e., intention-to-treat), with a conservative no-change-from-baseline imputation made for the missing values.

We computed the mean change in body weight, anthropometric, and body composition measures from baseline to 12 months and assessed the differences between each of the intervention and control groups. As these outcome measures were normally distributed, data transformation was not necessary. We used the generalized estimating equations modification of linear regression to account for the longitudinal nature of the data and correlation within individuals. As a secondary analysis, we compared the mean 12-month changes in outcomes by tertiles of adherence measures. Additional descriptive data are presented as mean ± s.d., unless otherwise noted. All statistical analyses were performed using SAS software version 9.1 (SAS Institute, Cary, NC).


The baseline characteristics of the 439 participants randomized to the NEW trial are shown in Table 1. Women were on average 58.0 ± 5.0 years old, obese (BMI 30.9 ± 4.0 kg/m2) with a high percent body fat (47.2 ± 4.3%), and had poor cardiorespiratory fitness (VO2max 22.9 ± 4.0 kg/ml/min). Approximately 85% of participants were non-Hispanic white, 65% were college graduates, and about half were full-time employed. With the exception of percent fat intake in the diet estimated by FFQ, there were no statistically significant differences in any demographic or lifestyle variables between the study groups at baseline. Forty women did not complete the study (D = 13, E = 11, D+E = 9, C = 7), with the primary reasons for study discontinuation being: dissatisfaction with randomization, work or family demands, or injury or other medical issues unrelated to the intervention. At 12 months, 399 participants completed follow-up visits for body weight and anthropometric assessments, 396 underwent a dual X-ray absorptiometry scan, and 370 participants completed a maximal treadmill test (Figure 1).

Table 1.  Baseline characteristics of NEW trial participants
inline image

Overall, adherence to the dietary weight loss and aerobic exercise interventions was excellent (Table 2). Estimated relative fat intake (percentage of total kcal/day) decreased by 18% in the diet-alone group and by 20% in the diet + exercise group. Changes in total kcal intake among D and D+E groups did not change significantly compared to controls, however the FFQ is not as robust for estimating this parameter (17). In both the diet-alone and diet + exercise groups, women attended an average of 27-diet behavior change sessions (86%). Women randomized to exercise alone achieved on average 80% of the target 225 min/week aerobic exercise over the 12-month trial, whereas women randomized to diet + exercise achieved 85%. Both groups significantly increased their average pedometer steps/day (E: 2,354 ± 2,749 steps/day, 42% increase; D+E: 3,408 ± 3,001 steps/day, 58% increase) and VO2max (E: 0.16 ± 0.36 l/min, 9% increase; D+E: 0.12 ± 0.34 l/min, 7% increase), compared to baseline.

Table 2.  Baseline and 12-month body weights, anthropometric measures, and body composition measures by intervention group
inline image

Body weight, anthropometric, and body composition changes were evaluated at 12 months and are presented in Table 2. Participants in the diet-alone group lost a mean 7.2 kg (−8.5%, P < 0.0001), those in the exercise alone group lost a mean 2.0 kg (−2.4%, P = 0.034), whereas those in the diet + exercise group lost a mean 8.9 kg (−10.8%, P < 0.0001), each compared to a 0.7 kg decrease among controls. The weight reductions for the diet-alone and diet + exercise groups both were significantly greater than that experienced by the exercise alone group (both P < 0.0001), though the difference between diet-alone and diet + exercise did not reach the adjusted level of statistical significance (P = 0.02). Waist circumference decreased significantly in all intervention groups compared to the nonsignificant 0.9 cm increase among controls (D: −4.5 cm P < 0.0001; E: −2.0 cm P = 0.001; D+E: −7.0 cm P < 0.0001). Furthermore, the decrease in waist circumference was significantly greater for the diet + exercise participants than for either the diet-alone or exercise-alone participants (P = 0.004 D+E vs. D and P < 0.0001 D+E vs. E), and women who were assigned to diet alone also experienced significantly greater decrease compared to those assigned to exercise alone (P = 0.004). Similarly, % body fat decreased among all intervention groups compared to controls, with an absolute percent change of −4.2% for diet alone, −1.6% for exercise alone, and −5.9% for diet + exercise (all P < 0.0001). Paralleling the findings for waist circumference, comparing these reductions among intervention groups revealed the rank-order for the reduction was also D+E > D > E (all P < 0.005). Finally, lean mass increased only among those participating in exercise alone (+0.3 kg). Although this increase did not differ significantly from control women, it was significantly greater compared to the reductions in lean mass among women in the diet alone (−0.8 kg, P < 0.0001 E vs. D) or diet + exercise (−0.4 kg, P = 0.003 E vs. D+E) groups.

