Abbreviations BMI: body mass index; CRF: cardiorespiratory fitness.
Human evolution has been dependent on a physically active lifestyle supplemented with nutritional fortification from uncultivated vegetables and wild game, exclusive of dairy products and high animal fat intake . Our genetic constitution has remained unchanged over the past 50 000 years ; thus, it is likely that an evolutionary mismatch in the patterns of nutrient intake and physical activity between our hunter–gatherer ancestors and those of modern industrialized societies underlies the global burden of chronic diseases such as diabetes and cardiovascular disease. Contemporary living environments in developed countries are characterized by low daily energy expenditure and an abundant and inexpensive calorie-dense food supply, making positive energy balance common. Indeed, the prevalence of sedentary living habits and obesity (body mass index [BMI] ≥30 kg/m2) among US adults is ≈30% for each [3,4]. Both conditions are associated with premature mortality, increased risk of chronic disease morbidity, and functional impairments [5,6] and both have been identified as major modifiable risk factors for coronary heart disease [7,8].
There are major challenges to disentangling the complex multifactorial etiology of adiposity, physical activity and health outcomes. Physical inactivity could be an antecedent or a consequence of obesity [9,10], but it is difficult to know how much of the increased morbidity and mortality seen in obese adults results from excessive body fat, physical inactivity, or both. An expert panel  recently concluded that available evidence  suggests that overweight and obese adults who achieve adequate levels of physical activity or cardiorespiratory fitness (CRF) have lower risk of morbidity and mortality than do their normal-weight but sedentary or unfit peers. Only recently, however, have a small number of prospective studies systematically examined the independent and joint associations of physical activity or CRF and adiposity with health outcomes.
In this brief report, we draw on recently published observational studies to illustrate the joint associations of physical activity, CRF, and adiposity with the risk of adverse health outcomes. Due to space limitations, we focus on studies with mortality outcomes. We begin with a comment on issues pertaining to exposure assessment that may underlie differences in the findings between studies on physical activity, CRF, adiposity, and health outcomes.
Assessing physical activity, fitness, and adiposity exposures
Differences among studies in the reported pattern and strength of association for activity and adiposity exposures and health outcomes depend, in part, on the methods of exposure assessment. Physical activity is a complex multidimensional behavior that is difficult to assess in free-living populations and for which a gold standard measurement does not exist. As such, a variety of methods have been used to assess physical activity and these measurements have a broad range of accuracy, reproducibility, and feasibility [5,13]. For example, self-administered or interview-based questionnaires have relatively low cost and administrative burden and can be used to obtain a crude categorization of activity status (e.g. sedentary versus active) or more detailed descriptions of activities (e.g. type, duration, frequency) and their estimated energy cost (e.g. kcal/wk). Issues pertaining to the calibration and transportability of a questionnaire's component physical activity domains and items across population subgroups (e.g. elderly, racial or ethnic minorities, women) must be considered when generalizing associations between health outcomes and activity levels obtained from a specific questionnaire . Response biases (e.g. recall bias, social desirability bias) limit the accuracy of self-reported physical activity exposures, particularly when used to estimate activity-related energy expenditure . Alternatively, direct monitoring of body movement and related energy expenditure can be performed using electronic motion sensors, heart rate monitors, portable indirect calorimeters, or some combination thereof . Direct physical activity assessment is not influenced by response biases and some methods provide an accurate quantification of the intensity of physical activity. Thus, direct monitoring may improve the accuracy of free-living physical activity assessment compared with self-reported methods. Lack of information on the type of activity being performed, potential changes in habitual physical activity behavior as a consequence of monitoring (e.g. reactivity), and the associated costs and administrative burden, however, have precluded using direct monitoring in most large-scale studies of health outcomes, at least to this point.
