Thereare several components to physical fitness, each of which can be measured with a variety of indicators (6). Each fitness indicator in turn is influenced to varying degrees by genes, the environment, and their interactions. Given that the genetic contribution to physical fitness varies from trait to trait and from population to population, the study of the inheritance of physical fitness is a complex and ambitious endeavor (5).
There is considerable evidence that indicators of physical fitness aggregate within families. Family studies have demonstrated significant familial resemblance for indicators of aerobic fitness (2,4,17,18,23–25), muscular strength and endurance (20,21,23–25,29), and flexibility (7,15,23). Estimates of generalized heritability vary among studies, and the respective roles of genes and the shared family environment and lifestyle on physical fitness have yet to be determined.
Heritability (h2) is a population parameter that provides the proportion of the observed variance in a trait that can be attributed to genetic factors. Sometimes genetic effects and familial environmental effects are combined into generalized heritability. However, population statistics such as h2 are difficult to interpret at the level of an individual. A related concept to heritability is that of familial risk. Familial risk estimates provide an indication of the risk to an individual of having a given characteristic (obesity, diabetes, being unfit) based on their relationship to a proband, or affected individual. For discrete traits, familial risk can be estimated by the lambda coefficient, λR=P(A‖R)]/[P(A)], where P(A) is the general population prevalence of a trait, and P(A‖R) is the prevalence among relatives of degree R of a proband who is affected (27). In addition, generalized relative risk ratios have been developed for quantitative traits (11). These are defined as λR(h,l)=[PR(l‖h)/P(l)], where P(l) is the probability that a randomly selected person in the general population has a trait value in the lth segment of the trait distribution, and PR(l‖h) is the probability that a person has a trait value in the lth segment given that a relative of type R has a trait value in the hth segment.
The concept of familial risk has been applied to the study of a variety of quantitative traits, including obesity in several populations (1,13,16,31); however, the familial risk for phenotypes related to physical fitness has yet to be determined. Knowledge of λR values for a given trait is not only useful for determining the risk profile at an individual level, but also for providing justification for complex and expensive molecular genetic studies. For example, Risch (27) has suggested that for traits with risks for first-degree relatives greater than five-fold, there exists the possibility of successfully testing for major gene effects. It may also be possible to maximize the efficiency of linkage studies by recruiting pairs of relatives with high λR values in addition to recruiting participants with extreme values of the trait (11,12,16,27) for such studies. Thus, the purpose of this study was to estimate the familial risks associated with being physically fit or unfit in the general Canadian population. To this end, data from the Canada Fitness Survey of 1981, a national representative sample of the Canadian population, were analyzed.
The 1981 Canada Fitness Survey, was based on a representative sample of the Canadian population, with participants from urban and rural areas of every province. A target sample of 13,440 households was chosen by Statistics Canada, and 11,884 (88%) households agreed to participate (10). In total, 23,400 people participated in one form or another, while the physical fitness measurements described here were available for 12,905 participants. Informed consent was obtained from all participants. Of the total sample, the family relationships were available for 11,680 participants from 4144 families which contained at least two individuals. In the Canada Fitness Survey, a reference individual was identified as the person with whom contact was first made by the survey team, and each individual in the family was identified by their relationship to the reference person. The present sample consists of 4144 referents, 2580 spouses, and 4956 first-degree relatives (fathers, mothers, sons, daughters, and siblings) of the referents.
Submaximal working capacity, muscular strength and endurance, and trunk flexibility were assessed using a battery of physical fitness tests following the standardized procedures of the Canada Fitness Survey (9).
The Canadian Aerobic Fitness Test (CAFT) was used to estimate submaximal physical working capacity. The CAFT is a progressive step test in which participants were asked to perform three 3-min bouts of exercise on double 20-cm-high steps. The cadence was set by a cassette tape with a progressive tempo. Heart rates were recorded after each exercise bout, and if a predetermined ceiling heart rate was not exceeded, the participant continued to the next level, for a maximum of three bouts. The beginning level was determined by the sex and age of the participant. The power output at each exercise level was determined using the following formula:MATH 1 where, body mass is the participant’s body mass (kg), AH is the ascent height (40 cm), 9.8 is the gravitational constant, and N is the number of ascents per minute. For participants who completed at least two levels of the test, the power output at each level was plotted against heart rate. The power output corresponding to a heart rate of 150 beats·min−1 (PWC150) was then interpolated from the least squares regression line.