We assessed measures of adherence for the combined diet + exercise group separately from either intervention alone. Changes in outcome measures at 12 months stratified by tertiles of percentage of in-person diet intervention session attendance are shown in Table 3 (D group) and Table 4 (D+E group). Women from the diet-alone group who participated in ≥106% of sessions (included optional visits) lost the most weight, with an 11.4 kg (13.5%) decrease, compared to losses of 8.3 kg (10.4%) among women attending 84–106% of sessions, 5.5 kg (6.3%) among those attending <84%, or a 0.5 kg loss (0.6%) among controls (all P trend <0.0001). Comparable findings were noted for BMI, waist circumference, and % body fat (all P trend ≤0.0001, Table 3). Similarly to the diet-alone participants, women from the diet + exercise group who attended the most in-person sessions also lost the most weight, with a 10.9 kg (13.4%) decrease, compared to losses of 9.4 kg (11.5%) among women attending 84–106% of sessions, and 8.1 kg (9.6%) among those attending <84% (all P trend <0.0001). Equivalent changes were observed for other all outcome measures (Table 4).

Table 3.  Baseline and 12-month weight, anthropometric, and body composition measures for diet group, stratified by adherence
inline image
Table 4.  Baseline and 12-month weight, anthropometric, and body composition measures for diet + exercise group, stratified by adherence
inline image

Changes in outcome measures at 12 months stratified by tertiles of reported exercise (exercise logbook min/week) are shown in Table 5 (E group) and Table 6 (D+E group). Women from the exercise-alone group who participated in ≥196 min/week physical activity (highly active) lost the most weight, with a 3.1 kg (3.9%) decrease, compared to losses of 2.0 kg (2.4%) among intermediate-active women (154–196 min/week), 1.9 kg (2.2%) among low-active women (<154 min/week) or a 0.5 kg loss (0.6%) among controls (P trend <0.0001 in reference to controls, P = 0.12 in reference to low-active group). Comparable findings were noted for BMI, waist circumference, and % body fat (P trend all <0.01 in reference to controls, Table 5); in addition these same trends were observed when adherence was assessed by change in cardiorespiratory fitness level or change in pedometer steps/week (data not shown). Highly active women from the diet + exercise group lost 10.5 kg (12.8%) decrease, similar to the weight loss achieved by active-intermediate women (10.5 kg or 12.7%), whereas low-active women lost 8.0 kg (9.6% all P trend <0.0001 in reference to controls, and P trend ≤0.01 in reference to low-active group). Equivalent changes were observed for other outcome measures (Table 6), and also when adherence was assessed by change in cardiorespiratory fitness level or change in pedometer steps/week (data not shown).

Table 5.  Baseline and 12-month weight, anthropometric, and body composition measures for exercise group, stratified by adherence
inline image
Table 6.  Baseline and 12-month weight, anthropometric, and body composition measures for diet + exercise group, stratified by adherence
inline image


We found that a year-long lifestyle-change program, incorporating either combined or separate dietary weight loss or moderate-to-vigorous aerobic exercise interventions, produced clinically important and significant reductions in body weight and improvements in body composition among overweight-to-obese postmenopausal women. Our dietary intervention was adapted primarily from the DPP for group-based delivery, whereas our exercise intervention was based on our previous successful physical activity programs (14,15). With excellent overall follow-up (>90%) and adherence to the interventions, we observed an 8.5% weight loss among women participating in diet alone, 2.4% weight loss among those participating in exercise alone, and 10.8% weight loss among those in the combined diet + exercise interventions. Furthermore, we observed that the relative reductions in % body fat measured by dual X-ray absorptiometry in each group differed significantly (at a Bonferroni-adjusted level of P < 0.05/6 to account for the multiple comparisons) with the following rank order: −12.4% (D+E) > −8.9% (D) > −3.3% (E) > −0.3 (C).

Our combined diet + exercise group effectively put to the test the recommendation of the NIH Obesity Education Initiative Expert Panel, which suggested a caloric deficit of 500–1,000 kcal/day using an individualized dietary strategy, along with 45 min of moderate-intensity physical activity 5 day/week. Most importantly, we tested this among postmenopausal women, for whom well-designed studies examining the longer-term benefits of lifestyle change are needed (12). When we examined the percentage of participants from our trial who achieved various clinically relevant weight loss targets (<5%, 5– <10%, and ≥10%), the majority (60%) of women in the diet + exercise intervention achieved ≥10%, whereas only 42% of those in diet alone, and 3% in exercise alone achieved that target (data not shown). Although adding 5 day/week of exercise requires effort, 20% more women attained the very relevant goal of 10% weight loss when they added exercise to diet modification. It is likely that the cost of exercise is offset by the cost savings due to health benefits arising from increased physical activity (22). Furthermore, following a recent review of studies examining exercise for weight loss the American College of Sports Medicine has now increased its recommended dose of exercise to 150–250 min/week, suggesting that particularly when exercise is combined with moderate caloric restriction these levels are necessary to augment weight loss (23). Similarly, the US Department of Health and Human Services Physical Activity Guidelines Committee found that optimal exercise levels for weight loss and weight loss maintenance could be up to 60 min/day (24).