Physical fitness is a set of physiological attributes that are enhanced through participation in regular physical activity . The major component of physical fitness that has been related to adverse health risks is CRF or ‘aerobic power’, as assessed by maximal or submaximal exercise testing. Although CRF is influenced by several factors including age, sex, health status, and genetics, the principal modifiable determinant is habitual physical activity levels. CRF is an objective and reproducible measure associated with recent physical activity patterns, accounting for up to 70% of the variance in detailed physical activity records . As CRF is less prone to misclassification due to response biases or behavior reactivity, it may better reflect the adverse consequences of a sedentary lifestyle than do self-reported or directly monitored activity exposures. For example, in one study  the age-adjusted relative risks for all-cause mortality among 1263 men with type 2 diabetes were 1.8 for physical inactivity but 2.9 for low CRF.
In most large population-based studies on adiposity and health outcomes, BMI (kg/m2) has been used as the exposure, and has generally been computed from reported heights and weights obtained from responses to a baseline health questionnaire . As with self-reported physical activity assessment, categorizing weight or adiposity using BMI computations from reported heights and weights is subject to response biases and thus misclassification . An important assumption that underpins the positive relationship between body size and health risk is that adiposity (e.g. excessive body fat) is the underlying putative risk factor . This leads to a second assumption regarding BMI as an index of health risk: that BMI provides a precise measure of body composition (e.g. fat mass, fat-free mass, and fat distribution). Body composition varies considerably with sex, age, race, and ethnicity, and differences in adiposity are not quantified with a criterion-referenced BMI scale [19,20]. One of the best examples of an inappropriate use of BMI is a report  showing that National Football League players are nearly all overweight, if not obese. BMI is an inappropriate measure of body composition because of the large muscle mass observed in exercise-trained athletes such as football players [22,23]. Feasibility issues and lack of a consensus definition for identifying high-risk exposures have limited the use of direct measures of body composition such as hydrodensitometry, dual X-ray absorptiometry, or computed tomography scans in large studies of health outcomes. Differential misclassification of self-reported physical activity or adiposity levels occurs in population subgroups; for example, obese individuals overestimate their physical activity levels and underestimate their weight, and sedentary individuals tend to overestimate their physical activity habits to a greater extent than do those who are regularly active.
Another issue that may influence differences between studies in the reported associations of physical activity and adiposity exposures with health outcomes is the distribution of each exposure within the study population sample. For example, suppose two studies examine associations between adiposity, defined by BMI levels, and a health outcome that is etiologically related to adiposity. If in one study a large number of participants are exposed to moderate and severe obesity (e.g. BMI ≥35 kg/m2) and in the other study the BMI distribution is truncated to normal and overweight phenotypes, the former may detect a stronger association for BMI with the health outcome. Likewise, when investigating an outcome that is causally associated with physical activity, a study that has a broad distribution of physical activity levels or related energy expenditure (or one that uses CRF as the exposure) may detect a stronger association with the outcome than a study that includes a narrow distribution of physical activity or energy expenditure. The influence of exposure distributions may be even greater when examining joint associations of adiposity and physical activity (or CRF) with health outcomes.
Current evidence on the joint associations of adiposity and physical activity with mortality
Recent prospective studies have systematically examined the independent and joint associations of physical activity or CRF and adiposity with health outcomes. Space limitations preclude a critical and exhaustive review of this literature here, and readers are referred elsewhere in this regard [12,24•]. Table 1 shows the results of seven large observational studies on physical activity, CRF, adiposity, and mortality that have been published since 2004 [25–28,29•,30,31]. In these studies, the highest mortality risk is observed in individuals who are obese and unfit or physically inactive. When examining mortality rates within a given BMI group, individuals with higher CRF or physical activity have a lower relative risk than their unfit and inactive peers. In some but not all of these studies, mortality risk was lower in the obese who were active or fit than in normal-weight individuals who were inactive or unfit. These observations are made in women and men, in high-risk populations with clinically manifest disease, when total or abdominal adiposity is cross-tabulated with physical activity or CRF, and they persist after accounting for several potential confounding factors.
