Exercise Effect on Weight and Body Fat in Men and Women

Transcription

1 Exercise Effect on Weight and Body Fat in Men and Women Anne McTiernan,* Bess Sorensen,* Melinda L. Irwin, Angela Morgan,* Yutaka Yasui, Rebecca E. Rudolph,* Christina Surawicz, Johanna W. Lampe,* Paul D. Lampe,*, ** Kamran Ayub, and John D. Potter* Abstract MCTIERNAN, ANNE, BESS SORENSEN, MELINDA L. IRWIN, ANGELA MORGAN, YUTAKA YASUI, REBECCA E. RUDOLPH, CHRISTINA SURAWICZ, JOHANNA W. LAMPE, PAUL D. LAMPE, KAMRAN AYUB, AND JOHN D. POTTER. Exercise effect on weight and body fat in men and women. Obesity. 2007;15: Objectives: The effect of national exercise recommendations on adiposity is unknown and may differ by sex. We examined long-term effects of aerobic exercise on adiposity in women and men. Research Methods and Procedures: This was a 12-month randomized, controlled clinical trial testing exercise effect on weight and body composition in men (N 102) and women (N 100). Sedentary/unfit persons, 40 to 75 years old, were recruited through physician practices and media. The intervention was facility- and home-based moderate-tovigorous intensity aerobic activity, 60 min/d, 6 days/wk vs. controls (no intervention). Results: Exercisers exercised a mean 370 min/wk (men) and 295 min/wk (women), and seven dropped the intervention. Exercisers lost weight (women, 1.4 vs. 0.7 kg in controls, p 0.008; men, 1.8 vs. 0.1 kg in controls, p 0.03), BMI (women, 0.6 vs. 0.3 kg/m 2 in controls, p Received for review January 23, Accepted in final form December 6, The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. *The Fred Hutchinson Cancer Research Center, Cancer Prevention Program, Seattle, Washington; Departments of Epidemiology, Medicine, and **Pathobiology, Seattle, Washington; Yale University School of Medicine, Department of Epidemiology and Public Health, New Haven, Connecticut; Department of Public Health Sciences, University of Alberta, Edmonton, Alberta, Canada; Veterans Administration Puget Sound Health Care System, Health Services Research and Development Program, Seattle, Washington; and Virginia Mason Medical Center, Seattle, Washington. Address correspondence to Anne McTiernan, Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue N, M4-B402, Seattle, WA amctiern@fhcrc.org Copyright 2007 NAASO 0.006; men, 0.5 kg/m 2 vs. no change in controls, p 0.03), waist circumference (women, 1.4 vs. 2.2 cm in controls, p 0.001; men, 3.3 vs. 0.4 cm in controls, p 0.003), and total fat mass (women, 1.9 vs. 0.2 kg in controls, p 0.001; men, 3.0 vs. 0.2 kg in controls, p 0.001). Exercisers with greater increases in pedometer-measured steps per day had greater decreases in weight, BMI, body fat, and intra-abdominal fat (all p trend 0.05 in both men and women). Similar trends were observed for increased minutes per day of exercise and for increases in maximal oxygen consumption. Discussion: These data support the U.S. Department of Agriculture and Institute of Medicine guidelines of 60 min/d of moderate-to-vigorous physical activity. Key words: physical activity, fitness, weight loss, exercise Introduction The worldwide epidemic of overweight and obesity has resulted in calls for prevention and treatment strategies. Scientific tests of weight reduction diets have shown modest long-term efficacy (1). The U.S. Department of Agriculture now recommends that to help manage body weight and prevent gradual, unhealthy body weight gain in adulthood, individuals should engage in 60 minutes of moderate-tovigorous-intensity activity on most days of the week while not exceeding caloric intake requirements (2). The Institute of Medicine recommends that adults and children spend a total of at least 1 hour each day in moderately intense physical activity (3). However, the long-term efficacy of this level of daily physical activity on weight control in the absence of dietary change has not been established. In contrast, the U.S. Center for Disease Control and American College of Sports Medicine recommend 30 min/d of at least moderate activity on most days of the week (4,5). Increased intra-abdominal obesity is associated with insulin resistance, type 2 diabetes, hypertension, dyslipide OBESITY Vol. 15 No. 6 June 2007

