OBJECTIVE: To evaluate prospectively the influence of habitual physical activity on body weight of men and women and to develop a model that defines the role of physical activity on longitudinal weight change.
DESIGN AND SETTING: Occupational cohort study conducted for a mean of 5.5 y.
SUBJECTS: A total of 496 (341 male and 155 female) NASA/Johnson Space Center employees who completed the 3 month education component of the employee health-related fitness program and remained involved for a minimum of 2 y.
MEASUREMENTS: Body weights were measured at baseline (T1) and follow-up (T2), and habitual physical activity was obtained from the mean of multiple ratings of the 11-point (0–10) NASA Activity Scale (NAS) recorded quarterly between T1 and T2. Other measures included age, gender, VO2 max obtained from maximal treadmill testing, body mass index (BMI), and body fat percentage.
RESULTS: Multiple regression demonstrated that mean NAS, T1 weight, aging and gender all influence long-term T2 weight. T1 age was significant for the men only. Independently, each increase in mean NAS significantly (P<0.01) reduced T2 weight in men (b=−0.91 kg; 95% CI:−1.4 to−0.42 kg) and women (b=−2.14 kg; 95% CI:−2.93 to−1.35 kg). Mean NAS had a greater effect on T2 weight as T1 weight increased, and the relationship was dose-dependent.
CONCLUSIONS: Habitual physical activity is a significant source of long-term weight change. The use of self-reported activity level is helpful in predicting long-term weight changes and may be used by health care professionals when counseling patients about the value of physical activity for weight control.
Adult Americans tend to gain weight1 and become increasingly overweight as they age.2 While gaining a slight amount of weight with increasing age has been associated with a longer life-span,3 large weight gains elevate the risks for disease and shorten life expectancy.4 When the body weight rises from normal to increasingly severe obesity, there is a parallel rise in the rate and relative incidence of serious medical conditions5 and mortality.6 Obesity in the North American population rose dramatically in the 1980s.7 In response the American Heart Association (AHA) declared obesity a serious public health threat (AHA, 1998) and the National Heart, Lung, and Blood Institute released the first federal guidelines for identifying, evaluating, and treating overweight and obese adults.8
In 1996 we reported the results of a longitudinal study9 on 1229 federal employees at the NASA/Johnson Space Center that documented a 10-y weight gain of 3.9 kg (8.6 lb) and 7.1 kg (15.7 lb) for males and females, respectively. These weight gains were associated with concomitant changes in blood lipids that raise the risks for cardiovascular disease. A major limitation of the NASA 10-y paper was that physical activity was not studied. This was a significant omission because regular physical activity may ameliorate many of the negative effects of weight gain,10,11,12 and it has been found to be one of the most consistent predictors of long-term weight loss maintenance.13,14,15,16,17 Cross-sectional studies have shown a negative relationship between body weight and physical activity, but they could not show whether physical activity was a cause or a consequence of body weight. It has been argued that solving this directional puzzle requires longitudinal research with multiple measurements of habitual activity taken over a lengthy period.18
The purpose of this study was to examine the influence of habitual physical activity on longitudinal change in body weight of men and women, and to develop a model to define the relationship between longitudinal weight change and physical activity.
The subjects of this occupational cohort were 341 male and 155 female NASA/Johnson Space Center employees. The ethnic composition (using the NASA classification method) of the sample was as follows: white, 90.5%; black, 3.2%; Hispanic, 4.4%; and Asian/Pacific Islander, 1.8%. Over 85% of the employees were college educated. These characteristics are representative of the total workforce. There were three inclusion criteria for this study. First, all were medically screened and declared healthy for vigorous physical activity in accordance with recommendations and guidelines of the American College of Sports Medicine.19 Second, all subjects completed the 3 month education component of the NASA/Johnson Space Center employee health-related fitness program. Third, all subjects remained involved in the fitness program for a minimum of 2 y.
