Journal Logo

BASIC SCIENCES: Epidemiology

Musculoskeletal Fitness and Weight Gain in Canada

Mason, Caitlin1; Brien, Susan E.1; Craig, Cora L.2; Gauvin, Lise3; Katzmarzyk, Peter T.1,4

Author Information
Medicine & Science in Sports & Exercise: January 2007 - Volume 39 - Issue 1 - p 38-43
doi: 10.1249/
  • Free


A positive relationship between musculoskeletal fitness (MSF) and health status has been demonstrated in adults (18,23) and is consistent with evidence that increased muscular strength, muscular endurance, and flexibility provide protective effects against disability (2,24) and premature mortality (8,16).

Greater MSF has also been associated with better metabolic functioning, including improvements in insulin sensitivity (12), lower incidence of metabolic syndrome (13), and improved metabolic profiles among type II diabetics (4,10), adding to a growing body of evidence suggesting that MSF may play a unique role in the management of obesit0y-related comorbidities. More recently, Fogelholm et al. (9) showed reduced functional muscle fitness despite an overall larger muscle mass and apparently stronger muscles in young, obese men. However, it is unknown whether deteriorated MSF is a consequence of obesity or whether low MSF imparts an increased risk of weight gain. Whether MSF has a distinct role in preventing obesity owing to higher basal metabolic rate attributable to increased muscle mass, greater participation in vigorous physical activity, higher total volume of physical activity, or through other pathways has apparently not yet been examined.

Given the increasing prevalence of obesity in Canada (14,29) and around the world, the primary prevention of obesity has assumed considerable importance for public health policy and practice. The purpose of this prospective study was, therefore, to examine the relationship between MSF and subsequent weight gain and the development of obesity in a representative sample of Canadians, to better understand the role of physical fitness in preventing unhealthy weight gain.



The Physical Activity Longitudinal Study (PALS) is a 20-yr follow-up cohort study of people who originally participated in the nationally representative 1981 Canada Fitness Survey (CFS) and the 1988 Campbell's Survey of Well-Being in Canada (5). The PALS cohort consists of approximately 4900 individuals aged 15 yr and older (family members born to women in the CFS were added to the cohort between 1988 and 2002) who provided data on a variety of health-related topics via self-administered questionnaire between September 2002 and April 2004. The sample used in the present analysis was limited to 606 PALS participants who were between the ages of 20 and 69 yr in 1981 and who completed both the questionnaire and physical measures components of the CFS. PALS was approved by the faculty of medicine's ethics review board of the University of Montreal, and informed written consent was obtained directly from participants in 1981 and 2002-2004.

Baseline measurements.

Anthropometric and physical fitness measurements were collected during household visits in 1981 using standardized equipment and procedures. The body mass index (BMI; kg·m−2) was calculated from measured heights and weights. Height was measured with a Harpenden tape to the nearest 0.1 cm, and body mass was measured, in light clothing, to the nearest 0.1 kg using a platform scale (Accuweigh Corporation, Columbia, MD).

The assessment of MSF included measures of grip strength, push-ups, sit-ups, and trunk flexibility. All measurements were made according to the standardized measurements of the Canada Fitness Survey, according to the protocols of the Canadian Standardized Test of Fitness (7). Grip strength was measured with a Stoelting adjustable dynamometer. Participants held the dynamometer at the level of the thigh in line with the forearm and were instructed to squeeze hard to exert maximum force. The maximum grip strength of three trials for each hand were summed to provide a single index of strength (kg). Muscular endurance was measured as the maximum number of sit-ups performed in 1 min and the maximum number of push-ups produced without time limit. Sit-ups were performed in the supine position, with knees flexed 90°. One complete sit-up entailed curling the trunk from the supine position to touching the elbows to the knees and returning to the start position. One complete push-up entailed keeping the back straight while straightening the elbows and returning to touch the nose to the floor. Men balanced from the toes, whereas women balanced from the knees. Trunk flexibility was assessed with a sit-and-reach test that required the participant to reach toward the toes with knees flat on the floor. A trunk flexibility of 25 cm is equivalent to touching the floor. The test was performed twice and the maximum value was recorded to the nearest 0.5 cm.