2 mia, and cardiovascular disease (6 11) and could be important in promoting cancers of the colon (12), breast (13), and endometrium (14). In contrast, subcutaneous fat may be more important in excessive non-gonadal production of sex hormones (15), which increase risk for breast and endometrial cancer (16,17). The purpose of this study was to assess, in a randomized, controlled clinical trial, the effect of a 12-month moderate-to-vigorous intensity exercise program (60 min/d, 6 d/wk) on weight, anthropometrics, and body composition (percentage body fat) and abdominal fat (total, intra-abdominal and subcutaneous) in women and men. Research Methods and Procedures Participants Eligible participants were 40 to 75 years old, engaged in 90 min/wk, over the previous 3 months, of moderate-tovigorous intensity sports, recreational, or walking exercise [or if report of exercise was questionable had a maximal oxygen consumption (VO 2max ) 1 indicating low fitness level (18)], drank less than 2 alcohol drinks/d, had no personal history of invasive cancer or other serious medical conditions such as cardiovascular disease or stroke, uncontrolled hypertension, or diabetes, had a normal response to a maximal exercise tolerance test, and had normal complete blood count and blood chemistries. Participants were recruited between 2001 and 2004 through gastroenterology practices (because other primary endpoints were related to colon cancer risk), media placements, flyers, a study web site, and referrals. From 9828 invitation letters, 2033 individuals (21%) responded, and of these, 956 (47%) were potentially eligible and were interviewed (Figure 1). There were 1328 responses to media placements, with 1092 interviews completed. Major reasons for ineligibility were unwillingness to be randomized (N 297), not sedentary (N 339), and insufficient time availability (N 48). Of those eligible, a total of 395 attended an information session, 311 were screened in clinic, and 202 were enrolled. Informed consent was obtained following the requirements of the Fred Hutchinson Cancer Research Center Institutional Review Board. Participants were paid $50 and $75 after completion of baseline and 12-month data collection, respectively. Randomization Participants were randomized by the study coordinator by a computerized program to exercise or control group (referred to as exercisers and controls, respectively), blocked on sex, use of non-steroidal anti-inflammatory medications ( 2 times/wk vs. less), smoking (yes vs. no), and among 1 Nonstandard abbreviations: VO 2max, maximal oxygen consumption; MNLTPA, Minnesota Leisure Time Physical Activity; MET, metabolic equivalent test; CT, computerized tomography; FFQ, food frequency questionnaire; CI, confidence interval. women, on menopausal status (pre- or peri- vs. postmenopausal) and current use (yes vs. no) of hormone replacement therapy. Endpoints were measured by staff blinded to randomization status. Exercise Data in All Participants We assessed the previous 3 months of physical activity type, frequency, and duration at baseline and 3, 6, 9, and 12 months with a physical-activity interview adapted from the Minnesota Leisure Time Physical Activity (MNLTPA) Questionnaire (19). For eligibility, adherence, and control contamination estimation, we included only sports, recreational, and walking activities with a metabolic equivalent test (MET) level 4.0 (20). Participants also wore Accusplit pedometers (Accusplit, Silicon Valley, CA) while awake, for 1 week at baseline and quarterly, and recorded their total daily steps in a log. We assessed VO 2max once each at baseline and 12 months (18). Participants completed a maximal-graded treadmill test, with heart rate and oxygen uptake continuously monitored by a MedGraphics automated metabolic cart (MedGraphics, St. Paul, MN). The protocol was a modified branching treadmill protocol (21). The test began at 3.0 miles/h and 0% grade. The speed or grade (2% increments) of the treadmill increased every 2 minutes (i.e., stage 2, 3.5 miles/h, 0% grade; stage 3, 3.5 miles/h, 2% grade; stage 4, 3.5 miles/h, 4% grade, etc.) until the participant reached volitional fatigue and reached a respiratory exchange ratio 1.0. For purposes of determining eligibility for individuals with questionable exercise level at baseline, we classified as low fitness persons with the following age/ gender results [i.e., fair or poor according to American College of Sports Medicine guidelines (18)]: women 40 to 49 years old, VO 2max 29.0 ml/kg per minute; women 50 years old, VO 2max 25.0 ml/kg per minute; men 40 to 49 years old, VO 2max 37.0 ml/kg per minute; men 50 to 59 years old, VO 2max 33.0 ml/kg per minute; and men 60 years old, VO 2max 31.0 ml/kg per minute. All participants previously passed a screening maximal Bruce Treadmill Test at baseline. Eight women and six men reported over 90 min/wk of moderate or vigorous exercise in the previous 3 months at baseline on the MNLTPA interview, among whom VO 2max results indicated that three women and five men were unfit by age and gender criteria (18). The remaining five women and one man were enrolled into the study in error. We analyzed the trial results data with and without all 14 who reported 90 min/wk of moderate or vigorous activity at baseline included and found no differences in the results. We, therefore, included them in the final sample. Anthropometrics and Body Composition Data We measured anthropometrics at baseline and 3 and 12 months in a standardized fashion in duplicate using methods OBESITY Vol. 15 No. 6 June

3 Figure 1: Recruitment of participants. from the Women s Health Initiative (22). Weight and height were measured to the nearest 0.1 kg and 0.1 cm, respectively, with a balance beam scale and stadiometer. Waist circumference was measured at the end of normal expiration over non-binding undergarments in a horizontal plane at the natural waist (minimal location on the torso, to the nearest 0.1 cm). Hip circumference was measured to the nearest 0.1 cm at the maximal circumference below the umbilicus. At baseline and 12 months, we assessed total and percentage body fat using a DXA whole-body scanner (GE Lunar, Madison, WI) and total, intra-abdominal, and subcutaneous abdominal fat with computerized tomography (CT) scans (General Electric model CT 9800 scanner; General Electric, Waukesha, WI) at the L4 to L5 space (at 125 kv and with a slice thickness of 8 mm). A technician measured abdominal fat areas in batches using a software application (sliceomatic; Tomovision, Montreal, Quebec, Canada) that identifies and measures each of the areas of interest by tracing lines around them and computing the circumscribed areas (23). The intra-batch and inter-batch coefficients of 1498 OBESITY Vol. 15 No. 6 June 2007

4 variation, respectively, were: 0.4% and 1.2% (total abdominal fat), 2.7% and 4.5% (intra-abdominal fat), and 1.1% and 2.5% (subcutaneous fat). Baseline, 3-Month, and 12-Month Follow-up Covariate Measures We collected information at baseline and 3 and 12 months on medical history, health habits, medications, and dietary patterns by a 120-item self-administered food frequency questionnaire (FFQ) (24). Intervention The intervention was a 12-month facility- and homebased exercise program. The prescription was 6 d/wk of 60 min/session of moderate-to-vigorous aerobic exercise gradually achieved over the first 12 weeks, with an additional 5 to 10 minutes of warm up, cool down, and stretching. Three days per week, participants exercised on treadmills, stationary bikes, elliptical machines, and rowers with an exercise specialist present at one of four facilities (one located at the Fred Hutchinson Cancer Research Center Prevention Center, A. McTiernan, Director, and three private health clubs). Participants were given Polar (Polar Electro Inc., Lake Success, NY) heart rate monitors, with a prescription corresponding to 60% to 85% of their maximal heart rate on their baseline VO 2max test, and they recorded maximal session heart rate in a daily activity log. Participants also were asked to exercise at home or gym an additional 3 d/wk with the same instructions regarding duration and heart rate goal ranges. A variety of strategies was used to achieve and maintain adherence, including regular monitoring and feedback, monthly progress review meetings with behaviorally trained exercise specialists, newsletters, incentives (e.g., water bottles), and group social events. Several methods were used to measure and monitor adherence, including facility class attendance (with data on type, length, and maximal heart rate entered on facility logs, verified by the exercise specialist, and data entered), home exercise logs (with data on type, length, and maximal heart rate entered, submitted weekly, reviewed by the exercise specialist, and data entered), quarterly MNLTPA interviews, quarterly use of pedometers for 1 week with logging of steps, and VO 2max treadmill tests at baseline and 12 months. Adherence was calculated weekly from logs as facility and home sessions completed, total minutes per week, MET minutes per week, and percentage of goal 360 min/wk of exercise. Good adherence was defined as meeting at least 80% of the overall minutes per week goal of moderate-to-vigorous exercise, such that for participants meeting less than this goal over 2 or more weeks, special assistance was used to help them increase their level of participation. After each facility session, participants made an entry into a facility log including their exercise duration and intensity. The exercise specialist monitored each participant s progress in relation to their individual heart rate ranges. The participants were taught that as they became more fit, it would take more effort to attain their moderateto-vigorous heart rate targets. The exercise specialist prescribed safe increases in the cardio machine settings to ensure that participants reached the desired heart rate ranges during their facility workouts. By recording intensity and duration for each session, participants became adept at monitoring their own progress. Exercisers were not assigned a weight loss program and were asked not to change their dietary habits during the trial. Control Group Controls were asked not to change their exercise or diet habits during the trial. After completing each quarter of the study, they used a pedometer to record number of steps per day for 7 days and participated in physical activity interviews. After completion of all end-of-study measures, they were given the opportunity to participate in exercise classes for 2 months. Statistical Analyses Primary analyses were based on assigned treatment at the time of randomization, regardless of adherence or retention status (i.e., intent-to-treat), and all participants data were included in the primary analyses. We computed the mean body composition measures of exercisers and controls at baseline and 3 and 12 months. The intervention effects were evaluated by the differences in the mean changes at 3 and 12 months between the intervention and control groups using the generalized estimating equation modification to linear regression models to account for the longitudinal nature of the data. Primary analyses were unadjusted on the basis of the randomized design of the study. End-of-study anthropometric data were available for 101 of 102 randomized men and 96 of 100 randomized women. End-of-study DXA scan data were available for 97 men and 93 women, and end-of-study CT scan data were available for 99 men and 93 women. These small numbers of participants with missing follow-up data were assumed to have no change from baseline, e.g., a no-change imputation was applied so that all participants were included in the primary intent-to-treat analyses. We also explored differential intervention effects by baseline age ( 55, 55 years old) and BMI ( 25, 25.0 to 29.9, and 30.0 kg/m 2 ). These effect modification analyses were planned a priori. As a preplanned secondary analysis, we grouped the mean changes of exercise adherence measures (VO 2max, mean number of minutes exercised per week from the daily activity logs, and steps per day from pedometer logs) from baseline to 12 months. We chose tertiles for the changes in VO 2max and number of minutes exercised per week. For the OBESITY Vol. 15 No. 6 June