Baseline data (T1) were obtained from the employees' medical records at entry into the program. After completing the 3 month educational component of the health-related fitness program, employees participated in a self-supervised exercise program that included periodic re-testing. The fitness tests were offered on a voluntary basis every 3 months (ie four times per year). The test battery included measurements of body weight, percentage fat, 1-min bent-knee sit-ups, 1-min push-ups, sit and reach test of lower back/hamstring flexibility, self-report physical activity measured by the NASA Activity Scale (NAS), and VO2max estimated from performance on a 1.5-mile run or a 1-mile walk. Procedures for administering these field tests have been published elsewhere.20 The standard for data inclusion was that the subjects had to return for testing for a period of 2 y or longer and during that period complete a minimum of 50% of the quarterly fitness tests. The years of employee participation ranged from 2 to 15 y (mean=5.5±3.4 y); the time period between the first and final weight measurements was also in this range. The NAS ratings were obtained each time the subject completed a fitness test battery in the period between the first and final weight measurements. The number of NAS ratings recorded per individual ranged from 5 to 60 (mean=14.2±10.4). The T2 data were obtained at each employee's final retest session.
All employees were medically screened for entrance into the health-related fitness program. The screening included a maximal treadmill stress test for men age 40 or over and women over age 50 y. Stress testing was done on younger employees if: (1) they exhibited symptoms or were diagnosed with cardiovascular disease; and (2) the medical examination documented two or more cardiovascular disease risk factors. A total of 287 employees (60%) had a stress test. For those who had a stress test, VO2max was measured by indirect calorimetry. These methods are fully outlined in other sources.21,22,23 A non-exercise method21 was used to estimate VO2max at baseline for those who did not have an exercise stress test. Percentage body fat was measured by the skin fold method.24,25 The VO2max and body composition data were used for descriptive purposes.
Level of physical activity was obtained by self-report with the NAS shown in the Appendix.
The scale enables subjects to rate their general activity behavior over the previous 30 days. The scale range is from 0 to 10, which is based on the total weekly minutes spent in exercise or the total weekly miles run or walked. A NAS of 0–1 represents very low activity. A rating of 2–3 represents regular recreation or work of modest effort in such activities as golf or yard work for a weekly total of between 30 min to 2 h. Ratings of 4–10 represent regular participation in aerobic exercise ranging from light to heavy exercise.
At each quarterly retest the employees used the NAS to rate their level of physical activity. The mean of all ratings between T1 and T2 was used to define the employee's level of physical activity. These data were used in two ways. First, the mean ratings were used to place the employee into a physical activity group. The groups defined by the mean NAS rating were: Inactive, ≤3.5; Moderately Active, >3.5–4.99; Active 5.0–6.5; and Very Active >6.5. Second, the employee's mean rating was used as a continuous variable for data analysis. While the rating was self-report, previous research has established the scale's validity.22,23
Analysis of variance (ANOVA) was used to evaluate gender differences on all variables studied. Multivariate analysis of variance (MANOVA) was used to study changes in body weight for gender and level of physical activity that occurred between T1 and T2. Wilks' lambda was the MANOVA test for significance. The Wilks' lambda procedure, which examines the multivariate dispersion among all of the groups of variables, is favored by statisticians because it is derived by the maximum likelihood technique.26 Multiple regression was used as a post hoc analysis to: (1) define the independent influence of physical activity on longitudinal weight change; and (2) provide an empirical model to define the level of physical activity needed for weight control. The dependent variable was T2 weight. The independent variables were T1 weight, T1 age, time differences between T1 and T2 tests (ie aging), and mean NAS.
Table 1 gives the subject baseline characteristics contrasted by gender. The males were significantly older than the females, the difference being 6 y.
The men and women differed significantly on height, weight, BMI, percentage body fat and VO2max. These are well-documented gender differences. When broken into age groups the weight and BMI of the men and women in this sample were very similar to large national samples, such as the third National Health and Nutrition Examination Survey (NHANES III).5 Using the American College of Sports Medicine standards,27 the men were in the average aerobic fitness and body fat categories for their age group; the women were average in aerobic fitness, but slightly poorer than average in body composition. The NAS at T1 was slightly, but significantly greater for the men. The mean NAS shows both the men and women increased their level of physical activity once they entered the fitness program and the significant gender difference in physical activity remained.
Table 2 gives the T1 and T2 body weight means. The time difference between T1 and T2 was 5.7 (±3.4) y for the men and 4.9(±3.2) y for the women. This difference was statistically significant (F(1, 495)=6.06; P=0.014). The male and female means shown in Table 2 are for the total sample and also contrasted by physical activity group. The actual mean annual change in weight for the total sample of men was 0.07(±1.58) kg and 0.27 (±1.72) kg for the women.