Baseline age, smoking status, alcohol consumption, household income, level of physical activity, and maximal oxygen uptake (V˙O2max) were used as covariates. Age was determined from birth and observation dates. Smoking status (nonsmoker, exsmoker, or current smoker), frequency of alcohol consumption (six-point scale ranging from complete abstention to consuming an alcoholic beverage at least once per day), and household income (self-reported family income before taxes for the year before the 1981 survey) were obtained by questionnaire.

Leisure-time physical activity levels were assessed in 1981 and in 2002-2004 using adaptations of the Minnesota Leisure-Time Physical Activity Questionnaire (28) that collected information about physical activity levels during the preceding 12 months. A list of physical activities was provided, and respondents indicated the number of occasions and the average duration of the activity bouts. Average daily leisure-time activity energy expenditure (AEE) was calculated as follows:

where Ni is the number of times the activity was performed, Di was the average duration in hours of the activity, and METsi was the estimated energy cost of the activity (kJ·kg−1·h−1).

Cardiorespiratory fitness was assessed by estimating maximal oxygen uptake (V˙O2max; METs) using a modified version of the Canadian Aerobic Fitness Test (CAFT), a submaximal step test with accepted validity for population surveys (27).


Self-reported height and weight were collected via self-administered questionnaire between 2002 and 2004 and were used to estimate BMI. Overweight and obesity were defined as BMI 25-29.9 kg·m−2 and ≥30kg·m−2, respectively, according to current Health Canada guidelines (22). Twenty-year weight change was calculated by subtracting measured weight in 1981 from self-reported weight in 2002-2004 and was subsequently dichotomized according to whether it exceeded 10 kg. A 10-kg threshold was chosen because it represents approximately one third of the cohort (34% of men and 38% of women gained ≥ 10 kg) and was in line with previous studies that have documented adverse changes in obesity-related risk factors with a weight gain of this magnitude (21).

Data analysis.

Two MSF composite scores were computed and retained for further analysis. The first was based on the Canadian Society for Exercise Physiology's guide for Physical Activity, Fitness and Lifestyle Approach (CPAFLA) (3), whereas the second was based on a principal-components analysis (PCA) of the individual MSF component scores.

Scores for each of the MSF components were categorized (excellent, very good, good, fair, needs improvement) according to age (10-yr groupings) and gender-specific cut points recommended in the CPAFLA. Because of differences in the sit-up protocol between the Canadian Standardized Test of Fitness and the CPAFLA Health-Related Fitness Test, sit-ups were alternatively scored on the basis of sex- and age group-specific quintiles derived from the original CFS sample (N = 9206). For example, individuals with scores corresponding to the highest quintile of their specific sex and age group were assigned a score of excellent, whereas those with scores corresponding to the lowest quintile of their sex and age group were scored as needing improvement. Next, overall MSF composite scores were determined on the basis of summed weighted scores of the individual components entered into the CPAFLA nomogram. For the purposes of analysis, the five groups created by the nomogram (excellent, very good, good, fair, needs improvement) were reduced to three by combining participants who achieved an excellent or very good overall score into a high-MSF group and those scoring fair or needing improvement into a low-MSF group. Individuals with an overall score of good were considered the average-MSF group.

The second MSF composite score was derived from a PCA of the MSF component scores. Briefly, the scores for sit-ups, push ups, flexibility, and grip strength were subjected to PCA, and the first principal component was retained for further analysis. The first principal component explained 44% of the variance in the original variables in women and 42% of the variance in men. The correlations between the original variables and the first principal component can be interpreted as loadings that represent their contribution to the overall score. The factor loadings were 0.12, 0.79, 0.82, and 0.61 in men and 0.50, 0.77, 0.72, and 0.62 in women for grip strength, sit-ups, push ups, and trunk flexibility, respectively. Thus, each original variable contributed similarly to the composite MSF score in both sexes, with the exception of grip strength, which did not contribute significantly to the composite MSF score in men. The PCA composite scores were further divided into tertiles to facilitate comparison with the CPAFLA-derived composite scores.