5 Figure 2: Percentage change and 95% CI in percentage body fat at 12 months from baseline by tertiles of duration and change in fitness level: women. 1 Percentage body fat measured by DXA. 2 Duration minutes per week spent in moderate-intensity sports activity by tertiles: low active, 250 min/wk (N 20); moderately active, 250 to 299 min/wk (N 18); and highly active, 300 min/wk (N 11). 3 Change in fitness by tertiles: low active, 5% change in VO 2max (N 15 5 missing); moderately active, 5% to 15% change in VO 2max (N 15); and highly active, 15% change in VO 2max (N 14). * Significantly different from low active group, p # Significantly different from control, p Significant trend among exercisers, p pedometer steps per day, we chose categories corresponding to 1,1to2,and 2 miles, respectively: 1760 steps, 1760 to 3520 steps, and 3520 steps. Tests for trend across the control and the three strata of the adherence in the intervention group were performed by placing the categorized adherence variable into the linear models as a continuous covariate. Age was considered as a possible confounder; however, no confounding by age was observed. Ninety-five percent confidence intervals (CIs) for the percentage changes from baseline to follow-up presented in Figures 2 to 5 were estimated using the bootstrap method. We also calculated the mean energy expenditure for the year-long intervention from the daily activity logs as: METs body weight (kilograms)/60 kg (1 kcal/kg per minute) minutes exercised per week 4.18 kj/kcal the mean energy expenditure kilojoules per week (20). The results for this analysis were similar to those for minutes per day of exercise from the daily activity logs, so they are not presented. Differences between women and men were tested by the gender intervention interaction term into the models. All statistical tests were two-sided. Statistical analyses were performed using SAS software (version 8.2; SAS Institute Inc., Cary, NC). Results Study Participants Baseline characteristics did not differ significantly between exercisers and controls (Table 1), including age, BMI, caloric intake, or percentage of calories from macronutrients. Men and women were of similar age (mean ages, 56.4 and 54.0, respectively), most were non-hispanic white, and 60% had a college degree or higher. The mean BMI was 28.7 kg/m 2 for women and 29.9 kg/m 2 for men. Men had higher levels of cardiopulmonary fitness than women, as expected from population norms (21). As reported in the MNLTPA interview that covered the previous 3 months, female and male exercisers significantly increased their mean amount of moderate or vigorous recreational physical activity at each time period (baseline, 3-, 6-, 9-, and 12-month mean of 25, 309, 364, 305, and 298 min/wk, respectively), whereas controls increased moderate or vigorous recreational physical activity level to a smaller degree (baseline, 3-, 6-, 9-, and 12-month mean of 29, 90, 95, 81, and 68 min/wk, respectively) (p 0.001). One male control indicated exercising at moderate or vigorous levels more than a mean 360 min/wk (over the 12 months of follow-up), i.e., he was a drop-in to the intervention. At baseline (pre-randomization), exercisers recorded a mean of 5963 steps/d from their pedometers, whereas controls recorded a mean of 6398 steps/d (p 0.05). Exercisers increased their steps per day by 3300 to 4000 at the follow-up points, whereas controls steps per day remained constant or decreased (p comparing change in exercisers vs. controls). Because exercisers used a variety of machines, the number of steps per day does not reflect all of their exercise activities. VO 2max increased a mean 2.5 ml/kg 1500 OBESITY Vol. 15 No. 6 June 2007

6 Figure 3: Percentage change and 95% CI in percentage body fat at 12 months from baseline by tertiles of duration and change in fitness level: men. 1 Percentage body fat measured by DXA. 2 Duration minutes per week spent in moderate-intensity sports activity: low active, 250 min/wk (N 12); moderately active, 250 to 299 min/wk (N 20); highly active, 300 min/wk (N 19). 3 Change in fitness by tertiles: low active, 5% change in VO 2max (N 14 3 missing); moderately active, 5% to 15% change in VO 2max (N 16); and highly active, 15% change in VO 2max (N 16). * Significantly different from low active group, p # Significantly different from control, p Significant trend among exercisers, p Figure 4: Percentage change and 95% CI in intra-abdominal body fat at 12 months from baseline by tertiles of duration and change in fitness level: women. 1 Intra-abdominal body fat measured by CT. 2 Duration minutes per week spent in moderate-intensity sports activity by tertiles: low active, 250 min/wk (N 20); moderately active, 250 to 299 min/wk (N 18); and highly active, 300 min/wk (N 11). 3 Change in fitness by tertiles: low active, 5% change in VO 2max (N 15 5 missing); moderately active, 5% to 15% change in VO 2max (N 15); and highly active, 15% change in VO 2max (N 14). * Significantly different from low active group, p # Significantly different from control, p Significant trend among exercisers, p OBESITY Vol. 15 No. 6 June