MANOVA showed there was a differential gender T1 to T2 weight change by exercise group (Wilks' Lambda 0.981, F(3, 489)=3.77, P=0.01). Figure 1 shows the trends in the mean weight changes (Δ=T2−T1) for men and women by NAS defined activity group.
The data illustrated in Figure 1 show a progressive inverse relationship between exercise group and weight change. The error bars give the 95% confidence interval for individual variability. The differential gender effect is particularly obvious in the inactive and very active groups. Inactive females gained about 7.1 kg more weight than the inactive men, and the very active women lost about 1.8 kg more weight than the very active men.
Table 3 presents the multiple regression analyses. Provided are the regression coefficients for the male and female equations and the 95% confidence intervals for each regression coefficient. Since the MANOVA results showed that NAS had a differential effect on weight changes, separate male and female regression models were developed to account for this gender difference.
The regression analyses produced different results for males and females. The independent variables found to have significant regression coefficients for men were T1 weight, time between T1 and T2 tests, NAS and T1 age. The significant independent variables for the women's model were T1 weight, time between T1 and T2 tests, and NAS. The T1 age regression coefficient for the female sample was−0.04, which was not significantly different from zero (F (1, 150)=0.96; P>0.05). Additionally, the interaction of T1 age with NAS, T1 weight, and Δ age were examined. The interactive terms added to the men's model accounted for 0.2% of T2 weight variance, which was not statistically significant (F (3, 334)=1.42; P>0.05). These interactive terms, along with T1 age also accounted for 0.2% of T2 weight variance of the female model, and, like the male model, this small change was not statistically significant (F (4, 147)=0.05; P>0.05). The measure of variability for these models is represented by the standard errors of estimate (±5.35 kg for men and±5.65 for women). These standard errors of estimate suggest that for the ‘average’ man and woman with the characteristics shown in Table 1 the estimated T2 weights may vary ±6% and ±8% of the T2 weights estimated by these models.
While the T1 weight regression coefficients, multiple correlations and standard errors of the male and female models were similar, the female regression coefficients for time difference and NAS were over twice the size of the male regression coefficients. Figure 2 shows these effects. Provided is the 5 y estimated weight change for low (NAS=0), moderate (NAS=5.5), and very high (NAS=10) levels of habitual physical activity. Individual variability for these estimated values are shown by the 95% confidence interval error bars. The T1 weight used in the regression models in Figure 2 were the male and female average weights of 82.2 and 67.0 kg. The age used in the men's equation was 36.1 y, the mean T1 age of the female sample. Figure 2 shows that the 5 y change for men and women for moderate activity levels was similar, but the males and females differed substantially for low and very high levels of activity. Inactive women can be expected to gain over twice the weight of inactive men while highly active women can be expected to lose almost three times the expected loss for highly active men.
The regression analyses showed longitudinal changes in weight of men and women are complex and due to initial weight, physical activity habit, aging and T1 age for men. To examine the effects of these independent variables on the longitudinal change in weight more closely, we used the male and female regression models to estimate the 5 y T2 weight for selected T1 weights, all NAS physical activity levels, and male ages of 25, 40 and 55 y. The differences between T2 and T1 weights were then normalized to a yearly change in weight (kg/y). Tables 4 and 5 show the projected yearly weight changes. The standard errors of prediction for the values in Tables 4 and 5 vary with each predicted value. The 95% confidence intervals for the ‘average’ predicted values are±0.11 kg/y for the men and ±0.18 kg/y for the women.
The significant negative T1 age coefficient for the men illustrated that older men gain less weight then younger men for the same T1 weight and level of physical activity. For example, the expected yearly weight gain of a man aged 25 y is 0.42 kg (∼0.9 lb) higher than a man 30 y his senior. Since T1 age was not a factor for women, Table 5 shows the expected weight change for level of exercise for women regardless of starting age.
These tables show that level of physical activity had a greater positive effect on weight change for higher T1 weight levels. For men, the influence of exercise increased with the age they started. To illustrate, if a 100 kg man exercises at an average NAS of 6, a 25 y old man can expect to lose about 0.15 kg per year while a man 30 y older loses nearly 0.6 kg per year. The estimated weight change for a 60 kg man following the same routine is a yearly gain of 0.73 kg at age 25, but only 0.31 if he started to exercise at age 55. In contrast, a 90 kg woman with a mean NAS of 6 would lose about 0.7 kg yearly while the weight loss rate of a 70 kg woman is only 0.2 kg per year.