General linear regression models were used to predict weight gain and BMI at follow-up on the basis of MSF composite scores. Next, logistic regression was used to predict the likelihood of obesity in 2002-2004 and of gaining at least 10 kg during follow-up associated with each individual component of MSF (grip strength, push-ups, sit-ups, trunk flexibility) and with both overall composite scores (CPAFLA and PCA) as measured in 1981. The group with high MSF served as the referent group in analyses involving the composite scores. All analyses were performed in the entire sample and in men and women separately. The regression models were tested under two adjustment strategies. First, with adjustment for sex (for analysis of overall sample), age, and baseline BMI; secondly, with adjustment for all covariates including age, BMI, level of physical activity, V˙O2max, smoking status, alcohol consumption, and income. All analyses were conducted using SAS systems and procedures (26).


Descriptive characteristics of the sample at baseline are provided in Table 1. On the basis of the CPAFLA health-related fitness test scores, approximately 26% of the sample had low MSF, 38% had average MSF, and 36% had high MSF.

Descriptive characteristics of men(N = 290) and women (N = 316) (20-69 yr)in the Physical Activity Longitudinal Study at baseline (1981).

During the 20-yr follow-up, the average BMI (SD) of the sample increased from 23.6 (3.2) kg·m−2 in 1981 (men: 24.6 (2.9) kg·m−2; women: 22.7 (3.1) kg·m−2) to 26.1 (4.0) kg·m−2 in 2002-04 (men: 26.5 (3.4) kg·m−2; women: 25.7 (4.6) kg·m−2), reflecting a mean weight gain of 7.4 (8.7) kg (men: 6.7 (8.4) kg; women: 8.1 (8.9) kg). Consequently, the prevalence of obesity (BMI ≥ 30 kg·m−2) also increased in the entire sample from 3.1% in 1981 to 15.2% in 2002-2004. Thirty-four percent of men and 38% of women gained at least 10 kg during follow-up. Thirty-nine percent of normal-weight individuals became overweight, whereas 7% of normal-weight and 27% of overweight individuals became obese by 2002-2004.

After adjustment for all covariates, higher BMI in 1981 was a significant predictor of obesity 20 yr later (i.e., in 2002-2004) in the overall group (OR: 1.65, 95% CI: 1.47-1.86) and separately in men (OR: 1.89, 95% CI: 1.54-2.34) and women (OR: 1.54, 95% CI: 1.33-1.78). However, baseline BMI was not associated with increased odds of gaining ≥ 10 kg during follow-up.

All individual components of MSF were significantly correlated with each other (P < 0.05) in the overall sample (Table 2). The strength of the correlations ranged from −0.09 between grip strength and trunk flexion to −0.49 between sit-ups and push-ups. All the correlations were positive, with the exception of grip strength and trunk flexibility. V˙O2max was moderately and positively associated with grip strength, push-ups, and sit-ups (r = 0.43-0.61, all P < 0.001), but not trunk flexibility (r = 0.06, P = 0.15). Baseline BMI (kg·m−2) was positively correlated with grip strength and negatively correlated with all other musculoskeletal fitness items (all P < 0.05).

Pearson correlation coefficients between MSF component items and baseline BMI, physical activity, and V˙O2max.

Results of the linear regression analyses revealed a significant inverse association between MSF and both BMI in 2002-2004 and weight gain during follow-up, after adjusting for age, sex, baseline BMI, and physical activity level (Table 3). Moreover, there was also a significant linear relationship between MSF and physical activity level at follow-up that was independent of age, sex, BMI, and baseline physical activity level.

β coefficients for linear regression models predicting 20-yr changes in BMI (kg m−2) and body weight (kg) associated with baseline (1981) musculoskeletal fitness scores among PALS participants 20-69 yr.

Results for the CPAFLA-derived composite MSF scores indicated that low MSF was not associated with significantly higher odds of being obese in 2002-2004 (P for trend = 0.09; Table 4); however, it was associated with higher odds of having gained at least 10 kg during follow-up (OR: 1.78, 95% CI: 1.14-2.79; P for trend = 0.01) independent of age, sex, and baseline BMI (Table 5). Furthermore, after additional adjustment for smoking, alcohol consumption, income, total physical activity, and cardiorespiratory fitness (V˙O2max), low MSF remained significantly predictive of weight gain of ≥ 10 kg, with odds of 1.72 (1.12-2.66) and 1.76 (1.09-2.85) among those with average and low MSF, respectively.