7 Figure 5: Percentage change and 95% CI in intra-abdominal body fat at 12 months from baseline by tertiles of duration and change in fitness level: men. 1 Intra-abdominal body fat measured by CT. 2 Duration minutes per week spent in moderate-intensity sports activity: low active, 250 min/wk (N 12); moderately active, 250 to 299 min/wk (N 20); and highly active, 300 min/wk (N 19). 3 Change in fitness by tertiles: low active, 5% change in VO 2max (N 14 3 missing); moderately active, 5% to 15% change in VO 2max (N 16); and highly active, 15% change in VO 2max (N 16). * Significantly different from low active group, p # Significantly different from control, p Significant trend among exercisers, p per minute (10.5%) in female exercisers and 3.3 ml/kg per minute (11%) in male exercisers and decreased in controls (p comparing exercisers with controls). Neither exercisers nor controls significantly changed mean total daily caloric intake, and there were no differences between exercisers and controls (data not shown). From the daily activity log data, 29 women (59.2%) and 42 men (82.3%) met 80% of the goal 360 min/wk over the 12 months. Combined, 71 (71.0%) met 80% of the goal. Five women (10.2%) and 14 men (27.5%) met or exceeded the goal of 360 min/wk over the entire 12 months. Combined, 19 (19.0%) of the exercisers met or exceeded the goal. Averaged over the entire 12 months, male exercisers completed a mean (standard deviation) 370 (86) min/wk (102.7% of goal) and median 318 min/wk, and female exercisers performed a mean (standard deviation) 295 (102) min/wk (82% of goal) and median 292 min/wk. The data were normally distributed. Only 2 of 51 male exercisers and 5 of 49 female exercisers dropped the intervention (all after 3 months). Anthropometrics After 12 months, female exercisers lost a mean 1.4 kg (a loss of 1.8%) vs. 0.7-kg gain (e.g., a gain of 0.9%) in controls (p 0.008); male exercisers lost 1.8 kg (a loss 1.9%) vs. 0.1-kg loss (a loss of only 0.1%) in controls (p 0.03) (Table 2). The weight loss occurred primarily in the first 3 months in men but after 3 months in women. BMI (kilograms per meter squared) decreased, on average, in female exercisers ( 0.6 vs. 0.3 kg/m 2 in controls, p 0.006) and in male exercisers ( 0.5 kg/m 2 vs. no change in controls, p 0.03). Waist circumference decreased a mean of 1.4 cm in female exercisers vs. 2.2-cm increase in controls (p 0.001) and decreased a mean of 3.3 cm in male exercisers vs. 0.4-cm decrease in controls (p 0.003). Hip circumference decreased a mean of 2.6 cm in female exercisers vs. 0.5-cm increase in controls (p 0.001) and decreased 1.7 cm in male exercisers vs. 0.1-cm decrease in controls (p 0.06). Differences in exercise intervention effects by sex were not statistically significant. Body Composition and Fat Distribution Changes Female exercisers experienced a mean 1.9-kg decline in total fat mass as measured by DXA scan, which represented a decline in percentage body fat from 43.3% to 41.5% vs. a mean gain of 0.2 kg fat in controls (p 0.001) (Table 2). Male exercisers experienced an average 3.0-kg loss of total fat mass, which represented a decline in percentage body fat from 31.5% to 28.8% vs. a slight increase in controls (p 0.001) (Table 2). Female exercisers experienced a mean 5.8-cm 2 decrease ( 5.5%) in intra-abdominal fat area and a mean 18.0-cm 2 decrease ( 4.8%) in subcutaneous abdominal fat area vs. mean 1.6% increase and 0.4% decrease, respectively, in 1502 OBESITY Vol. 15 No. 6 June 2007

8 Table 1. Baseline demographic and anthropometric characteristics Women Men Exercisers (N 49): mean (SD) range or N (%) Controls (N 51): mean (SD) range or N (%) p Exercisers (N 51): mean (SD) range or N (%) Controls (N 51): mean (SD) range or N (%) p Age (yrs) 54.4 (7.1) 43.0 to (5.6) 42.5 to (6.7) 40.3 to (7.6) 40.3 to Race American Indian/Alaskan Native 1 (2.0) 1 (2.0) (0.0) 0 (0.0) 0.55 Asian or Pacific Islander 4 (8.2) 1 (2.0) 2 (3.9) 3 (5.9) African-American 1 (2.0) 2 (3.9) 0 (0.0) 0 (0.0) Hispanic 1 (2.0) 0 (0.0) 0 (0.0) 0 (0.0) White, non-hispanic 42 (85.7) 47 (92.2) 48 (94.1) 48 (94.1) Other 0 (0.0) 0 (0.0) 1 (2.0) 0 (0.0) Education College degree or more 28 (57.1) 34 (66.7) (64.7) 28 (54.9) 0.31 Height (cm) (6.9) 153 to (6.4) 151 to (6.3) 166 to (8.0) 163 to Weight (kg) 78.0 (17.8) 56.3 to (12.8) 58.6 to (14.9) 68.0 to (18.2) 66.8 to BMI (kg/m 2 ) 28.9 (5.5) 21.2 to (4.8) 20.3 to (3.7) 23.3 to (4.8) 21.3 to Waist circumference (cm) 87.5 (13.6) 65.6 to (11.4) 63.5 to (9.8) 83.0 to (12.5) 82.3 to Hip circumference (cm) (13.3) 93.1 to (10.1) 92.0 to (8.3) 91.0 to (8.8) 94.3 to Total fat mass (kg)* 34.4 (13.7) 33.9 (10.4) (9.54) 29.0 (9.47) 0.50 Body fat (%)* 43.3 (7.2) 27.4 to (6.8) 28.7 to (6.4) 12.4 to (5.9) 14.5 to Total abdominal fat (cm 2 )* 711 (233) 327 to (187) 334 to (166) 482 to (207) 445 to Intra-abdominal fat (cm 2 )* 106 (60.8) 0 to (55.8) 12.9 to (66.3) 32.7 to (79.1) 45.3 to Subcutaneous fat (cm 2 )* (178.2) 113 to (144.8) 90.2 to (105.9) 122 to (131.1) 83.9 to Steps per day 5959 (2567) 1272 to (2931) 1717 to (2778) 1422 to (3236) 1610 to Moderate-to-vigorous exercise in past 3 months (min/wk) from MNLTPA interview 26.8 (47.7) 0 to (55.3) 0 to (58.3) 0 to (59.7) 0 to VO 2max (ml/kg/min) 23.8 (5.09) 14.9 to (4.34) 17.5 to (5.92) 20.7 to (6.72) 15.0 to Energy intake (kcal)* 1543 (720) 552 to (555) 610 to (639) 528 to (566) 662 to SD, standard deviation; MNLTPA, Minnesota Leisure Time Physical Activity; Vo 2max, maximal oxygen consumption; CT, computerized tomography; FFQ, food frequency questionnaire. * One female control was missing baseline DXA, CT, and FFQ. One male control and two male exercisers were missing baseline DXA and FFQ. These participants were excluded from the analysis for these variables. Student s t test comparing baseline characteristics for female exercisers vs. controls. Student s t test comparing baseline characteristics for male exercisers vs. controls. OBESITY Vol. 15 No. 6 June