The primary purpose of this study was to examine the influence of physical activity on longitudinal changes in body weight. Analysis of our data showed that changes in body weight are a function of several factors: initial weight, time between tests, gender and physical activity habit. Baseline age was a factor for the men only. Previous studies are in general agreement that age is positively associated with body weight and negatively associated with physical activity.18 However, there is no consensus in the literature about exercise affecting body weight differentially by age. Some studies with large cross-sectional28 and longitudinal29 data sets have indicated that the effect of exercise/fitness on weight and weight change does not differ by baseline age. Conversely, one large cross-sectional study showed that the weight control effect from exercise increased with age for both men and women,30 and a population-based study drawing on both cross-sectional and longitudinal data of men and women supported our finding that baseline age significantly added to the prediction of weight change in men only.31 The age coefficient (b=−0.05 kg) in the prediction equation of this population-based study was similar to the age coefficient in our male equation (b=−0.07 kg). These low coefficients may help explain the apparent contradictions in the literature about the effect or no effect from baseline age. If there is an age interaction, the age contribution is either insignificant or very small, particularly in comparison to starting weight and exercise level.
One explanation for the gender difference in the T1 age effect is the cross-sectional association between weight and age. The correlation between T1 age and weight for men was low, 0.06 (95% CI:−0.04–0.17) while the same correlation for the women was higher, 0.17 (95% CI: 0.02–0.32). To study this more closely, we examined the cross-sectional aging effect of weight with two large NASA databases used to study the role of aging in VO2max.22,23 The correlation between age and body weight for over 1500 NASA male employees (aged 25–70 y) was 0.05 (95% CI:−0.01 to 0.10). In contrast, the correlation found with 407 NASA women who ranged in age from 20 to 64 y was higher, 0.25 (95% CI: 0.16–0.34). This suggests that the failure to find a T1 aging effect with women is that of a gender difference in the cross-sectional aging effect on body weight. The T1 weight of women is somewhat reflective of their age. As a predictor of future weight (ie T2 weight) of women T1 weight is more potent than age.
Multiple regression of our data revealed that physical activity level is a significant source of weight change, but differential effects should be expected based on initial weight and gender. For the same level of physical activity, heavier men and women exhibit the most favorable changes in body weight. For equal weight loss over time, individuals with lower T1 weights need to exercise more than heavier individuals. The high value of T1 weight in predicting T2 weight may be explained by the fact that the energy cost of exercise is largely dependent on body weight. For example, metabolic calculations32 show that, in a mile of walking, a 90 kg person (198 lb) will burn about 439 kJ (105 kcal), but a 70 kg person (154 lb) walking the same distance will burn only 335 kJ (80 kcal). This may be one of the reasons why overweight people tend to lose a substantial amount of weight with exercise early on and then lose smaller amounts of weight over time. For example, a 90 kg person (198 lb) walking 20 miles per week burns enough energy to lose 0.45 kg (one pound) in about 1.7 weeks. The time required to lose the same weight with this walking regimen is about 1.9 weeks for an 80 kg individual (176 lb) and 2.2 weeks for a 70 kg person (154 lb). These calculations illustrate that, as the body weight drops, the rate of loss also will drop unless the individual progressively increases the walking mileage.
The role of gender is demonstrated in our equations by the influence of time between tests and habitual activity, both of which exert roughly twice the impact on women as compared to men. Tables 4 and 5 show that when men and women are matched for T1 body weight and physical activity, women tend to gain more weight with low level exercise (NAS=0–4), but they lose more weight with high level exercise (NAS=6–10). The reason for the gender effect is speculative; men may require more activity for weight control because of their greater relative leanness and the associated demand for higher metabolic activity to maintain their proportionately greater muscle mass. Previous research has shown that women reduce the amount and intensity of physical activity as they age at about twice the rate of aging men,33 but compared to men women can achieve weight loss and weight maintenance with exercise that is lower in intensity and duration.28,31,34 Empirical research also shows that women tend to gain more weight than men over time.2,5,9 Our models suggest that this is due primarily to the lower T1 female weight and possibly to a lower female activity habit.