The degree to which higher levels of physical activity and CRF modify obesity-related mortality risk varies among studies. For example, in the Aerobics Center Longitudinal study (ACLS), [28,31–33] higher CRF eliminates the increased mortality risk in obese adults, whereas in other studies [26,29•] higher levels of self-reported physical activity have attenuated but not completely eliminated the excess mortality that is associated with obesity. Possible explanations for these differences may be due to chance or may be related to exposure measurement issues that were discussed earlier. In the ACLS, CRF is assessed with maximal exercise testing and BMI is from measured heights and weights; whereas in other studies, physical activity is assessed by questionnaire [26,29•] and BMI is from self-reported values . Overweight and obese individuals underreport their weight and overestimate their physical activity level, which may result in an underestimation of the true association between physical activity and mortality risk. Another complex issue is the gene–environment interaction for body fat and health risk. Although physical activity is a major environmental factor that influences the degree to which bad genes express unfavorable phenotypes, CRF may be a better indicator of the combination of genetics and behavior and thus stronger than physical activity as a predictor of health outcomes under a variety of circumstances, including obesity.
We believe, however, that the health focus should not be whether mortality risk is attenuated or completely eliminated by higher levels of physical activity or CRF. Instead, attention should be given to the overwhelming trend across available data for lower mortality risk in all members of the population, normal weight or obese, who are active and fit than in those who are sedentary or unfit [5,12]. Both for clinical interventions and for public health programs, the focus should be on healthful lifestyle behaviors. Just as weight is monitored in some manner at each physician visit, so should attention be given to monitoring and promoting adequate levels of physical activity and CRF and a healthful diet. Everyone should be encouraged to follow a dietary pattern that emphasizes fruits, vegetables, and whole grains; limits intake of saturated and trans fat; and includes a wide variety of foods. Everyone also should be encouraged to engage in physical activities that are at least moderately intense for 30 minutes or longer on 5 or more days per week. It is now recognized that physical activity–related energy expenditure, or the total dose of physical activity, is more important for health benefits than is the specific type of physical activity performed (e.g. walking, running, cycling) . Moderate-intensity physical activity is associated with an energy expenditure of 3 kcal/kg/h or more (≥65% maximal heart rate), which for most individuals is equivalent to brisk walking at a pace of 3.5 miles per hour, or other activities that increase heart rate and breathing well above resting levels but do not cause one to strain (e.g. you can still talk comfortably during the activity). Observational  and experimental [35,36] data suggest that this dose of physical activity also is a sufficient stimulus to achieve moderate levels of CRF in apparently healthy adults. These recommendations clearly lead to improved health and function and will provide benefits whether or not they result in weight loss.
Recent epidemiologic studies suggest that higher levels of physical activity or CRF may offset much of the excess mortality risk that is associated with overweight and obesity in adults. Physical activity and body composition each are complex issues that are difficult to assess with precision in large-scale studies. The relatively crude self-reports of physical activity or of height and weight for BMI computation that typically are used in epidemiologic studies are prone to misclassification and are likely to underestimate the true associations of physical activity or adiposity with health outcomes. Misclassification of physical activity or adiposity levels may be particularly high among individuals who are sedentary or who are obese. This limitation of physical activity studies has been at least partially overcome by studies in which the exposure has been an objective measure of CRF. Although there is a genetic component to fitness, as for virtually anything else one can measure in humans including adiposity, the major modifiable determinant of CRF is physical activity over the weeks or months prior to the fitness assessment. Therefore, we conclude that to evaluate the independent and combined exposures of physical activity and BMI on health outcomes, it is necessary to develop new projects in which the exposures are assessed accurately. Although laboratory measures of fitness or body composition are more costly and logistically challenging than questionnaire-based physical activity or BMI assessment, we believe that fitness and body composition assessments should be included more often in epidemiologic investigations. A newer approach to objective assessment of physical activity involves using accelerometers, although this too is more costly and burdensome than self-report questionnaires. Just as new and more costly technology is being used in epidemiologic studies to improve the quantification of, for example, cardiovascular disease risk predictors , we believe that it is important to make the additional effort to include accurate measures of physical activity and adiposity in epidemiologic studies to improve the precision of estimating associations between these exposures and health outcomes. It is only by use of accurate assessment of activity and body habitus that we will be able to advance our understanding of the crucial public health issues related to these topics.
