Journal Logo


Physical Fitness, Adiposity, and Metabolic Risk Factors in Young College Students


Author Information
Medicine & Science in Sports & Exercise: June 2010 - Volume 42 - Issue 6 - p 1039-1044
doi: 10.1249/MSS.0b013e3181c9216b
  • Free


Approximately 31% of students in the U.S. college population are overweight or obese (1), and both men and women appear to gain weight during their freshman year (9,24). This time in college coincides with the beginning of the trajectory of the nearly twofold increase in overweight and obesity in the general U.S. population from 34.3% at ages 12-19 yr to 57.1% at ages 20-30 yr (22). Concurrent with this rise in obesity in youth, metabolic risk factors for cardiovascular disease and type 2 diabetes are increasingly apparent in adolescents and young adults. The prevalence of these risk factors also increases dramatically between the ages of 12-19 and 20-30 yr (8). However, few studies have systematically examined these risk factors in a college population that straddles the transition from adolescence to adulthood.

Physical activity and physical fitness are strong determinants of health outcomes. In fit individuals, adiposity decreases (26), muscle health is enhanced through increased insulin sensitivity (13), and there is decreased circulating levels of inflammatory mediators (17). In total, physical activity and fitness confer a lower metabolic risk profile in youth (3). It is estimated that 40%-50% of college students are physically inactive (1,18), and there are no recent estimates of their physical fitness levels. There is some evidence that obese youth who are fit demonstrate lower metabolic risk factors and systemic inflammation than unfit obese youth and exhibit levels similar to normal weight fit children (7,11,12). Although body mass index (BMI) was examined in these studies as a measure of "fatness," the results are suggestive that physical fitness may decrease metabolic stress independent of adiposity.

In this study, we examined the relative impact of fitness and adiposity on metabolic risk factors in a cohort of adolescents during their early college years. The purpose of this study was threefold: 1) to assess anthropometry, physical fitness, and metabolic risk factors (specifically serum lipids and glucose) in college students; 2) to determine whether physical fitness or percent body fat is associated with these risk factors; and 3) to elucidate whether physical fitness alone or in conjunction with body fatness can further influence metabolic risk factors in college students.


In 1998, the Tufts Longitudinal Health Study (TLHS) began to follow the health and health-related behaviors of undergraduate university students. From 1998 to 2007, members of the incoming freshmen class were mailed a 43-item Health Behavior Survey (9) and an informed consent form in late July/August before arriving at Tufts (baseline). Consenting students were invited to an annual April TLHS assessment approximately 8 months later (assessment day) to repeat the Health Behavior Survey, to complete additional diet and health questionnaires, to participate in anthropometric and physical fitness measurements, and to give a blood sample. Blood samples were collected in 2000, 2001, 2002, 2004, 2006, and 2007, and all data were pooled for this cross-sectional analysis. All procedures were performed in accordance with standards developed for research with human subjects and approved by the Tufts University institutional review board.


Anthropometric variables included weight, height, BMI, and body composition. Height was measured without shoes using a portable stadiometer (Model 214; Seca Weighing and Measuring Systems, Hanover, MD), with the head in the Frankfurt plane made with a right angle height procedure (19). It was recorded to the closest 1/8th inch. Weight was measured without shoes in light clothing on a portable balance beam scale (Healthometer, Boca Raton, FL) and recorded to the closest 0.25 lb. These students are entering young adulthood and span the age range where BMI is determined from either the Centers for Disease Control growth charts (4) or the adult definition of overweight. To categorize students as overweight or obese, we used age- and gender-specific BMI cutoffs (85th and 95th percentiles for overweight and obese, respectively) for students 20 yr and younger. For students older than 20 yr, the adult definition of overweight as greater than or equal to 25 kg·m−2 and of obese as greater than or equal to 30 kg·m−2 was used.

For calculation of total fat mass and percent body fat, total body resistance (ohms) and reactance (ohms) were measured with the subject lying supine using a bioelectrical impedance analyzer with four surface electrodes as previously described (BIA model 101; RJL Systems, Detroit, MI) (27). Fat-free mass was then calculated using ethnicity- and gender-specific prediction equations for men and women using a two-compartment model: fat mass was calculated by subtraction from total body mass and used along with total body mass to calculate percent body fat (19).

Physical fitness.