9 Table 2. Baseline, 3-month, and 12-month body composition measures Baseline 3 Months 12 Months Exercisers* mean (SD) Controls mean (SD) Exercisers Controls Exercisers Controls Mean (SD) (% ) Mean (SD) (% ) Mean (SD) (% ) Mean (SD) (% ) Women Hip Weight (kg) 78.0 (17.8) 77.9 (12.8) 77.6 (17.1); p ( 0.5) 78.7 (12.8) 0.8 ( 1.0) 76.6 (17.4); p 0.008** 1.4 ( 1.8) 78.6 (13.5) 0.7 ( 0.9) Waist circumference (cm) 87.5 (13.6) 85.2 (11.4) 85.8 (13.2); p ( 1.9) 86.7 (13.1) 1.5 ( 1.8) 86.1 (13.5); p 0.001** 1.4 ( 1.6) 87.4 (12.4) 2.2 ( 2.6) circumference (cm) (13.3) (10.1) (12.3); p ( 1.6) (10.2) 0.6 (.5) (12.9); p 0.001** 2.6 ( 2.4) (10.1) 0.5 ( 0.5) BMI (kg/m 2 ) Hip 28.9 (5.48) 28.5 (4.80) 28.7 (5.27); p ( 0.7) 28.8 (4.80) 0.3 ( 1.1) 28.3 (5.31); p 0.006** 0.6 ( 2.1) 28.8 (5.16) 0.3 ( 1.1) Total fat mass (kg) 34.4 (13.6) 33.9 (10.4) 32.5 (13.4); p 0.001** 1.9 (5.8) 34.1 (10.5) 0.2 ( 0.6) Body fat (%) 43.3 (7.18) 43.0 (6.80) 41.5 (7.89); p 0.002** 1.8 ( 4.2) 42.9 (6.73) 0.1 ( 0.2) Total abdominal fat (cm 2 ) (232.5) (186.6) (225.6); p 0.13** 17.1 ( 2.4) (201.9) 0.2 (0.0) Intra-abdominal fat (cm 2 ) (60.8) (55.8) (58.8); p 0.10** 5.8 ( 5.5) (59.6) 1.6 ( 1.6) Subcutaneous fat (cm 2 ) (178.2) (144.9) (168.0); p 0.45** 18.0 ( 4.8) (153.4) 1.5 ( 0.4) Men Weight (kg) 94.8 (14.9) 97.4 (18.2) 93.6 (14.8); p ( 1.3) 97.8 (17.9) 0.4 ( 0.4) 93.0 (14.8); p 0.03** 1.8 ( 1.9) 97.3 (17.4) 0.1 ( 0.1) Waist circumference (cm) (9.8) (12.5) (10.8); p ( 1.9) (13.2) 0.6 ( 0.6) 99.5 (10.6); p 0.003** 3.3 ( 3.2) (12.4) 0.4 ( 0.4) circumference (cm) (8.3) (8.8) (8.30); p ( 0.9) (8.37) 0.8 ( 0.7) (9.30); p 0.06** 1.7 ( 1.6) (8.49) 0.1 (0.0) BMI (kg/m 2 ) 29.7 (3.7) 30.1 (4.8) 29.3 (3.75); p ( 1.3) 30.2 (4.69) 0.1 ( 0.3) 29.2 (3.89); p 0.03** 0.5 ( 1.7) 30.1 (4.64) 0.0 (0.0) Total fat mass (kg) 30.3 (9.54) 29.0 (9.47) 27.3 (10.4); p 0.001** 3.0 ( 9.9) 29.2 (9.52) 0.2 ( 0.7) Body fat (%) 31.5 (6.4) 29.7 (5.9) 28.8 (7.71); p 0.001** 2.7 ( 8.6) 29.9 (5.78) 0.2 ( 0.7) Total abdominal fat (cm 2 ) (165.6) (207.1) (178.5) p 0.02** 43.3 ( 5.4) (202.9) 9.4 ( 1.1) Intra-abdominal fat (cm 2 ) (66.3) (79.1) (76.6); p 0.45** 12.2 ( 7.5) (72.2) 6.2 ( 3.5) Subcutaneous fat (cm 2 ) (105.9) (131.1) (105.8); p 0.001** 36.3 ( 11.2) (128.3) 4.4 ( 1.4) SD, standard deviation; Vo 2max, maximal oxygen consumption; CT, computerized tomography. * Women, N 49. Women, N 51 for physical exam measures and VO 2max ; N 50 for DXA and CT measures. Men, N 51 for physical exam measures, VO 2max, and CT measures; N 49 for DXA measures. Men, N 51 for physical exam measures, VO 2max, and CT measures; N 50 for DXA measures. p Value comparing change from baseline to 3 months in exercisers vs. controls. ** p Value comparing change from baseline to 12 months in exercisers vs. controls OBESITY Vol. 15 No. 6 June 2007