Odds of obesity associated with baseline (1981) musculoskeletal fitness scores among PALS participants 20-69 yr.
Twenty-year odds of weight gain of ≥ 10 kg associated with baseline (1981) musculoskeletal fitness scores among PALS participants 20-69 yr.

Although using the first component of the PCA as a continuous independent variable showed MSF to be inversely associated with both obesity and weight gain of ≥ 10 kg, no significant effect of low MSF was detected when this variable was analyzed by tertiles (Tables 4 and 5).

When the component measures of MSF were examined individually using logistic regression, only push-ups were significantly predictive of obesity at follow-up in the overall sample (Table 4). These elevated odds were attenuated in the sex-specific analyses; however, low grip strength became predictive of obesity when women were examined alone. Both push-ups and sit-ups were significant predictors of weight gain of ≥ 10 kg in the overall sample initially; however, the elevated odds associated with sit-ups were attenuated to nonsignificance after full multivariate adjustment (Table 5).


To our knowledge, this is the first study to prospectively examine the relationship between MSF and weight gain in a population cohort. Previous reports have provided evidence that physical activity and cardiorespiratory fitness are associated with lower risk of future weight gain (6,19,25). Our findings suggest that low MSF, at least as defined according to the CPAFLA, is also protective against substantial (≥ 10 kg) weight gain, independent of cardiorespiratory fitness. Although the odds ratios for obesity (BMI ≥ 30 kg·m−2) at follow-up did not reach statistical significance, they tended to be higher among individuals with low MSF. Future studies should examine these relationships in other cohorts.

During the 20-yr follow-up of this cohort, an increase in the average BMI of the sample and in the prevalence of obesity was observed, consistent with documented temporal trends among Canadians (14). However, in the present sample, MSF was inversely related to weight gain and to BMI at follow-up, suggesting that age-related weight gain may be more pronounced among individuals with low levels of muscular functioning. In fact, individuals with low MSF, as categorized by the CPAFLA, had 78% increased odds of substantial weight gain (≥ 10 kg) during follow-up compared with those with high MSF, after adjusting for sex, baseline age, BMI, total physical activity energy expenditure, cardiorespiratory fitness, smoking, alcohol consumption, and income. Surprisingly, these elevated odds were of similar magnitude to those observed among individuals with average MSF. Currently, the CPAFLA suggests that the greatest MSF-related health gains are incurred by moving out of the "needs improvement" category (3). Because of the small number of PALS participants in this category, we could not conduct valid analysis to examine this group independently. The results of our grouped analysis, however, suggest that in counseling using the CPAFLA, the benefits of MSF with respect to preventing weight gain may be limited to those with at least very good MSF levels.

The exact mechanism(s) through which musculoskeletal functioning is associated with weight gain has yet to be elucidated. One possible explanation for the observed relationship between MSF and weight gain may be through increased levels of physical activity, because high levels of muscular fitness could represent an active lifestyle or participation in more vigorous types of physical activity. However, the uniqueness of the PALS dataset with respect to its measures of cardiorespiratory fitness and average physical activity energy expenditure allowed us to minimize potential confounding attributable to varied levels of physical activity and cardiorespiratory fitness in this sample. Unfortunately, in the absence of interim measures in the follow-up period, the extent to which musculoskeletal and cardiorespiratory fitness are related as a result of participation in physical activity is unknown. Because maintaining adequate strength, endurance, and flexibility facilitate one's ability to carry out activities of daily living and to participate in physical activity, low musculoskeletal fitness may ultimately be both a cause and a consequence of weight gain.

In light of only a modest correlation between musculoskeletal and cardiorespiratory fitness (r = 0.35 between PCA-derived score and V˙O2max) in our sample, it seems plausible that higher levels of muscular functioning also act through biological pathways independent of physical activity to prevent weight gain over time. Resistance-training studies support the conclusion that the benefits of increased musculoskeletal functioning include not only increases in fat-free mass but also improved insulin sensitivity, glycemic control, and reduced abdominal adipose tissue (11,20,30). Thus, the apparent protective effect of MSF against substantial weight gain in this cohort may operate through one or more of these putative pathways.