To assess physical fitness, we used the Queens College Step Test protocol (20). Students were asked to step onto and off of a platform (height = 16.25 inches) at a rate synchronized to the tone of a metronome for 3 min (metronome set to appropriate steps per minute: women at 88 steps per minute and men at 96 steps per minute). Within 5 s of completing the test, students had their pulse taken for 15 s using the radial artery, and standard equations were used for estimation of maximal oxygen uptake (V˙O2max) (20). V˙O2max results were then assigned to "low," "medium," or "high" fitness on the basis of criterion-referenced standards from a similar age group (6).

Blood draw and blood lipid profiles.

Blood was drawn after a 12-h overnight fast. Serum lipid profiles (total cholesterol, HDL cholesterol, and triglycerides) and glucose were measured by ACE™ Clinical Chemistry Systems (Schiapparelli Biosystems, Fairfield, NJ) using standard reagent kits. LDL cholesterol was determined by the Friedewald equation (14).


All variables were checked for normality and logarithmically transformed where necessary (triglycerides). Data were stratified by gender before analysis. After initial bivariate correlation analyses, multiple linear regression analysis was used to assess whether either physical fitness or body fatness was associated with altered serum glucose or lipids. The initial models included blood lipids (total cholesterol, triglycerides, HDL, or LDL) or glucose as the dependent variables and age, ethnicity, and either percent body fat or fitness as independent variables. Age and race/ethnicity data were collected by self-report during the spring assessment and used as covariates as they often relate to body fat percentage and physical fitness. Ethnicity was categorized into "white" and "nonwhite" because of the small percentage of nonwhite groups in this population. To examine whether percent body fat or fitness was more strongly associated with outcomes, we compared the beta coefficients of these variables (controlling for any significant interactions where necessary). Finally, an F-test was used to determine whether the addition of fitness to the models that already included percent body fat increased the ability of the model to predict blood glucose and lipid levels.

Logistic regression analyses were performed to determine whether percent body fat or fitness was predictive of "at-risk" levels of serum glucose and blood lipids (glucose >100 mg·dL−1, triglycerides >200 mg·dL−1, cholesterol >240 mg·dL−1, LDL >160 mg·dL−1, and HDL <40 mg·dL−1 for men and <50 mg·dL−1 for women). Log-likelihood tests were used to determine whether the addition of physical fitness to the model with percent body fat further contributed to any changes in serum glucose or blood lipids.

Dichotomous variables were created to define fatness using percent body fat (>23% for women and >19% for men as considered above "ideal" range [25]). The physical fitness tertiles ("low," "medium," and "high" fitness [6]) were combined into a dichotomous variable where low levels are considered "unfit" and medium and high levels are considered "fit." Using the categorical variables for both body fatness and physical fitness, students were grouped into four categories: 1) "fit and not fat," 2) "fit and fat," 3) "not fit and not fat," and 4) "not fit and fat." Multiple regression analyses were used to determine whether the groups that include fit students have healthier serum glucose and blood lipids compared with groups that include "unfit" regardless of body fatness. All regression analyses controlled for age and ethnicity.

All analyses were performed by SPSS statistical software (SPSS 14.0; Chicago, IL). Results are presented as means ± SD, unless otherwise noted. Significance was set at P < 0.05.


The mean age of students examined in this study was 19.4 ± 1.1 yr (median = 19.0; n = 564, 65.7% women). Approximately 23.7% of students who participated were nonwhite (11.5% Asian/Pacific American, 3.7% black/African, 3.7% Hispanic/Latin American, 4.2% multiracial, 0.6% other).

Table 1 presents the average BMI and body composition of the subjects compared with U.S. averages collected through the National Health and Nutrition Examination Survey (NHANES) (5) as well as the recommended ranges for BMI (2) and body fatness. Overweight and obesity was 16.2% among students, notably lower than the national average for similar ages. Using predictive V˙O2max testing, male college students had higher physical fitness compared with U.S. norms, whereas women demonstrated no difference (Table 1) (6).

Anthropometrics and physical fitness of college students participating in the Tufts Longitudinal Health Study (TLHS) relative to recommendations and U.S. data for this age group.

Nearly one-quarter of students presented at-risk levels of HDL, and one-third had LDL cholesterol levels above the recommendation (Table 2). Triglycerides were above the recommendation in 11% of the students. Men had lower total cholesterol and HDL levels compared with women.

Serum glucose and lipid profiles of college students.