10 controls, but these differences were not statistically significant (Table 2). Total abdominal fat area decreased a mean of 17.1 cm 2 ( 2.4%) in female exercisers vs. no substantial change in controls, but the differences were also not statistically significant. Intra-abdominal fat area decreased to a greater degree in male exercisers ( 12.2 cm 2, 7.5%) than controls ( 6.2 cm 2, 3.5%), although the differences also were not statistically significant (Table 2). Subcutaneous fat area decreased, on average, 36.3 cm 2 ( 11.2%) in male exercisers vs. a decrease of 4.4 cm 2 ( 1.4%) in controls (p 0.001). Total abdominal fat decreased 43.3 cm 2 ( 5.4%) in male exercisers vs. 9.4 cm 2 ( 1.1%) in controls (p 0.02). Differences in exercise intervention effects by sex were not statistically significant. Effects of Age, Menopausal Status, and Baseline BMI on Intervention Efficacy The effect of exercise on adiposity variables did not differ significantly by age ( 55 vs. 55 years) or, among women, by menopausal status. Overweight and obese female exercisers lost more weight (mean, 2.1 and 1.9 kg, respectively) than normal-weight women (0.4 kg). Normal-weight male exercisers gained a mean 1.4 kg, whereas overweight and obese male exercisers lost weight (mean, 2.8 and 1.9, respectively). Tests of interactions between exercise and adiposity variable changes comparing normal-weight vs. overweight or obese individuals were not statistically significant (Table 3). Exercise Effect by Intervention Adherence and Change in Cardiopulmonary Fitness We used three measures of adherence to assess effect of adherence on changes in adiposity and body composition: mean minutes per day of moderate or vigorous activities reported in daily activity logs, mean increase over baseline in steps per day reported on pedometer logs that participants used when wearing Accusplit Pedometers for 1 week at baseline and months 3, 6, 9, and 12 and change from baseline to 12-month VO 2max. Missing daily activity log data values were set to 0, and all participants were included. A total of five male and eight female exercisers did not have any pedometer log data so were deleted from the analysis for steps per day. Other missing pedometer log data were set to the corresponding baseline value. All 202 study participants completed baseline VO 2max tests, and 48 male exercisers and 44 female exercisers completed 12-month VO 2max tests. Data for participants with missing 12-month VO 2max tests were deleted from the analysis by VO 2max change. In female exercisers, losses in weight, BMI, percentage body fat, and intra-abdominal fat increased with increasing tertiles of mean minutes exercised per week, although these effects of adherence were only statistically significant comparing those exercising for a mean of 250 to 299 min/wk vs. those exercising less than a mean of 250 min/wk on change in percentage body fat (p 0.05) (Figures 2 and 4). In male exercisers, the mean number of minutes exercised per week was variably associated with change in body composition variables, and the greatest effect was observed in those exercising 250 to 299 min/wk (Figures 3 and 5). In male exercisers, tests for trend were statistically significant for effect of increased adherence on greater loss in weight (p 0.03), BMI (p 0.03), total fat mass (p 0.01), percentage body fat (p 0.001), total abdominal fat (p 0.02), and intra-abdominal fat (p 0.01). Comparing exercisers at the three adherence levels as measured by the daily activity logs vs. the controls, those exercising a mean 250 to 299 min/wk had statistically significantly greater loss in percentage body fat (women, p 0.002; men, p 0.03), and those exercising more than a mean 300 min/wk had statistically significantly greater loss in percentage body fat (women, p 0.01; men, p 0.001) and intra-abdominal fat (women, p 0.03; men, p 0.04) (Figures 2 to 5). Results classified by weekly energy expenditure largely paralleled the results for mean minutes of moderate-vigorous intensity exercise per week. Exercisers with greater pedometer-measured steps per day had greater decreases in weight, hip circumference, BMI, body fat, and intra-abdominal fat (all p trend 0.05 in both men and women) (Table 4). Female exercisers who recorded a mean increase over baseline of 1760 to 3520 steps/d lost 1.2 kg, and those who reported 3520 steps/d lost a mean 2.3 kg, vs. a loss of just 0.1 kg in those who increased 1760 steps/d (p trend including controls 0.02). A similar, non-significant trend was observed for waist circumference. Female exercisers whose steps increased by 1760, 1760 to 3520, and 3520 steps/d experienced decreases in BMI of 0.2, 0.5, and 0.8 kg/m 2, respectively (p trend 0.01). Both total fat mass and percentage body fat decreased significantly with greater increases in steps per day, such that women who increased 3520 steps/d experienced mean decreases in fat mass of 3.5 kg and percentage body fat of 3.4%, and female exercisers who reported less increase in steps per day experienced smaller fat losses (p trend for both fat mass and percentage body fat). Change in intra-abdominal were also significantly related to increase in steps per day. Female exercisers whose steps increased by 1760 steps/d experienced an increase of 0.9 cm 2 of intra-abdominal fat, whereas those whose steps increased by 1760 to 3520 and 3520 steps/d experienced decreases in intra-abdominal fat of 7 and 13 cm 2, respectively (p trend 0.02). Male exercisers who recorded a mean increase over baseline of 1760 to 3520 steps/d lost a mean 0.3 kg, and those who reported 3520 steps/d lost a mean 3.9 kg, vs. a loss of 1.4 kg in those who increased 1760 steps/d (p trend including controls 0.001). Mean decreases in waist circumference also were greater in those whose mean increase in steps/d was 3520, but the test for trend was not statis- OBESITY Vol. 15 No. 6 June