Because the classification of obesity for this analysis was based on BMI, it is possible that some subjects may have been misclassified as a result of high levels of fat-free mass. Given that only two obese participants had high MSF at baseline and that this type of misclassification at follow-up would be expected to yield findings opposite those reported here, the impact of this potential misclassification seems small. Nevertheless, it may help explain why low MSF was a significant predictor of weight gain of ≥ 10 kg but not of obesity, and it should be considered when interpreting the results. In addition, measured heights and weights were not obtained in the 2002-2004 phase of the PALS; rather, these values were self-reported. Because self-reported heights and weights tend to be idealized (1), this self-reporting likely resulted in an underestimation of true weight gain and misclassification of some obese participants, biasing the results towards the null. A recently released report from Statistics Canada using data based on measured heights and weights from the Canadian Community Health Survey indicated that, in 2004, 23.1% of Canadians were obese (29), a figure that is considerably higher than what was reported in PALS. This underestimate, combined with small cell sizes, particularly for sex-specific and multivariate-adjusted analyses, may account for some of the borderline significant findings.

Many previous studies examining the health benefits of MSF have focused on elderly or unhealthy populations; however, this is the first attempt to examine the importance of MSF with regards to the prevention of weight gain in a large, population-based sample that includes a wide range of ages and health statuses. However, a limitation of the analysis was the lack of dietary information for inclusion as a covariate, which may have provided further resolution of the relationship between MSF and weight gain. Unfortunately, dietary data were not collected in the 1981 Canada Fitness Survey.

The results of this study support increasing evidence that resistance training and other activities that promote MSF have a unique role in promoting health maintenance and disease prevention. Beyond the typically recognized benefits to overall health and functionality, increasing MSF may also help prevent unhealthy weight gain. Given the mounting toll of obesity on public health in Canada (15,17), MSF deserves greater attention as a determinant of future weight gain and as a possible area for intervention. Future studies should investigate potential mechanisms through which musculoskeletal functioning affects metabolism and may protect against weight gain over time.

This study was supported by a grant from the Social Sciences and Humanities Research Council of Canada and a New Emerging Team Grant from the Canadian Institutes of Health Research and the Heart and Stroke Foundation of Canada.