The regression coefficients in our male/female models for estimating T2 weight from time between tests were 0.37 (men) and 0.70 (women). This suggests that, excluding the influence of the other variables, the aging effect is about 0.37–0.70 additional kilograms in body weight per year of aging. However, the individual weight change records with aging were highly variant, and the actual mean weight changes for the men and women in our sample were lower than these values. This suggests that aging alone does not explain future weight, and that other factors, such as activity habit, may be applied to alter or limit what may otherwise seem to be an inevitable aging consequence. Previous studies revealed that with aging, physical fitness (VO2max) tends to decline and body weight goes up, but these changes are due primarily to the typical age-related decrease in physical activity and secondarily to aging.22,23,35
One limitation to assessing the role of activity habit on weight change in this study was the use of self-report level of physical activity. While it would be better to have objective data, we have published data supporting the NAS scale's validity.22,23 The cross-sectional correlations with measured VO2max of the NASA employees measured in these studies were found to be 0.58 and 0.63 for males and females, respectively. Longitudinal analysis showed that reported changes in NAS were associated with changes in measured VO2max in the expected direction. The validity of the NAS and a similar scale with cross-sectional data has been reported by other investigators.36,37 Blair and associates38 showed that self-report level of physical activity was sensitive to detecting longitudinal changes in aerobic fitness.
In our study there were no dietary data reported. Although diet is obviously important for health, the practice of dieting has produced a very poor record in long-term weight loss and weight control. When the actual long-term effects of physical activity and dieting were compared in weight control studies, the contribution of dieting ranged from negligible to slight, but exercise emerged as the leading player.39,40 Skender and colleagues17 studied the 2 y effects of diet and exercise on weight loss. In their randomized experimental study they showed that exercise, not diet, was the determinate of long-term weight loss. They did find that diet produced a substantial 1 y weight loss. In the first 12 months the diet group lost 6.8 kg (15 lb) while the combined diet and exercise group lost nearly 9.1 kg (20 lb). In contrast, the exercise only group lost just 2.7 kg (6 lb). During the final 12 months, the diet group not only gained all the weight back, but they also gained an additional 0.9 kg (2 lb). During the second year the diet and exercise group gained back most of the weight they lost. The exercise only group maintained their weight loss during the second year. The longitudinal research published by Kasch and associates41 support the effect of habitual physical activity on weight control. They reported that the 25 y longitudinal weight loss of physically active men was 3.4 kg (−0.14 kg/y), while their sample of inactive men gained 3.2 kg (0.13 kg/y). Our results are consistent with these findings that physical activity is the key element in the maintenance of long-term weight loss.
The subjects in our study were not typical of the general population. They were motivated to the degree that they were long-term adherents to an employee wellness program (HRFP). Compared to the nationally representative NHANES data5 the HRFP subjects exhibited less weight gain over time, but their weight changes were similar to those reported by a longitudinal study of active members of the Aerobics Center in Dallas, Texas.29 The average yearly change in weight of the HRFP subjects was about 0.07 kg (0.15 lb) and 0.27 kg (0.60 lb) for the men and women, respectively, compared with the average yearly weight increase for the Aerobics Center sample of 0.08 kg (0.18 lb) for the men and 0.20 kg (0.44 lb) for the women. In general, the HRFP subjects tended to maintain a moderate to high level of physical activity. The mean NAS throughout the study shown in Table 1 (women=5.0, men=5.6) represents an activity equivalence of walking between 7.5 and 8.6 miles per week. Approximately 72% of the men and 52% of the women reported that at T2 they were exercising at a level higher than this (NAS>6). The special nature of the HRFP sample should be considered when making inferences from our data.
The cardinal lesson of this study is that a substantial but manageable amount of exercise is required for weight control. Our models show the habitual activity needed to maintain weight for 5 y in the average man and woman in our sample is represented by mean NASs of 6.0 and 5.6, respectively. A 10 y public health survey of 79 236 adult Americans showed that the exercise requirement to lose or maintain weight was a weekly habit of jogging at least an hour or walking at least 4 h.34 This amount of activity rates a 6 on our NAS. The joint consensus recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine is 30 min or more of moderate-intensity physical activity, such as walking, done on ‘most, preferably all, days of the week’.42 Assuming an exercise frequency of at least 5 days per week, this amount of activity also rates a 6 on our NAS. Notably, the joint consensus recommendation addressed the exercise requirement for overall good health, not just weight control, but the exercise requirement for weight control revealed in our study is essentially the same as that needed for health.