11 Table 3. Select baseline and 12-month body composition measures, stratified by baseline BMI Women Baseline 12 Months Exercisers Controls Exercisers Controls Mean (SD) Mean (SD) Mean (SD) (% ) Mean (SD) (% ) Weight (kg) BMI 25* 64.4 (6.4) 65.7 (4.6) 64.0 (6.43); p ( 0.6) 66.1 (3.96) 0.4 ( 0.6) BMI 25 to (6.5) 74.3 (6.1) 68.5 (6.09); p ( 3.0) 75.2 (5.75) 0.9 ( 1.2) BMI (16.1) 90.4 (10.5) 93.2 (16.2); p ( 2.0) 91.1 (12.9) 0.7 ( 0.8) BMI (kg/m 2 ) BMI 25* 23.5 (1.3) 22.9 (1.1) 23.3 (1.05); p ( 0.9) 23.1 (1.11) 0.2 ( 0.9) BMI 25 to (1.1) 27.4 (1.4) 26.5 (1.68); p ( 2.9) 27.7 (1.46) 0.3 ( 1.1) BMI (3.9) 33.7 (2.7) 33.9 (3.87); p ( 2.0) 34.0 (4.00) 0.3 ( 0.9) Intra-abdominal fat (cm 2 ) BMI 25* 50.6 (36.0) 63.2 (29.3) 45.4 (27.0); p ( 10.3) 58.8 (26.3) 4.4 ( 7.0) BMI 25 to (49.3) 92.4 (42.8) 97.0 (44.6); p ( 8.5) 94.9 (39.7) 2.5 ( 2.7) BMI (45.0) 139 (59.2) 148 (45.2); p ( 2.6) 144 (67.0) 5.0 ( 3.6) Men Baseline 12 Months Exercisers Controls Exercisers Controls Mean (SD) Mean (SD) Mean (SD) (% ) Mean (SD) (% ) Weight (kg) BMI 25* 75.4 (5.4) 73.2 (6.8) 76.8 (10.9); p ( 1.9) 74.6 (6.31) 1.4 ( 1.9) BMI 25 to (6.9) 88.7 (8.2) 85.3 (6.07); p ( 3.2) 89.5 (8.91) 0.8 ( 0.9) BMI (12.1) (15.6) (12.2); p ( 1.8) (15.4) 1.2 ( 1.1) BMI (kg/m 2 ) BMI 25* 24.4 (0.7) 23.4 (1.3) 24.8 (2.44); p ( 1.6) 23.9 (1.26) 0.5 ( 2.1) BMI 25 to (1.5) 27.3 (1.4) 27.1 (1.67); p ( 2.9) 27.5 (1.65) 0.2 ( 0.7) BMI (2.7) 33.9 (3.7) 32.2 (3.09); p ( 1.8) 33.6 (3.82) 0.3 ( 0.9) Intra-abdominal fat (cm 2 ) BMI 25* 101 (64.1) 91.6 (57.0) 84.9 (80.2); p ( 15.9) 114 (82.5) 22.4 ( 24.5) BMI 25 to (59.0) 150 (61.6) 111 (54.1); p ( 18.4) 138 (55.2) 12 ( 8.0) BMI (48.6) 219 (71.5) 201 (57.5); p (0.0) 210 (60.7) 9.0 ( 4.1) SD, standard deviation; CT, computerized tomography. Exercise effect on weight in women with baseline BMI 25 vs. exercise effect in women with BMI 25 to 29.9 and BMI 30, interaction p 0.17, 0.32, respectively. In men, interaction p 0.14, 0.19, respectively. Exercise effect on BMI in women with baseline BMI 25 vs. exercise effect in women with BMI 25 to 29.9 and BMI 30, interaction p 0.16, 0.30, respectively. In men, interaction p 0.17, 0.87, respectively. Exercise effect on intra-abdominal fat in women with baseline BMI 25 vs. exercise effect in women with BMI 25 to 29.9 and BMI 30, interaction p 0.46, 0.74, respectively. In men, interaction p 0.45, 0.11, respectively. * Women, N 14 controls with physical exam measures; N 13 controls with CT measures; N 16 exercisers. Men, N 6 controls; N 7 exercisers. Women, N 18 controls; N 14 exercisers. Men, N 20 controls; N 20 exercisers with physical exam measures and CT measures. Women, N 19 controls; N 19 exercisers. Men, N 25 controls with physical exam measures and CT measures; N 24 exercisers with physical exam measures and CT measures. p-value comparing change from baseline to 12 months in exercisers vs. controls within BMI strata OBESITY Vol. 15 No. 6 June 2007

12 Table 4. Change in body composition measures, stratified by change in steps per day, in women and men Baseline mean (standard deviation) 12 Months mean (standard deviation) P Trend Women Weight (kg) Controls 77.9 (12.8) 78.6 (13.5) Increase in steps/d 75.5 (18.0) 75.1 (18.5) to 3520 Increase in steps/d 76.9 (19.1) 75.7 (17.4) Increase in steps/d 75.0 (16.6) 72.7 (15.1) 2.3 Waist circumference (cm) Controls 85.2 (11.4) 87.4 (12.4) Increase in steps/d 84.6 (12.6) 84.0 (12.4) to 3520 Increase in steps/d 86.8 (15.4) 86.1 (15.8) Increase in steps/d 84.7 (11.2) 82.6 (10.3) 2.1 Hip circumference (cm) Controls (10.1) (10.1) Increase in steps/d (12.8) (14.2) to 3520 Increase in steps/d (13.2) (10.8) Increase in steps/d (12.5) (11.5) 2.0 BMI (kg/m 2 ) Controls 28.5 (4.80) 28.8 (5.16) Increase in steps/d 27.8 (5.09) 27.6 (5.15) to 3520 Increase in steps/d 28.3 (6.23) 27.8 (5.47) Increase in steps/d 28.2 (4.83) 27.4 (4.45) 0.8 Total fat mass (kg) Controls 33.9 (10.4) 34.1 (10.5) Increase in steps/d 32.4 (13.5) 31.6 (13.8) to 3520 Increase in steps/d 33.2 (14.3) 31.3 (13.2) Increase in steps/d 32.4 (13.3) 28.9 (11.8) 3.5 Body fat (%) Controls 43.0 (6.80) 42.9 (6.73) Increase in steps/d 41.8 (6.45) 41.1 (7.23) to 3520 Increase in steps/d 41.8 (7.57) 40.2 (7.74) Increase in steps/d 42.7 (7.57) 39.3 (8.10) 3.4 Total abdominal fat (cm 2 ) Controls 710 (187) 710 (202) Increase in steps/d 656 (213) 664 (207) to 3520 Increase in steps/d 705 (220) 670 (216) Increase in steps/d 677 (211) 655 (200) 22 Intra-abdominal fat (cm 2 ) Controls 103 (55.8) 104 (59.6) Increase in steps/d 83.1 (47.0) 84.0 (39.7) to 3520 Increase in steps/d 124 (76.5) 117 (81.0) Increase in steps/d 102 (61.1) 89.4 (54.8) 13 Subcutaneous fat (cm 2 ) Controls 374 (145) 372 (153) Increase in steps/d 349 (165) 382 (178) to 3520 Increase in steps/d 349 (159) 315 (139) Increase in steps/d 339 (181) 332 (170) OBESITY Vol. 15 No. 6 June