Maximal grip strength was measured using a Stoelting adjustable dynamometer. Participants held the dynamometer in line with the forearm at the level of the thigh and were instructed to squeeze vigorously as to exert maximal force. Two attempts were made with each hand, with the best score being recorded. The scores for the right and left hands were summed to provide a single index of strength. Muscular endurance was assessed by the maximum number of sit-ups performed in 60 s (n·min−1). Sit-ups were performed from a supine position with legs flexed 90° at the knees.
Trunk flexibility was assessed using a sit-and-reach test. The test measured how far a participant could reach towards the toes, with the knees flat on the floor. The test was repeated twice and the maximum value was recorded to the nearest 0.5 cm. A trunk flexibility score of 25 cm is equivalent to touching the floor.
The sample was divided into two-year sex-specific age groups from 7–19 yr (with the exception of 17- to 19-yr-olds), and into decades from 20 to 69 yr to preserve sample sizes (see Table 1). Given the strong relationships between PWC150, grip strength and body mass in this sample, these variables were adjusted for body mass using regression procedures. Briefly, PWC150 and grip strength were regressed on body mass within the sex- and age-groups described above. The residuals of the regressions were retained for further analyses. Body mass explained between 10% and 90% of the variance in PWC150 and between 8% and 60% of the variance in grip strength by sex and age group.
Several percentile cut-offs were determined for each variable to differentiate those in the upper and lower portions of the population distribution. For youth 7–19 yr of age, the following sex- and age-group specific cut-offs were used: 5th, 15th, 25th, 75th, 85th, and 95th. For adults, the sex-specific cut-offs were determined for 20- to 29-yr-olds and were applied as the cut-offs across all adult age groups.
For the purpose of the present study, probands were defined as reference individuals who were either above or below each percentile cut-off, respectively, for each indicator of fitness. The prevalences of being physically fit or unfit based on the above percentile cut-offs were calculated for groups of spouses and first-degree relatives of probands. The prevalence rate for each group of spouses and first-degree relatives was standardized for age and sex by weighting each age and sex group prevalence by the number of spouses or first-degree relatives in the group relative to the total sample and adding (22 groups in total). Standardized risk ratios (SRR) for groups of spouses and first-degree relatives of fit and unfit probands were calculated by dividing the standardized prevalence rate for each group by the standardized population prevalence rates, calculated using the same weightings determined for each group of spouses and first-degree relatives, respectively, to ensure a similar age and sex profile in both samples. Thus, the SRR is an estimate of the risk that a spouse or first-degree relative of a proband who is either above or below a specified cut-off will also be above or below the same cut-off. Estimates of 95% confidence intervals were calculated using the following formula (28):MATH 2
The age- and sex-standardized prevalence rates for spouses and first-degree relatives of probands, and age- and sex-matched (weighted) samples of the population are presented in Table 2, along with the associated SRR. Figures 1–4 present the SRR for high and low fitness for spouses and first-degree relatives of probands at each percentile cut-off. The SRR for spouses and first-degree relatives of probands exceeding the 95th percentile are 1.63 and 1.81 for PWC150, 2.38 and 3.16 for grip strength, 2.63 and 3.98 for sit-ups, and 2.59 and 3.56 for trunk flexibility, respectively. The SRR for spouses and first-degree relatives of probands below the 5th percentile are 1.54 and 1.34 for PWC150, 1.83 and 1.85 for grip strength, 1.13 and 1.53 for sit-ups, and 1.42 and 1.84 for trunk flexibility, respectively. The SRR decrease as the cut-off values move toward the center of the distribution, and in general, the familial risk of being physically fit (95th, 85th, or 75th) is slightly higher than the risk of being unfit (5th, 15th, or 25th) in both spouses and first-degree relatives.
With the exception of PWC150, first-degree relatives of physically fit individuals generally have higher risks at each percentile cut-off than spouses, and the difference tends to decrease in the mid-range of the distribution. The greatest differences in risk are clearly at the 95th percentile for each variable. For PWC150, the familial risks are similar for spouses and first-degree relatives across all percentile cut-offs.