The literature shows that weight gain is common with advancing years, even for those who maintain a regular exercise habit. Two recent studies, one on cross-sectional survey data from a large group of male runners,43 the other on longitudinal data from members of a large health/fitness center,29 showed that long-term maintenance of the level of exercise and aerobic power attenuates but does not prevent aging-related weight gain. The authors of both studies suggested that without exercise there would be even greater weight gain, and preventing weight gain may actually require an increase in exercise with aging. These findings are not entirely incongruent with ours. Our subjects reported a substantial long-term increase in exercise habit. The men maintained a higher mean activity than the women, and that habit was associated with a non-significant weight gain after a mean of 5.7 y. The lower activity habit of the women was apparently not enough to prevent a gain of 1.4 kg over a shorter, 4.9 y, time period. Only those men and women who maintained a high level of exercise (NAS≥6.5) lost weight.
In conclusion, the ease of obtaining the predictor variables (body weight, age, gender and self-report physical activity), the high multiple correlations and low standard errors all suggest our models provide simple, practical and reasonably accurate means for estimating future weight. The only controllable variable in the models is habitual activity, substantiating the extreme importance of exercise in weight control.
Roberts J . Weight by height and age of adults Vital and Health Statistics, Series 11, no. 14 National Center for Health Statistics: Hyatsville, MD 1986.
Williamson D, Kahn HS, Remington PL, Anda RF . The 10-y incidence of over-weight and major weight gain in US adults Arch Intern Med 1990 150: 665–672.
Paffenbarger RJ, Hyde RT, Wing AL, Hsieh C . Physical activity, all cause mortality, and longevity of college alumni New Engl J Med 1986 314: 605–613.
Blair S, Shaten J, Brownell K, Collins G, Lissner L . Body weight change, all-cause mortality, and cause-specific mortality in the multiple risk factor intervention trial Ann Intern Med 1993 119: 749–757.
National Institutes of Health (NIH)/National Heart, Lung, and Blood Institute . Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity: the evidence report Obes Res 1998 6 (Suppl 2).
Lee I, Manson JE, Hennekens CH, Paffenbarger RS . Body weight and mortality, a 27-year follow-up of middle-aged men JAMA 1993 270: 2823–2828.
Klegal KM, Carroll MD, Kuczmarski RJ, Johnson CL . Overweight and obesity in the United States: prevalence and trends, 1960–1994 Int J Obes Relat Metab Disord 1998 22: 39–47.
National Institutes of Health (NIH) . First federal obesity clinical guidelines released NIH News Release 17 June 1998.
Wier LT, Jackson AS, Ayers GA . Longitudinal assessment of 10-y weight change in a large federal workforce. (Abstract) Med Sci Sports Exerc 1996 28 (Suppl): S43,252.
Grundy SM, Blackburn G, Higgins M, Lauer R, Perri MG, Ryan D . Physical activity in the prevention and treatment of obesity and its comorbidities: evidence report of independent panel to assess the role of physical activity in the treatment of obesity and its comorbidities Med Sci Sports Exerc 1999 31: 1493–1500.
Barlow CE, Kohl HW, Gibbons LW, Blair SN . Physical fitness, mortality, and obesity Int J Obes Relat Metab Disord 1995 19 (Suppl 4): S41–S44.
US Department of Health and Human Services (USDHHS) . Physical activity and health: a report of the Surgeon General US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion: Atlanta, GA 1996.
Foreyt JP, Goodrick GG . Attributes of successful approaches to weight loss and control Appl Prev Psychol 1994 3: 209–215.
Kayman S, Bruvold W, Stern JS . Maintenance and relapse after weight loss in women: behavioral aspects Am J Clin Nutr 1990 52: 800–807.
Klem ML, Wing RR, McGuire MT, Seagle HM, Hill JO . A descriptive study of individuals successful at long-term maintenance of substantial weight loss Am J Clin Nutr 1997 66: 239–246.
Pavlou KN, Krey S, Steffee WP . Exercise as an adjunct to weight loss and maintenance in moderately obese subjects Am J Clin Nutr 1989 49: 1115–1123.