Bivariate correlation analyses indicated that having a higher percent body fat was positively associated with higher cholesterol, triglycerides, and LDL and lower HDL levels in both men and women (data not shown; P < 0.05). Conversely, fitness was inversely associated with elevated glucose, total cholesterol, triglycerides, and LDL for both genders (data not shown; P < 0.05).

When fitness and body fat were analyzed separately in linear regression analyses, higher percent body fat was associated with increased total cholesterol and LDL levels in both men and women and with increased triglyceride levels and decreased HDL levels in women (Table 3). Greater fitness was associated with higher HDL and lower triglyceride levels in women and lower serum glucose in men. On the basis of comparisons of the beta coefficients for the models with fitness and percent body fat, percent body fat is a better predictor of serum triglyceride levels in women. Adding fitness to the model with percent body fat improved the model for HDL in women (fitness was nearly significant at P = 0.052) and the model for serum glucose in men (fitness is significant at P = 0.009).

Fitness and fatness as predictors of serum glucose and blood lipids in college students.

The results of logistic regression analyses to examine at-risk levels of blood demonstrated that in women, higher percent body fat was associated with elevated triglycerides (β = 0.14, P < 0.01), and fitness was inversely associated with elevated total cholesterol (β = −0.09, P = 0.03) and suboptimal HDL (β = −0.08, P = 0.03). Adding fitness with body fatness did not improve the model for triglycerides but did with total cholesterol and HDL (χ2 = 4.89 and 4.73, respectively, P < 0.05). No differences were found in men.

To further demonstrate how fitness and fatness in combination impact blood lipids, we grouped students into four categories on the basis of a combination of their physical fitness (fit/unfit) and percent body fat (not fat/fat) as described in the Methods section. Within these categories, 31.7% were "fit and not fat," 36.3% were "fit and fat," 9.2% were "unfit and not fat," and 22.8% were "unfit and fat."

Gender had an impact on the effect of fitness and fatness on glucose, cholesterol, and HDL and LDL (P < 0.05). The "fit/not fat" groups had a better metabolic risk profile compared with the "unfit/fat" group: "fit/not fat" men had lower glucose, total cholesterol, and LDL levels, and "fit/not fat" women had lower cholesterol, triglycerides, and LDL and higher HDL (data not shown; P < 0.05). We found no differences between the "fit/fat" and the "unfit/not fat" groups. In men, the "fit/fat" group had lower LDL levels compared with the "unfit/fat" group (P < 0.05), and in women, the "fit/fat" group had lower triglyceride and higher HDL levels compared with the unfit/fat group (data not shown; P < 0.01).


Attending college is a major time of transition for adolescents who are establishing independent physical activity and nutrition habits. This transition can set the stage for future lifestyle behaviors surrounding these practices (21). Their behavioral choices may influence the development of overweight and obesity and other factors that increase risk of disease later in life. Chronic disease is not yet evident in our young college population, and they appear healthier than the U.S. norm for the same age group. However, it is striking that many students are not in the optimal range for BMI, even more students do not meet the percent body fat recommendations, and almost one-third already display suboptimal or at-risk levels of at least one metabolic risk factor. These results reinforce the need to further examine the health of college students because habits that will impact their risk for the development of chronic diseases may be established during this critical transition period.

We found that the majority of students were physically fit, which is higher than the recent estimates of the U.S. college population (1,18). Of the fit students, approximately the same proportion were either within or exceeded the recommended levels for percent body fat, demonstrating how these youth can be both "fit" and "fat." This allowed us to examine whether these "fit" and "fat" students had more favorable metabolic profiles than those were "unfit." Indeed, fit individuals, even if they were "fat," had blood glucose and lipids closer to the recommended levels, although the lipid results varied by gender. Similarly, when body fat was predictive of metabolic risk factors, such as HDL and triglycerides in women, adding fitness to the model increased the strength of the association. Others have demonstrated a decreased cardiovascular disease risk score in young adolescents who are fitter as measured by a 1.6-km run time and have less body fat as assessed by skinfold thicknesses (11). This highlights the importance of trying to influence students to make lifestyle changes that could impact both their fitness and body fatness to decrease metabolic risk.

