Secondary Logo

Journal Logo

Genetic Effects on Physical Activity: Results from the Swedish Twin Registry


Medicine & Science in Sports & Exercise: August 2006 - Volume 38 - Issue 8 - p 1396-1401
doi: 10.1249/01.mss.0000228941.17034.c1
BASIC SCIENCES: Epidemiology

Purpose: The aim of this study was to investigate the genetic effects on leisure-time physical activity using data from the Swedish Twin Registry.

Methods: We investigated 13,362 twin pairs (5334 monozygotic and 8028 dizygotic pairs) aged 14-46 yr. Information on leisure-time physical activity was obtained by questionnaire. Correlations and odds ratios of physical activity were calculated for males, females, and monozygotic and dizygotic twins, respectively. Structural equation modeling was used to estimate the contribution of genetic effects as well as common and nonshared environmental factors on leisure-time physical activity.

Results: About one third of the twins reported that they exercised regularly (26% in females and 39% in males). The correlations of physical activity were twice as high in monozygotic compared with dizygotic twins, suggesting the presence of genetic effects. The variation in physical activity due to heritage was 57% (95% confidence interval (CI) = 0.49-0.63) in males and 50% (95% CI = 0.49-0.55) in females. The common environmental influence on physical activity was very small compared with the influence from environmental factors unique to the individual.

Conclusions: Our study establishes heredity as an important component behind individual differences in physical activity in adult men and women. This may be one reason behind difficulties in convincing people to adopt an active lifestyle. Still, this study shows that there is a substantial influence on physical activity from environmental factors unique to the individual.

1Division of Epidemiology, Stockholm Centre of Public Health, and Department of Epidemiology, Institute of Environmental Medicine, Karolinska Institutet, Stockholm, SWEDEN; 2Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, SWEDEN; and 3Department of Surgical Sciences, Section of Orthopaedics, University Hospital, Uppsala, SWEDEN

Address for correspondence: Sofia Carlsson, Ph.D., Division of Epidemiology, Stockholm County Council, Norrbacka, S-171 76 Stockholm, Sweden; E-mail:

Submitted for publication September 2005.

Accepted for publication February 2006.

Despite growing evidence of major beneficial health effects and public health recommendations, many adults do not exercise regularly. According to recent publications, the prevalence of a sedentary lifestyle in the European Union ranges from about 40% in Sweden to almost 90% in Portugal (20). The development of health promotion strategies to increase physical activity in adults may be improved by better knowledge about the determinants of physical activity.

It has been shown that the level of physical activity that people engage in during their leisure time is influenced by social, demographic, psychological, and physiological factors (7,17,20). Evidence from a fairly small number of family and twin studies suggests that genetic factors also influence physical activity (reviewed in Frederiksen and Christensen (5)). However, the magnitude of this influence is not clear; previous studies suggest that genetic effects explain a small, 29% in the Quebec family study (15), to moderately large, 62% in the Finnish Twin Cohort (6), part of the individual variation in mean physical activity levels. Data in adolescents suggest that the genetic component to physical activity is stronger in men than in women (2,3,11). Whether this holds true in adults is not known. Previous studies were based on men (6,9) or analyzed men and women together (5,15) so there is little gender-specific information and no data on females separately. It has been suggested that the genetic component is most important for vigorous physical activity (6,9), but few studies have addressed the respective genetic effects on high and low level of physical activity.

Twins offer a natural study population for genetic epidemiology because monozygotic (MZ) twins are genetically identical, whereas dizygotic (DZ) twins, like ordinary siblings, share half of their segregating genes. Thus, by comparing concordance of physical activity between DZ and MZ twin pairs, the magnitude of genetic effects can be investigated. In contrast to family studies, it is possible to disentangle the familial pattern into hereditary or environmental factors. The Swedish Twin Registry is currently the largest in the world and provides a unique opportunity for evaluating genetic factors (10). The aim of this study was to investigate the genetic effects on physical activity using this data base.

Back to Top | Article Outline


The Swedish Twin Registry.

The Swedish Twin Registry includes three cohorts collected in slightly different ways. The analyses of this study were based on the middle cohort, which was established in 1970. This cohort includes all twins born from 1926 through 1958 who were alive and living in Sweden in 1970 (12). In 1972, all same-sexed pairs received a questionnaire including questions on lifestyle factors such as physical activity. Reponses that allowed a determination of zygosity were received from 16,120 pairs: 2,936 MZ male pairs; 3,337 MZ female pairs; 4,833 DZ male pairs; and 5,014 DZ female pairs. The response rate was 88%. The analyses of this study are based on data from the 13,362 twin pairs with complete information on physical activity. The Swedish Twin Registry has been approved by both the data inspection and the ethical committees at Karolinska Institutet.