We thank Melba Morrow, MA, for editorial assistance.
References and recommended reading
Papers of particular interest, published within the annual period of review, have been highlighted as:
• of special interest
•• of outstanding interest
Additional references related to this topic can also be found in the Current World Literature section in this issue (p. 643).
1 Eaton SB, Konner M, Shostak M. Stone agers in the fast lane: chronic degenerative diseases in evolutionary perspective. Am J Med 1988; 84:739–749.
2 Macaulay V, Richards M, Hickey E, et al
. The emerging tree of West Eurasian mtDNAs: a synthesis of control-region sequences and RFLPs. Am J Hum Genet 1999; 64:232–249.
3 Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends in obesity
among US adults, 1999–2000. JAMA 2002; 288:1723–1727.
4 Ham SA, Yore MM, Fulton JE, Kohl HW. Prevalence of no leisure-time physical activity
: 35 states and the District of Columbia, 1988–2002. MMWR Morb Mortal Wkly Rep 2004; 53:82–86.
5 US Department of Health and Human Services. Physical activity
and health: a report of the Surgeon General. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease
Prevention and Health Promotion; 1996.
6 National Institutes of Health. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity
in adults: the evidence report. Rockville, MD: National Institutes of Health; 1998.
7 Fletcher GF, Blair SN, Blumenthal J, et al
. Position statement – statement on exercise
– benefits and recommendations for physical activity
programs for all Americans: a statement for health professionals by the committee on Exercise
and Cardiac Rehabilitation of the Council on Clinical Cardiology, American Heart Association. Circulation 1992; 86:340–344.
8 Eckel RH. Obesity
and heart disease: a statement for healthcare professionals from the Nutrition Committee, American Heart Association. Circulation 1997; 96:3248–3250.
9 Di Pietro L, Dziura J, Blair SN. Estimated change in physical activity
level (PAL) and prediction of 5-year weight change in men: the Aerobics Center Longitudinal Study. Int J Obes Relat Metab Disord 2004; 28:1541–1547.
10 Petersen L, Schnohr P, Sorensen TI. Longitudinal study of the long-term relation between physical activity
in adults. Int J Obes Relat Metab Disord 2004; 28:105–112.
11 Grundy SM, Blackburn G, Higgins M, et al
. Physical activity
in the prevention and treatment of obesity
and its comorbidities. Med Sci Sports Exerc 1999; 31:S502–S508.
12 Blair SN, Brodney S. Effects of physical inactivity and obesity
on morbidity and mortality
: current evidence and research issues. Med Sci Sports Exerc 1999; 31:S646–S662.
13 LaMonte MJ, Ainsworth BE, Reis JP. Measuring physical activity
. In: Zhu W, Woods T, editors. Measurement theory and practice in kinesiology. Champaign, IL: Human Kinetics; 2006. pp. 237–272.
14 LaMonte MJ, Ainsworth BE. Quantifying energy expenditure and physical activity
in the context of dose response. Med Sci Sports Exerc 2001; 33:S370–S378.
15 Paffenbarger RS Jr, Blair SN, Lee I-M, Hyde RT. Measurement of physical activity
to assess health effects in free-living populations. Med Sci Sports Exerc 1993; 25:60–70.
16 Wei M, Gibbons LW, Kampert JB, et al
. Low cardiorespiratory fitness
and physical inactivity as predictors of mortality
in men with type 2 diabetes. Ann Intern Med 2000; 132:605–611.
17 Nieto-Garcia FJ, Bush TL, Keyl PM. Body mass definitions of obesity
: sensitivity and specificity using self-reported weight and height. Epidemiology 1990; 1:146–152.
18 Bray GA, Gray DS. Obesity
part I: pathogenesis. West J Med 1988; 149:429–441.
19 Jackson AS, Stanforth PR, Gagnon J, et al
. The effect of sex, age and race on estimating percentage body fat from body mass index: the Heritage Family Study. Int J Obes Relat Metab Disord 2002; 26:789–796.