Skender ML, Goodrick K, DelJunco DJ, Reeves RS, Darnell L, Gotto AM, Foreyt JP . Comparison of 2-year weight loss trends in behavioral treatments of obesity: diet, exercise, and combination interventions J Am Diet Assoc 1996 96: 342–346.
DiPietro L . Physical activity, body weight and adiposity: an epidemiologic perspective Exerc Sports Sci Rev 1996 86: 999–1004.
American College of Sports Medicine (ACSM) . ACSM's guidelines for exercise testing and prescription 5th edn.Williams and Wilkins: Baltimore, MD 1995.
Baumgartner TA, Jackson AS . Measurement for evaluation in physical education and exercise science 6th edn. Wm. C. Brown: Dubuque 1999.
Jackson AS, Blair SN, Mahar MT, Wier LT, Ross RM, Stuteville JE . Prediction of functional aerobic capacity without exercise testing Med Sci Sports Exerc 1990 22: 863–870.
Jackson AS, Beard EF, Wier LT, Ross RM, Stuteville JE, Blair SN . Changes in aerobic power of men ages 25–70 y Med Sci Sports Exerc 1995 27: 113–120.
Jackson AS, Wier LT, Aers GW, Beard EF, Stuteville JE, Blair SN . Changes in aerobic power of women, ages 20 to 64 y Med Sci Sports Exerc 1996 28: 884–891.
Jackson AS, Pollock ML . Generalized equations for predicting body density of men Br J Nutr 1978 40: 497–504.
Jackson AS, Pollock ML, Ward A . Generalized equations for predicting body density of women Med Sci Sports Exerc 1980 12: 175–182.
Pedhazur EJ . Multiple regression in behavioral research explanation and prediction 3rd edn. Harcourt Brace College: New York 1997.
Gettman L . Fitness testing In: American College of Sports Medicine Resource manual for guidelines for exercise testing and prescription, 2nd edn Lea & Febiger: Philadelphia, PA 1993 229–246.
Williamson D, Madans J, Anda RF, Kleinman JC, Kahn HS, Beyers T . Recreational physical activity and 10-year weight change in a US national cohort Int J Obes Relat Metab Disord 1993 17: 279–286.
DiPietro L, Kohl HW III, Barlow CE, Blair SN . Improvements in cardiorespiratory fitness attenuate age-related weight gain in healthy men and women: the Aerobics Center longitudinal study Int J Obes Relat Metab Disord 1998 22: 55–62.
DiPietro L, Williamson DF, Casperson CJ, Eaker E . The descriptive epidemiology of selected physical activities and body weight among adults trying to lose weight: the Behavioral Risk Factor Surveillance System survey, 1989 Int J Obes Relat Metab Disord 1993 17: 69–76.
French S, Jeffery RW, Forster JL, McGovern PG, Kelder SH, Baxter JE . Predictors of weight change over two years among a population of working adults: the Healthy Worker Project Int J Obes Relat Metab Disord 1994 18: 145–154.
American College of Sports Medicine (ACSM) . Appendix D: Metabolic calculations ACSM's guidelines for exercise testing and prescription Williams & Wilkins: Baltimore, MD 1985 269–287.
Folsom AR, Caspersen CJ, Taylor HL, Jacobs DR Jr, Luepker RV, Gomez-Marin O, Gillum RF, Blackburn H . Leisure time physical activity and its relationship to coronary risk factors in a population-based sample Am J Epidemiol 1985 121: 570–579.
Kahn H, Tatham LM, Rodriguez C, Calle EE, Thun MJ, Heath CW . Stable behaviors associated with adults' 10-year change in body mass index and likelihood of gain at the waist Am J Public Health 1997 87: 747–754.
Meredith C, Zackin MJ, Frontera WR, Evans WJ . Body composition and aerobic capacity in young and middle-aged endurance-trained men Med Sci Sports Exerc 1987 19: 557–563.
Heil D, Freedson PS, Ahlquist LE, Price J, Rippe JM . Non-exercise regression models to estimate peak oxygen consumption Med Sci Sports Exerc 1995 27: 599–606.
Jones N, Madrides L, Hitchcock C, Chypcar T, McCartney N . Normal standards for an incremental progressive cycle ergometer test Am Rev Resp Dis 1985 131: 700–708.