Other research has demonstrated comparable degrees of weight loss to those we found with our interventions, though studies vary widely with respect to type of diet and exercise interventions, the inclusion of a control group, their duration, size, and measures of potential greater metabolic relevance such as body fat and lean mass (25,26,27,28,29,30). The MONET study also targeted postmenopausal women (N = 173), but only examined diet vs. diet + resistance training over 6 months, and found equivalent weight loss between these groups (31). We found that exercise had an important effect of maintaining lean body mass that did not occur with diet alone, and this effect could certainly have important implications as older people are at increased risk for sarcopenia (32). Although it would be difficult to determine the relative impact of different diets and doses of exercise within a single study, it has been shown that most diets are effective regardless of composition as long as participants are adherent (33,34). We found adherence, whether to the diet or exercise intervention, had a dose-dependent impact on weight loss and change in body composition among our population of postmenopausal women.

The degree of adherence to lifestyle change has been clearly shown to positively influence all weight and body composition-related outcomes. We found that greater adherence to our exercise intervention, whether measured using a self-reported variable of minutes exercised per week (in relation to the goal of 225), or more objective measures such as pedometer-recorded steps per week or cardiorespiratory fitness (measured by an exercise tolerance test), was associated with greater weight reduction or improvement in body composition. Greater adherence to the exercise prescription in both the DPP and the Look AHEAD trial also predicted greater rates of achieving weight loss goals (19,35). The same relationship with adherence and weight loss was found when we examined in-person session attendance for our diet intervention. The level of adherence overall was very high (over 90% attendance for both D and D+E), but we were still able to see greater improvement in body weight and composition outcomes among women in the highest adherence groups. Alhassan et al. found that adherence to diet was a key predictor of weight loss in the “A to Z” study of different popular diets, similar findings were also reported by Greenberg et al. for a randomized low fat vs. low carbohydrate diet, and again for the DPP and Look AHEAD (19,35,36,37). Further research is warranted to improve our understanding of behavioral and biological determinants of adherence, particularly those that could be modifiable.

One limitation of our study was that exercise performed at home was self-reported, in comparison with that directly observed at our facility. Although facility sessions allowed for direct observation of participants' activity, the applicability of our findings to less-intensive or costly community programs to promote physical activity is not as clear. Still, the significant increases in VO2max observed only among women randomized to the exercise groups (E and D+E) support the internal validity of our data, and that these women did raise their physical activity level. Diet was also self-reported by women in the dietary weight loss arms, and although the FFQ is a valid measure for assessing diet, it has acknowledged limitations. The significant weight loss experienced by women in the diet interventions (nearly 9% and 11% in D and D+E, respectively) again supports that these women did reduce their caloric intake. Another potential factor which may have introduced some additional variability to the weight loss achieved by the study groups is contact time with study staff. Study contact time may impact weight loss (38), and it varied to some degree by intervention group and was the least, by design, for women in the control group.

Strengths of our study included our large study size with sufficient power to examine differences not only in comparison to a no-lifestyle-change control group, but also relative to differences among the elements of lifestyle change (diet and exercise, alone and in combination). We examined the impacts of our interventions on weight, but also on body fat measured by dual X-ray absorptiometry. Body fat does not increase linearly with body weight, and measuring body fat directly may allow for better estimation of health impacts (39). Our study population consisted of postmenopausal women, and our focus on this group was purposeful, given their particularly high rates of weight gain and obesity (1,40). Furthermore, we experienced excellent adherence to the interventions, and low drop-out rates. Finally, in contrast to numerous published short-term studies examining the impacts of diet and exercise, our study was a full year in duration. Ongoing, long-term follow-up of participants is now underway, and factors contributing to the successful maintenance of weight loss will be examined in the future.

In conclusion, we have demonstrated a successful implementation of a group-based modification of the DPP dietary weight loss intervention, along with moderate-vigorous aerobic physical activity, among a population at high risk for ongoing weight gain and the negative metabolic consequences thereof. Although our diet and exercise interventions had beneficial effects on weight loss and body composition when delivered in isolation, the greatest effects were found in the combined intervention group, where 60% of participants achieved ≥10% weight loss at 1 year. Our combined diet + exercise intervention group essentially followed the current recommendations by the NIH Obesity Education Initiative Expert Panel (5), providing the highest level of support for these recommendations: a randomized clinical trial that examined each element singly and in combination vs. a no-lifestyle-change control. Yet, despite the now overwhelming evidence for the benefits of lifestyle-induced weight loss, there are still major barriers to implementing these programs, particularly at the primary care level where the most benefit stands to be made (41,42). Providing group-based nutrition education along with exercise likely outweighs the costs of health consequences that come from untreated obesity, assuming these changes can be sustained over the long-term (43,44). We are conducting ongoing follow-up of our NEW trial participants to examine the behavioral and physiologic predictors of long-term weight loss maintenance in a free-living environment post-trial, with the goal that identification of these factors will help to guide the successful design and implementation of obesity treatment programs with the greatest potential to impact public health.


This study was funded by National Cancer Institute (NCI) NIH grants R01 CA102504 and U54-CA116847. K.E.F. received support from NIH 5KL2RR025015-03, A.K. was supported by NCI R25 CA94880. The authors thank the study participants for their time and dedication to the study.


The authors declared no conflict of interest.