20 Adams TD, LaMonte MJ, Gress RE, Hunt SC. Gender differences in percentage body fat for a given BMI extend into the severely obese [abstract]. Obes Res 2003; 11(Suppl):A131.
21 Harp JB, Hecht L. Obesity
in the National Football League. JAMA 2005; 293:1061–1062.
22 Wilmore JH, Haskell WL. Body composition and endurance capacity of professional football players. J Appl Physiol 1972; 33:564–567.
23 Welham WC, Behnke AR. The specific gravity of healthy men: body weight divided by volume and other characteristics of exceptional athletes and of naval personnel. JAMA 1942; 118:498–501.
24• Weinstein AR, Sesso HD. Joint effects of physical activity
and body weight on diabetes and cardiovascular disease. Exerc Sport Sci Rev 2006; 34:10–15. A concise overview of selected prospective cohort studies that systematically examined the combined association of self-reported physical activity
and BMI with either cardiovascular events or diabetes incidence. Recommendations for future research in this area are provided.
25 Stevens J, Evenson KR, Thomas O, et al
. Associations of fitness and fatness with mortality
in Russian and American men in the Lipids Research Clinics Study. Int J Obes Relat Metab Disord 2004; 28:1463–1470.
26 Hu FB, Willett WC, Li T, et al
. Adiposity as compared with physical activity
in predicting mortality
among women. N Engl J Med 2004; 351:2694–2703.
27 Hu G, Tuomilehto J, Silventoinen K, et al
. Joint effects of physical activity
, body mass index, waist circumference and waist-to-hip ratio with the risk of cardiovascular disease among middle-aged Finnish men and women. Eur Heart J 2004; 25:2212–2219.
28 Church TS, LaMonte MJ, Barlow CE, Blair SN. Cardiorespiratory fitness
and body mass index as predictors of cardiovascular disease mortality
among men with diabetes. Arch Intern Med 2005; 165:2114–2120.
29• Hu G, Tuomilehto J, Silventoinen K, et al
. The effects of physical activity
and body mass index on cardiovascular, cancer and all-cause mortality
among 47 212 middle-aged Finnish men and women. Int J Obes Relat Metab Disord 2005; 29:894–902. A large well established cohort study that examined the risk for several fatal endpoints in relation to the independent and combined effects of self-reported physical activity
and BMI that was computed from measured heights and weights in women as well as men. The study highlights that these associations are robust against adjustment for several potential confounding factors, exclusion of early deaths, and advanced age.
30 Holmes MD, Chen WY, Feskanich D, et al
. Physical activity
and survival after breast cancer diagnosis. JAMA 2005; 293:2479–2486.
31 Katzmarzyk PT, Church TS, Janssen I, et al
. Metabolic syndrome, obesity
, and mortality
: impact of cardiorespiratory fitness
. Diabetes Care 2005; 28:391–397.
32 Lee CD, Blair SN, Jackson AS. Cardiorespiratory fitness
, body composition, and all-cause and cardiovascular disease mortality
in men. Am J Clin Nutr 1999; 69:373–380.
33 Blair SN, Barlow CE, Kampert JB, et al
. Joint association of cardiorespiratory fitness
and body mass index with mortality
in women. Med Sci Sports Exerc 2005; 37(5 Suppl):S284–S285.
34 Stofan JR, DiPietro L, Davis D, et al
. Physical activity
patterns associated with cardiorespiratory fitness
and reduced mortality
: the Aerobics Center Longitudinal Study. Am J Public Health 1998; 88:1807–1813.
35 King AC, Haskell WL, Taylor CB, et al
. Group vs home-based exercise
training in healthy older men and women: a community-based clinical trial. JAMA 1991; 266:1535–1542.
36 Duncan GE, Anton SD, Sydeman SJ, et al
. Prescribing exercise
at varied levels of intensity and frequency: a randomized trial. Arch Intern Med 2005; 165:2362–2369.
37 Bild DE, Bluemke DA, Burke GL, et al
. Multiethnic study of atherosclerosis: objectives and design. Am J Epidemiol 2002; 156:871–881.