Serum glucose, HDL, and triglycerides are important metabolic risk factors, and at-risk levels comprise part of the definition for the metabolic syndrome in adolescents (8). Individually, both fitness and fatness differentially impacted these three biomarkers. Physical fitness was an important predictor of blood glucose in men and predictive of HDL and triglycerides in women. Body fatness also predicted these blood lipids in women, but fitness was a stronger predictor over percent body fat. Furthermore, if we examine the students with at-risk levels of blood glucose and blood lipids in each gender, being unfit remained a better predictor of low HDL in women. This is an important association as low levels of HDL have been linked to inflammatory pathways (15). Women therefore appear to demonstrate a slightly greater benefit of higher fitness on metabolic risk factors. It is possible, however, that we lacked sufficient power to detect these associations in men because fewer men had a percent body fat that fell outside the recommended range, and the majority was fit.

Examination of the concurrent effects of physical fitness and body fatness on metabolic risk biomarkers indicated that there was a clear benefit to being "fit and not fat," albeit again differently for men and women. Even more important is that there were benefits of fitness even when women were categorized as "fat" compared with those who were "unfit and fat." Similar results were found for blood pressure in women in a younger adolescent population (10). As the majority of our population was physically fit and of a healthy weight and body composition, a larger sample size of students classified into the two "unfit" groups may have enabled us to detect more subtle differences in serum glucose and lipid levels among all groups. Others have encountered this issue in cross-tabulating subjects into fit and fat groups as well, and it will be important to examine these factors in a less fit population (10,12). However, even among a relatively "healthy" distribution of students, at-risk levels of serum lipids are still evident.

This study has several limitations. First, because of the cross-sectional design, we cannot infer causality. Second, the use of a field test for determination of V˙O2max is not ideal but is valid and the most practical for a large longitudinal study such as this. Third, although we believe that a body composition measurement for percent body fat adds value over BMI measurements used in several studies to assess fatness (7,11,12), the use of bioelectrical impedance is not without measurement error. In addition, although there are normative standards for percent body fat, there are no set criteria for levels that are associated with increased disease risk. Others have chosen different cut points on the basis of their given sample (10). Further, although the sample size was large, the students participating in TLHS were self-selected and may represent a healthier sample of the college population. In addition, 56% of incoming freshman students from 1998 to 2007 were non-Hispanic white, whereas our sample was 76% non-Hispanic white. It has yet to be determined if these associations are similar for other ethnic and racial groups. Finally, several studies indicate that many students have poor dietary behaviors (16,18,24). Future studies are necessary to examine the impact that diet, in addition to body composition and fitness, has on these risk factors for chronic disease.

In summary, these results indicate that 1) metabolic risk factors are prevalent in this young college population, although they are healthier than the U.S. norms for fitness, weight status, and body fatness; 2) being either overweight or less physically fit predisposes students to greater metabolic stress; 3) being physically fit regardless of body fatness may confer additional health benefits; and 4) these effects vary by gender for each risk factor. Similar findings for the impact of fitness and fatness on metabolic risk have been reported by others (7,10-12), but none to our knowledge has examined these parameters during this critical transition period between adolescence and young adulthood. Practitioners should not only follow the specific guidelines set forth by the American Academy of Pediatrics for the prevention, assessment, and treatment of adolescent overweight and obesity (2) but also realize the importance of encouraging both a healthy body composition and physical fitness in college-aged youth as each appears to play important and differing roles in biochemical parameters associated with increased disease risk. In future studies, it may be important to examine whether it is more feasible for youth to increase their fitness levels than to make changes in their percent body fat to impact their metabolic health. Comprehensive strategies that promote the adoption of healthy lifestyle behaviors to support optimal health and the early prevention of disease during this critical period are needed.

The authors thank Tufts University and the Rosenberg Foundation for their financial support of the Tufts Longitudinal Health Study. The authors especially thank the Tufts University staff who assisted with all aspects of data collection, data entry, and data analysis over the 9-yr study duration.

The authors declare no conflict of interest.

Results of the present study do not constitute endorsement by the American College of Sports Medicine.