13 Table 4. Continued Baseline mean (standard deviation) 12 Months mean (standard deviation) P Trend Men Weight (kg) Controls 97.4 (18.2) 97.3 (17.4) Increase in steps/d 97.3 (15.7) 95.9 (15.9) to 3520 Increase in steps/d 90.4 (13.7) 90.1 (13.8) Increase in steps/d 97.5 (17.0) 93.6 (17.0) 3.9 Waist circumference (cm) Controls (12.5) (12.4) Increase in steps/d (9.58) (9.92) to 3520 Increase in steps/d (10.3) 98.1 (11.6) Increase in steps/d (10.3) 99.0 (11.3) 5.9 Hip circumference (cm) Controls (8.81) (8.49) Increase in steps/d (8.83) (9.39) to 3520 Increase in steps/d (7.42) (11.8) Increase in steps/d (9.28) (7.95) 4.2 BMI (kg/m 2 ) Controls 30.1 (4.84) 30.1 (4.64) Increase in steps/d 30.2 (3.76) 29.8 (4.14) to 3520 Increase in steps/d 28.6 (3.70) 28.5 (3.65) Increase in steps/d 30.2 (4.18) 29.1 (4.47) 1.1 Total fat mass (kg) Controls 29.0 (9.47) 29.2 (9.52) Increase in steps/d 32.2 (9.86) 29.1 (10.4) to 3520 Increase in steps/d 28.6 (9.86) 27.1 (11.2) Increase in steps/d 30.7 (10.6) 26.0 (11.4) 4.7 Body fat (%) Controls 29.7 (5.95) 29.9 (5.78) Increase in steps/d 32.3 (5.76) 29.7 (6.82) to 3520 Increase in steps/d 30.9 (6.98) 29.4 (8.48) Increase in steps/d 31.2 (7.42) 27.2 (8.59) 4.0 Total abdominal fat (cm 2 ) Controls 821 (207) 812 (203) Increase in steps/d 818 (163) 796 (170) to 3520 Increase in steps/d 756 (173) 740 (191) Increase in steps/d 826 (180) 742 (197) 84 Intra-abdominal fat (cm 2 ) Controls 177 (79.1) 171 (72.2) Increase in steps/d 139 (49.6) 145 (61.7) to 3520 Increase in steps/d 173 (86.8) 166 (98.6) Increase in Steps/d 165 (59.4) 130 (65.2) 35 Subcutaneous fat (cm 2 ) Controls 324 (131) 319 (128) Increase in steps/d 353 (114) 312 (105) to 3520 Increase in steps/d 277 (85.6) 249 (88.6) Increase in steps/d 345 (119) 300 (118) OBESITY Vol. 15 No. 6 June 2007

14 tically significant. The greatest decrease in BMI was also in those whose increase in steps per day was 3520 (p trend 0.001). Both total fat mass and percentage body fat losses were greatest in male exercisers whose increase in steps per day was 3520, but the group reporting increases of 1760 to 3520 steps/d had lower mean loss of body fat and percentage body fat than did those reporting 1760 steps/d. In male exercisers, both intra-abdominal and subcutaneous abdominal fat decreased significantly with increasing number of added steps per day over baseline. Male exercisers whose steps increased by 1760 to 3520 and 3520 steps/d experienced mean decreases in intra-abdominal fat of 7 and 35 cm 2, respectively, whereas male exercisers whose steps per day increased 1760 experienced a mean gain of 6 cm 2 in intra-abdominal fat (p trend 0.03). Examination of the change in body composition among exercisers by tertiles of change in VO 2max ( 5% increase, 5% to 15% increase, 15% increase) showed that, among women, those in the top two tertiles for increase in VO 2max experienced the greatest declines in weight, waist circumference, and percentage fat mass (p trend all 0.05) (Figure 2). Those in the top tertile for increase in VO 2max experienced a statistically significant decrease of 16.6 cm 2 of intra-abdominal fat ( 15.6%, p trend 0.03) (Figure 4). Among men, greater increases in VO 2max were associated with increasing loss of weight (p trend 0.001) and declining waist circumference (p trend 0.001), BMI (p trend 0.001), percentage body fat (p trend 0.001), total abdominal fat (p trend 0.002), subcutaneous abdominal fat (p trend 0.001), and intra-abdominal fat (p trend 0.06) compared with controls (Figures 3 and 5). Similar trends were observed when exercisers in the top tertiles of VO 2max increase were compared with those in the bottom tertile. Adverse Effects At the end of the trial, we asked all participants the following question: In the past 12 months, have you had an injury that prevented you from doing your usual daily activities (including strains, sprains, bursitis, fractures, and other injuries to muscle, tendon, bone, joint, or ligament)? There were no differences in percentage of participants responding yes to this question between exercisers (28%) and controls (27%). One exerciser experienced a recurrence of an old medial meniscus tear with scar tissue buildup, which required surgical repair. Discussion This randomized, controlled clinical trial showed that moderate-to-vigorous intensity exercise, such as that recommended by the U.S. Department of Agriculture (2), results in significant weight and fat loss over 12 months in sedentary individuals. The mean weight loss (1.4-kg decrease in female exercisers vs. 0.7-kg gain in female controls, 1.8-kg decrease in male exercisers vs. 0.1-kg decrease in male controls) was modest but is similar to that observed with some low-fat dietary interventions over a comparable time frame (25). This is promising because some individuals may prefer to control weight through exercise rather than (or as well as) through restrictive diets. We observed significant differences in reported physical activity levels in exercisers vs. controls at 3, 6, 9, and 12 months follow-up. In the MNLTPA interview, controls reported increases from baseline activity level, but they did not increase VO 2max, on average, and their recorded pedometer steps per day did not change from baseline, which suggests that controls, on average, over-reported activity levels in the interview. The exercisers indicated exercising a mean 298 min/wk in the previous 3 months at the 12- month interview vs. 24 min/wk in the 3 months before baseline. This level of increase in caloric expenditure would be expected to produce 0.15 kg of weight loss per week (assuming an approximate 4 kcal expended per minute of moderate activity) or 7.8 kg over 1 year if caloric intake remained stable. That the exercisers lost only a mean 1.4 kg (women) and 1.8 kg (men) suggests that some exercisers increased caloric intake. Our measure of dietary intake, the FFQ, does not accurately capture caloric intake (26). However, we observed no difference between exercisers and controls in change in the FFQ-measured caloric intake over the 12 months of study; thus, it is unlikely that differential change in diet by study arm accounted for the observed differences in weight loss between exercisers and controls. Exercisers vs. controls experienced statistically significant decreases in BMI, waist and hip circumferences, fat mass and percentage body fat, and, among men, total and subcutaneous abdominal fat. The amount of intra-abdominal fat loss was substantial ( 5.5% in female exercisers vs. 1.6% in controls; 7.5% in male exercisers vs. 3.5% in controls). A few studies have shown losses in total body weight from exercise training without dieting (27 32), whereas others have reported no weight loss (33,34), although few of these have been long-term (e.g., 12 months) (27,35). Few trials have tested exercise effect on intraabdominal fat (27,30,31,36). We observed comparable effects of the exercise intervention in men and women. Men experienced greater loss of overall and abdominal fat, but the differences between sexes were not statistically significant. We did not observe statistically significant differences in the exercise effect on body composition by age, or among women, by menopausal status. Individuals who were overweight or obese at baseline lost more weight than those of normal weight, although the differences in weight change by baseline BMI were not significant. A benefit of including normal-weight individuals is to assess whether exercise can prevent weight gain because yearly weight gain is a major contributor to lifelong OBESITY Vol. 15 No. 6 June