A finding of significant familial risk implies that family members of affected individuals are at increased risk of also being affected, i.e., the trait “runs” in families. However, a significant familial risk does not mean that genetic factors are responsible for the observed risk. Significant familial risks could be due to genetic factors, a shared living environment and lifestyle, or a combination of the two. Higher SRR for first-degree relatives than for spouses at each percentile cut-off would suggest the contribution of genetic factors to the familial risk.
The SRR for indicators of physical fitness are significant in the general Canadian population. Spouses and first-degree relatives of fit or unfit probands have an increased risk of also being fit or unfit, respectively, compared with the general population. Higher SRR for first-degree relatives of fit individuals than for spouses for flexibility, muscular strength, and endurance suggest the possible contribution of genetic factors to the familial risk for these traits; however, the similar risks observed for PWC150 in spouses and first-degree relatives suggests that environmental factors are particularly important in explaining the familial risk for submaximal working capacity.
The findings of significant familial risk for fitness are consistent with those of Pérusse et al. (23), who used a path analytical model to estimate the transmission of physical fitness from parents to offspring in the Canadian population (Canada Fitness Survey) through both biological and cultural paths. Estimates of transmissibility (generalized heritability) were 28% for PWC150, 37% for grip strength, 37% for sit-ups, and 48% for trunk flexibility. However, the model could not differentiate between genetic and environmental (cultural) transmission. The results of the present study extends the interpretation of the transmissibility estimates by suggesting that the familial aggregation observed for PWC150 could be largely explained by the shared family environment and lifestyle characteristics, such as physical activity levels, which also aggregate within families (22,26). This is especially true because any genetic effect on PWC150 that is also common to body mass may have been removed by adjusting PWC150 for body mass. On the other hand, the familial risk for first-degree relatives for the other indicators of physical fitness may be partially explained by genetic factors, particularly at the upper end of the distribution (95th percentile); there is little indication of a substantial genetic effect for aerobic fitness in this study.
There is consistent evidence that indicates that PWC150 aggregates within families, especially after adjustment for body mass (4,22,24), but the evidence for a genetic effect per se is less convincing. For example, spousal, parent-offspring, and sibling correlations were 0.21, 0.14, and 0.25 for PWC150, respectively, in the Québec Family Study (24), whereas the corresponding correlations were 0.17, 0.17, and 0.26, respectively, in the Canada Fitness Survey (23). The higher correlations between siblings than between parents and offspring demonstrated in these studies suggest the role of shared lifestyle in explaining the familial aggregation for PWC150. In addition, the results of a BETA path-analysis of the Québec Family Study sample indicated that nongenetic inheritance could explain 100% of the transmissibility of PWC150 from parents to offspring (25).
The results for PWC150, a measure of submaximal exercise capacity, are in direct contrast to studies of maximal aerobic capacity. Twin and family studies have provided little support for the contribution of genetic factors to PWC150 (4,23–25), whereas there is consistent evidence for a genetic effect on maximal aerobic capacity (2,3,8,14,17,19). Variation in the genes associated with aerobic fitness may only become important during maximal effort when the underlying physiological systems are stressed to the limits of their capacity.
Both family and twin studies have suggested a significant genetic component to the expression of muscular strength and endurance. A large family study from Poland indicated significant familial resemblance in grip strength; however, relatively high spousal correlations (right grip r = 0.15, left grip r = 0.26) suggests either assortative mating for strength or the role of the shared family environment in explaining the family resemblance (29). Results of a BETA path analysis in the Québec Family Study indicated that 63% and 55% of the variance in muscular strength (quadriceps) and muscular endurance (sit-ups), respectively, was explained by transmission from parents to offspring. Between 20% and 30% of the transmissible variance was accounted for by genetic factors, while the remaining 70% to 80% of the variance was either not transmitted or was culturally inherited (25).
The results of two twin studies are in agreement with the family studies in implicating the role of genes in the expression of muscular strength. One study reported that 39 pairs of DZ twins had significantly greater within-pair variance than 55 pairs of MZ twins (F = 4.28, P < 0.01) for muscular strength (composite of hand grip, knee stretch, and arm bent) expressed per unit2 of height, indicating a significant genetic effect (8). Additionally, a multivariate genetic analysis of 25 MZ and 16 DZ twin pairs demonstrated that between 50% and 78% of the variance in isometric strength at various elbow joint angles was due to additive genetic effects (30).