1. American College Health Association. National College Health Assessment Spring 2006 reference group data report (abridged). J Am Coll Health. 2007;55:195-206.
2. Barlow SE, Expert Committee. Expert committee recommendations regarding the prevention, assessment, and treatment of child and adolescent overweight and obesity: summary report. Pediatrics. 2007;120:S164-92.
3. Brage S, Wedderkopp N, Ekelund U, et al. Features of the metabolic syndrome are associated with objectively measured physical activity and fitness in Danish children: the European Youth Heart Study (EYHS). Diabetes Care. 2004;27:2141-8.
4. Centers for Disease Control and Prevention. CDC table for calculated body mass index values for selected heights and weights for ages 2 to 20. [cited Dec. 3, 2007]. Available from:
5. Chumlea W, Guo S, Kuszmarski R, et al. Body composition estimates from NHANES III bioelectrical impedance data. Int J Obes. 2002;26:1596-609.
6. Cureton K, Warren G. Criterion-referenced standards for youth health-related fitness tests: a tutorial. Res Q Exerc Sport. 1990;61:7-19.
7. DuBose K, Eisenmann J, Donnelly J. Aerobic fitness attenuates the metabolic syndrome score in normal-weight, at-risk-for-overweight, and overweight children. Pediatrics. 2007;120:e1262-8.
8. Duncan GE, Li SM, Zhou XH. Prevalence and trends of a metabolic syndrome phenotype among U.S. adolescents, 1999-2000. Diabetes Care. 2004;27:2438-43.
9. Economos C, Hildebrandt M, Hyatt R. College freshman stress and weight change: differences by gender. Am J Health Behav. 2008;32:16-25.
10. Eisenmann J, Welk G, Ihmels M, Dollman J. Fatness, fitness, and cardiovascular disease risk factors in children and adolescents. Med Sci Sports Exerc. 2007;39(8):1251-6.
11. Eisenmann JC, Katzmarzyk PT, Perusse L, Tremblay A, Despres JP, Bouchard C. Aerobic fitness, body mass index, and CVD risk factors among adolescents: the Quebec family study. Int J Obes. 2005;29:1077-83.
12. Eisenmann JC, Welk GJ, Wickel EE, Blair SN. Combined influence of cardiorespiratory fitness and body mass index on cardiovascular disease risk factors among 8-18 year old youth: the Aerobics Center Longitudinal Study. Int J Pediatr Obes. 2007;2:66-72.
13. Ferguson MA, Gutin B, Le NA, et al. Effects of exercise training and its cessation on components of the insulin resistance syndrome in obese children. Int J Obes. 1999;23:889-95.
14. Friedewald W, Levy R, Fredrickson D. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18:499-502.
15. Hamer M. The relative influence of fitness and fatness on inflammatory factors. Prev Med. 2007;44:3-11.
16. Huang T, Harris K, Lee R, Nazir N, Born W, Kaur H. Assessing overweight, obesity, diet and physical activity in college students. J Am Coll Health. 2003;52:83-6.
17. Isasi CR, Deckelbaum RJ, Tracy RP, Starc TJ, Berglund L, Shea S. Physical fitness and C-reactive protein level in children and young adults: the Columbia University BioMarkers Study. Pediatrics. 2003;111:332-8.
18. Keating X, Guan J, Pinero J, Bridges D. A meta-analysis of college students' physical activity behaviors. J Am Coll Health. 2005;54:116-25.
19. Lohman T. Advances in body composition assessment. In: Current Issues in Exercise Science. Champaign (IL): Human Kinetics; 1992; Monograph 3.
20. McArdle WD, Katch F, Pechar G. Reliability and interrelationships between maximal oxygen uptake, physical work capacity and step test scores in college women. Med Sci Sports Exerc. 1972;4(4):182-6.
21. Nelson M, Story M, Larson N, Neumark-Sztainer D, Lytle L. Emerging adulthood and college-aged youth: an overlooked age for weight-related behavior change. Obesity. 2008;16:2205-11.
22. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the United States, 1999-2004. JAMA. 2006;295:1549-55.
23. Pate R, Wang C-Y, Dowda M, Farrell S, O'Neill J. Cardiorespiratory fitness levels among US youth 12-19 years of age. Arch Pediatr Adolesc Med. 2006;160:1005-12.
24. Racette S, Deusinger S, Strube M, Highstein G, Deusinger R. Weight changes, exercise and dietary patterns during freshman and sophomore years of college. J Am Coll Health. 2005;53:245-51.
25. The Cooper Institute. The Cooper Institute Personal Trainer Course. [cited Dec. 16, 2007]. Available from:
26. Watts K, Jones T, Davis E, Green D. Exercise training in obese children and adolescents: current concepts. Sports Med. 2005;35:375-92.
27. Wetter AC, Economos CD. Relationship between quantitative ultrasound, anthropometry and sports participation in college aged adults. Osteoporosis Int. 2004;15:799-806.


©2010The American College of Sports Medicine