1. Booth, M. L., C. Hunter, C. J. Gore, A. Bauman, and N. Owen. The relationship between body mass index and waist circumference: implications for estimates of the population prevalence of overweight. Int. J. Obes. Relat. Metab. Disord. 24:1058-1061, 2000.
2. Brill, P. A., C. A. Macera, D. R. Davis, S. N. Blair, and N. Gordon. Muscular strength and physical function. Med. Sci. Sports Exerc. 32:412-416, 2000.
3. Canadian Society for Exercise Physiology. The Canadian Physical Activity, Fitness and Lifestyle Approach. 3rd ed. Ottawa, Canada: Canadian Society for Exercise Physiology, pp. 36-51, 2003.
4. Cauza, E., U. Hanusch-Enserer, B. Strasser, et al. The relative benefits of endurance and strength training on the metabolic factors and muscle function of people with type 2 diabetes mellitus. Arch. Phys. Med. Rehabil. 86:1527-1533, 2005.
5. Craig, C. L., L. Gauvin, S. Cragg, et al. Towards a social epidemiological perspective on physical activity and health: the aims, design, and methods of the Physical Activity Longitudinal Study (PALS). J. Phys. Act. Health 3:272-284, 2005.
6. DiPietro, L., H. W. Kohl 3rd, C. E. Barlow, and S. N. Blair. 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. 22:55-62, 1998.
7. Fitness Canada. Canadian Standardized Test of Fitness, Ottawa, Canada: Government of Canada, pp. 41, 1981.
8. FitzGerald, S. J., C. E. Barlow, J. B. Kampert, J. R. Morrow, A. W. Jackson, and S. N. Blair. Muscular fitness and all-cause mortality: prospective observations. J. Phys. Act. Health 1:7-18, 2004.
9. Fogelholm, M., J. Malmberg, J. Suni, M. Santtila, H. Kyrolainen, and M. Mantysaari. Waist circumference and BMI are independently associated with the variation of cardio-respiratory and neuromuscular fitness in young adult men. Int. J. Obes. (Lond.) 30:962-969, 2006.
10. Honkola, A., T. Forsen, and J. Eriksson. Resistance training improves the metabolic profile in individuals with type 2 diabetes. Acta Diabetol. 34:245-248, 1997.
11. Ibanez, J., M. Izquierdo, I. Arguelles, et al. Twice-weekly progressive resistance training decreases abdominal fat and improves insulin sensitivity in older men with type 2 diabetes. Diabetes Care 28:662-667, 2005.
12. Ishii, T., T. Yamakita, T. Sato, S. Tanaka, and S. Fujii. Resistance training improves insulin sensitivity in NIDDM subjects without altering maximal oxygen uptake. Diabetes Care 21:1353-1355, 1998.
13. Jurca, R., M. J. Lamonte, C. E. Barlow, J. B. Kampert, T. S. Church, and S. N. Blair. Association of muscular strength with incidence of metabolic syndrome in men. Med. Sci. Sports Exerc. 37:1849-1855, 2005.
14. Katzmarzyk, P. T. The Canadian obesity epidemic, 1985-1998. CMAJ 166:1039-1040, 2002.
15. Katzmarzyk, P. T., and I. Janssen. The economic costs associated with physical inactivity and obesity in Canada: an update. Can. J. Appl. Physiol. 29:90-115, 2004.
16. Katzmarzyk, P. T., and C. L. Craig. Musculoskeletal fitness and risk of mortality. Med. Sci. Sports Exerc. 34:740-744, 2002.
17. Katzmarzyk, P. T., and C. Mason. Prevalence of class I, II and III obesity in Canada. CMAJ 174:156-157, 2006.
18. Kell, R. T., G. Bell, and A. Quinney. Musculoskeletal fitness, health outcomes and quality of life. Sports Med. 31:863-873, 2001.
19. Lewis, C. E., D. E. Smith, D. D. Wallace, O. D. Williams, D. E. Bild, and D. R. Jacobs Jr. Seven-year trends in body weight and associations with lifestyle and behavioral characteristics in black and white young adults: the CARDIA study. Am. J. Public Health 87:635-642, 1997.
20. Miller, W. J., W. M. Sherman, and J. L. Ivy. Effect of strength training on glucose tolerance and post-glucose insulin response. Med. Sci. Sports Exerc. 16:539-543, 1984.
21. Norman, J. E., D. Bild, C. E. Lewis, K. Liu, and D. S. West. The impact of weight change on cardiovascular disease risk factors in young black and white adults: the CARDIA study. Int. J. Obes. Relat. Metab. Disord. 27:369-376, 2003.
22. Office of Nutrition Policy and Promotion. Canadian Guidelines for Body Weight Classification in Adults. Ottawa, Canada: Health Canada, 2003. Available at:
23. Payne, N., N. Gledhill, P. T. Katzmarzyk, V. Jamnik, and S. Ferguson. Health implications of musculoskeletal fitness. Can. J. Appl. Physiol. 25:114-126, 2000.
24. Rantanen, T., J. M. Guralnik, D. Foley, et al. Midlife hand grip strength as a predictor of old age disability. JAMA 281:558-560, 1999.
25. Rissanen, A. M., M. Heliovaara, P. Knekt, A. Reunanen, and A. Aromaa. Determinants of weight gain and overweight in adult Finns. Eur. J. Clin. Nutr. 45:419-430, 1991.
26. SAS Institute. SAS/STAT Software: Changes and Enhancements through Release 8.2. Cary, NC: SAS Institute, 2001-2002.
27. Shephard, R. Fitness of a Nation: Lessons from the Canada Fitness Survey. New York: Karger, pp. 35-40, 1984.
28. Taylor, H. L., D. R. Jacobs Jr., B. Schucker, J. Knudsen, A. S. Leon, and G. Debacker. A questionnaire for the assessment of leisure time physical activities. J. Chronic. Dis. 31:741-755, 1978.
29. Tjepkema, M. Adult obesity. Health Rep. 1717(3):9-25, 2006.
30. Treuth, M. S., A. S. Ryan, R. E. Pratley, et al. Effects of strength training on total and regional body composition in older men. J. Appl. Physiol. 77:614-620, 1994.


©2007The American College of Sports Medicine