The results of the present study indicate that first-degree relatives of fit individuals have greater familial risks for muscular strength and endurance than spouses, although spousal risks are still significant. As with other indicators of fitness, the differences in risk are greatest at the 95th percentile, which suggests that genetic factors are influencing the expression of strength at the upper end of the distribution more so than in the mid-range. Thus, it appears as though the family or shared environment is important in defining familial risk to a certain extent. However, genetic factors become increasingly important in the expression of muscular strength at the upper end of the distribution.
Estimates of the heritability of trunk flexibility from various studies range from 18% to 70% (5). The estimated transmissibility of trunk flexibility from parents to offspring was 66% in a sample of Mennonites (7), and 48% in the Canada Fitness Survey (23), suggesting significant familial aggregation. A comparison of familial risks among spouses and first-degree relatives in the Canada Fitness Survey suggests that there is a genetic component to the familial aggregation of trunk flexibility, and that genetic factors may have more influence at the upper end of the distribution than in the mid-range. It should be noted that with flexibility being joint specific, the present findings may not be applicable to all flexibility phenotypes.
The results of the present study suggest that the genetic basis may be different for extreme fitness levels as compared to moderate levels of fitness. This has serious implications for the design of genetic studies aimed at determining the molecular basis of physical fitness. Families or individuals ascertained through high levels of fitness may provide more power in genetic studies than families or individuals ascertained through moderate levels of fitness.
The results of the present study are unique in that they indicate that the risk of being physically fit or unfit in the Canadian population is in part familial. Additionally, the pattern of SRR suggests that genetic factors are important in explaining the familial risks associated with trunk flexibility and muscular strength and endurance, but much less so for submaximal work capacity (adjusted for body mass). On the other hand, spousal risks are significant for all indicators of fitness, which could be due to assortative mating for fitness phenotypes, or from shared environmental influences associated with activity patterns and other lifestyle components during cohabitation.
Special thanks to Cora Craig and her colleagues at the Canadian Fitness and Lifestyle Research Institute (CFLRI) for making available the 1981 Canada Fitness Survey database. Research support was provided, in part, by a grant from the Medical Research Council of Canada (MT-13960) and NIH grant GM28719. C. Bouchard is funded, in part, by the George A. Bray Chair in Nutrition.
1. Allison, D. B., M. S. Faith, and J. S. Nathan. Risch’s lambda values
for human obesity. Int. J. Obes. 20:990–999, 1996.
2. Bouchard, C., E. W. Daw, T. Rice, et al. Familial resemblance for O2max
in the sedentary state: the HERITAGE Family Study
. Med. Sci. Sports Exerc. 30:252–258, 1998.
3. Bouchard, C., R. Lesage, G. Lortie, et al. Aerobic performance in brothers, dizygotic and monozygotic twins. Med. Sci. Sports Exerc. 18:639–646, 1986.
4. Bouchard, C., G. Lortie, J.-A. Simoneau, C. Leblanc, G. Thériault, and A. Tremblay. Submaximal power output in adopted and biological siblings. Ann. Hum. Biol. 11:303–309, 1984.
5. Bouchard, C., R. M. Malina, and L. Pérusse. Genetics
of Fitness and Physical Performance. Champaign, IL: Human Kinetics, 1997, pp. 400.
6. Bouchard, C., and R. J. Shephard. Physical activity, fitness and health: the model and key concepts. In: Physical Activity, Fitness and Health, C. Bouchard, R. J. Shephard, and T. Stephens (Eds.). Champaign, IL: Human Kinestics, 1994, pp. 77–88.
7. Devor, E. J., and M. H. Crawford. Family resemblance for neuromuscular performance in a Kansas Mennonite community. Am. J. Phys. Anthropol. 64:289–296, 1984.
8. Engström, L.-M., and S. Fischbein. Physical capacity in twins. Acta Genet. Med. Gemellol. 26:159–165, 1977.
9. Fitness Canada. Standardized Test of Fitness: Operations Manual, 2nd Ed. Ottawa, ON: Ministry of Fitness and Amateur Sport, 1981.
10. Fitness, Canada. A User’s Guide to CFS Findings: A Technical Reference Work Describing the CFS Sample, Data Items, and Forms of Data Access. Ottawa, ON: Ministry of Fitness and Amateur Sport, 1983.