Blair S, Kohl HW, Barlow MS, Paffenbarger RS, Gibbons LW, Macera CA . Changes in physical fitness and all-cause mortality: a prospective study of healthy and unhealthy men JAMA 1995 273: 1093–1098.
Avila P, Hovell MF . Physical activity training for weight loss in Latinas: a controlled trial Int J Obes Relat Metab Disord 1994 18: 476–482.
Bild D, Sholinsky P, Smith DE, Lewis CE, Hardin JM, Burke GL . Correlates and predictors of weight loss in young adults: the CARDIA study Int J Obes Relat Metab Disord 1996 20: 47–55.
Kasch FW, Boyer JL, VanCamp SP, Verity LS, Wallace J . The effect of physical activity and inactivity on aerobic power in older men (a longitudinal study) Phys Sports Med 1990 18: 73–81.
Pate R, Pratt M, Blair SN, Haskell WL, Macera CA, Bouchard C, Buchner D, Ettinger W, Health GW, King AC, Kriska A, Leon AS, Marcus BH, Morris J, Paffenbarger RS, Patrick K, Pollock ML, Rippe JM, Sallis J, Wilmore JH . Physical activity and public health—a recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine JAMA 1995 273: 402–407.
Williams PT . Evidence for the incompatibility of age-neutral overweight and age-neutral physical activity standards for runners Am J Clin Nutr 1997 65: 1391–1396.
This work was partially supported by a Minority Scientist Development Award from the American Heart Association and with funds contributed by the AHA, Puerto Rico Affiliate.
Appendix NASA activity scale (NAS)
Appendix NASA activity scale (NAS)
Code for physical activity status
Select the appropriate number (0–10) in the space for physical activity code according to which of the following best describes your activity level for the previous month:
Do not exercise regularly, ie
Avoid walking or exertion, eg always use elevator, drive whenever possible instead of walking.
Walk for pleasure, routinely use stairs, occasionally exercise sufficiently to cause heavy breathing or perspiration.
Participate regularly in recreation or work requiring modest physical activity, such as golf, horseback riding, calisthenics, table tennis, bowling, weight-lifting or yard work.
10–60 min per week.
Over 1 h per week.
Participate regularly in heavy physical exercise, eg, running or a comparable activity such as brisk walking, indoor biking, swimming, cycling, rowing, skipping rope, running in place, or engaging in vigorous aerobic exercise such as tennis, basketball, or handball.
Run less than 1 mile per week or walk less than 1.3 miles per week or spend less than 30 min per week in comparable physical activity.
Run 1–5 miles per week or walk 1.3–6.9 miles per week or spend 30–60 min per week in comparable physical activity.
Run 6–10 miles per week or walk 7–13.9 miles per week or spend 1–3 h per week in comparable physical activity.
Run 11–15 miles per week or walk 14–20 miles per week or spend 4–6 h per week in comparable physical activity.
Run 16–20 miles per week or walk 21–26.9 miles per week or spend 7–9 h per week in comparable physical activity.
Run 21–25 miles per week or walk 27–33.9 miles per week or spend 10–12 h per week in comparable physical activity.
Run over 25 miles per week or walk over 34 miles per week or spend over 12 h per week in comparable physical activity.
About this article
Cite this article
Wier, L., Ayers, G., Jackson, A. et al. Determining the amount of physical activity needed for long-term weight control. Int J Obes 25, 613–621 (2001). https://doi.org/10.1038/sj.ijo.0801586
- weight control
- body weight
- physical activity
Accuracy of Nonexercise Prediction Equations for Assessing Longitudinal Changes to Cardiorespiratory Fitness in Apparently Healthy Adults: BALL ST Cohort
Journal of the American Heart Association (2020)
Physiological effects of weightlessness: countermeasure system development for a long-term Chinese manned spaceflight
Frontiers of Medicine (2019)
Comparison of non-exercise cardiorespiratory fitness prediction equations in apparently healthy adults
European Journal of Preventive Cardiology (2019)
The Examination of Patient-Reported Outcomes and Postural Control Measures in Patients With and Without a History of ACL Reconstruction: A Case Control Study
Journal of Sport Rehabilitation (2018)
Comparative Effectiveness of Plantar-Massage Techniques on Postural Control in Those With Chronic Ankle Instability
Journal of Athletic Training (2017)