11. Gu, C., and D. C. Rao. A linkage strategy for detection of human quantitative-trait loci. I. Generalized relative risk ratios and power of sib pairs with extreme trait values. Am. J. Hum. Genet. 61:200–210, 1997.
12. Gu, C., A. Todorov, and D. C. Rao. Combining extremely concordant sibpairs with extremely discordant sibpairs provides a cost effective way to linkage analysis of quantitative trait loci. Genet. Epidemiol. 13:513–533, 1996.
13. Katzmarzyk, P. T., L. Pérusse, D. C. Rao, and C. Bouchard. Familial risk of obesity and central adipose tissue distribution in the general Canadian population. Am. J. Epidemiol. 149:933–942, 1999.
14. Klissouras, V., F. Pirnay, and J. M. Petit. Adaptation to maximal effort: genetics
and age. J. Appl. Physiol. 35:288–293, 1973.
15. Kovar, R. Sledovani podobnosti mezi rodici a jejich potomky v nekterych motorickych projech. Teorie a Praxe Telesne Vychovy 29:93–98, 1981.
16. Lee, J. H., D. R. Reed, and R. A. Price. Familial risk ratios for extreme obesity: implications for mapping human obesity genes. Int. J. Obes. 21:935–940, 1997.
17. Lesage, R., J.-A. Simoneau, J. Jobin, J. Leblanc, and C. Bouchard. Familial resemblance in maximal heart rate, blood lactate and aerobic power. Hum. Hered. 35:182–189, 1985.
18. Lortie, G., C. Bouchard, C. Leblanc, et al. Familial similarity in aerobic power. Hum. Biol. 54:801–812, 1982.
19. Maes, H., G. Beunen, R. Vlietinck, et al. Heritability of health- and performance-related fitness. In: Kinanthropometry IV, W. Duquet and J. A. P. Day (Eds.). London: Spon, 1993, pp. 140–149.
20. Malina, R. M., and W. H. Mueller. Genetic and environmental influences on the strength
and motor performance of Philadelphia school children. Hum. Biol. 53:163–179, 1981.
21. Montoye, H. J., H. L. Metzner, and J. K. Keller. Familial aggregation of strength
and heart rate response to exercise
. Hum. Biol. 47:17–36, 1975.
22. Pérusse, L., C. Leblanc, and C. Bouchard. Familial resemblance in lifestyle components: results from the Canada Fitness Survey
. Can. J. Public Health 79:201–205, 1988.
23. Pérusse, L., C. Leblanc, and C. Bouchard. Inter-generation transmission of physical fitness in the Canadian population. Can. J. Sport Sci. 13:8–14, 1988.
24. Pérusse, L., C. Leblanc, A. Tremblay, et al. Familial aggregation in physical fitness, coronary heart disease risk factors, and pulmonary function measurements. Prev. Med. 16:607–615, 1987.
25. Pérusse, L., G. Lortie, C. Leblanc, A. Tremblay, G. Thériault, and C. Bouchard. Genetic and environmental sources of variation in physical fitness. Ann. Hum. Biol. 14:425–434, 1987.
26. Pérusse, L., A. Tremblay, C. Leblanc, and C. Bouchard. Genetic and environmental influences on level of habitual physical activity and exercise
participation. Am. J. Epidemiol. 129:1012–1022, 1989.
27. Risch, N. Linkage strategies for genetically complex traits. I. Multilocus models. Am. J. Hum. Genet. 46:222–228, 1990.
28. Rothman, K. J. Modern Epidemiology.
Boston: Little, Brown and Company, 1986, pp. 358.
29. Szopa, J. Familial studies on genetic determination of some manifestations of muscular strength
in man. Genetica Polonica 23:65–79, 1982.
30. Thomis, M. A., M. Van Leemputte, H. H. Maes, et al. Multivariate genetic analysis of maximal isometric muscle force at different elbow angles. J. Appl. Physiol. 82:959–967, 1997.
31. Ziegler, A., H. Schäfer, and J. Hebebrand. Risch’s lambda values
for human obesity estimated from segregation analysis. Int. J. Obes. 21:952–953, 1997.
Keywords:©2000The American College of Sports Medicine
FAMILY STUDY; EXERCISE; STRENGTH; AEROBIC FITNESS; GENETICS; LAMBDA VALUES; CANADA FITNESS SURVEY