Back to Top | Article Outline


Zygosity of twins (MZ, DZ, and unclassified) was classified based on the question: "Were you as children as alike as two peas in a pod?" When both twins answered affirmatively, they were defined as MZ, and when both answered no, they were defined as DZ. This method has been proven to correctly classify more than 95% of the twins (10).

Back to Top | Article Outline

Physical activity.

Information on physical activity was collected by a question on average physical activity during leisure time in the past year, with seven response options ranging from "almost never" to "exercise very much" (Table 2). We defined low physical activity as exercising "almost never" to "hardly ever," moderate physical activity as "very little" to "quite a lot," and high physical activity was defined as exercising "a lot" to "very much."

Back to Top | Article Outline

Other "covariates."

The questionnaire included questions on height and weight that were used to calculate body mass index as the weight in kilograms divided by the square of height in meters. Subjects were also asked whether they had ever had any long-term or serious illness.

Back to Top | Article Outline

Data analyses.

The first step of the analysis was to determine whether genetic effects were present. Because the variable of interest was ordinal, the similarity within twins with regard to physical activity was assessed by the polychoric correlation coefficients (sometimes referred to as "correlation in liability") separately for MZ and DZ twin pairs. The interpretation of these analyses is that when correlation is higher in MZ twins than in DZ twins, genetic effects are likely to explain some of the variation in physical activity levels.

To investigate the question of whether both high and low physical activity are inherited, we used multivariate logistic regression, conditional on pair membership, to estimate odds ratios (OR) with 95% confidence interval (CI) of low physical activity (categories 1-2, exercise "almost never" to "hardly ever") and high physical activity (categories 6-7, exercise "a lot" to "very much"). The remaining three middle categories ("exercise very little" to "exercise quite a lot") were used for reference purposes. The ratio of the MZ and DZ OR were interpreted as indicators of genetic effects on high and low physical activity.

Correlations and OR can test whether genetic effects are important. However, they are not informative about the magnitude of the genetic effects. Therefore, in a third step, we used data from both MZ and DZ twins simultaneously in structural equation models (13), which made it possible to quantify how much of the variance in physical activity could be explained by additive genetic effects, common environmental effects, and unique environmental effects. Common environmental effects reflect twin similarity that is not explained by heritable genetic influence but that results from factors shared by the family unit such as values, dietary habits, leisure-time activities, socioeconomic situation, and neighborhood. Unique environmental effects are those explained by environmental factors unique to the individual.

The analyses of this study were based on the usual assumptions made in twin studies: that the genetic similarity is half as great for DZ as for MZ twin pairs, whereas common environmental influences contribute equally to making MZ and DZ pairs similar, and that twins are representative of the general population (13).

Back to Top | Article Outline



There were 5334 MZ (2440 male and 2894 female) and 8028 (3804 male and 4224 female) DZ twin pairs. The age range was 14-46 yr. Only modest differences in age, body mass index, prevalence of disease, and overweight were seen between zygosity groups (Table 1) in both males and females.



Back to Top | Article Outline

Physical activity levels for male and female subjects.

MZ and DZ twins had similar distributions across the seven levels of physical activity, with about 20% (males) to 25% (females) reporting that they hardly ever exercised (Table 2). Males were more active than females; 16% (DZ) to 17% (MZ) of male twins reported high physical activity compared with 5% (DZ) to 6% (MZ) of female twins (Table 2). These results show that twins and cotwins tended to fall in the same or adjacent categories of physical activity and that there were few twin pairs with contrasted reported physical activity.



Back to Top | Article Outline

Physical activity in MZ versus DZ twin pairs.

The polychoric correlation of physical activity for MZ males was 0.62, whereas the correlation for DZ males was 0.31 (Table 3). The correlation of physical activity in females was 0.58 and 0.30 for MZ and DZ, respectively. The correlation of physical activity was thus about twice as high in MZ compared with DZ twins (Table 3). To assess the magnitude of the genetic component, structural equation analyses were performed. These analyses indicate that about half of the variation in physical activity, 57% in males and 50% in females, was attributable to genetic factors. Less than 7% of the variation in physical activity was explained by common environmental factors. Consequently, the influence of unique environmental factors on physical activity was 40% (95% CI = 37-42%) in males and 44% (95% CI = 43-46%) in females. When these analyses were stratified by age, the genetic component was 64% in males aged 14-28 yr and 40% in males aged 29-46 yr. In females, the corresponding estimates were 51 and 41%, respectively (Table 4).





Back to Top | Article Outline

High versus low physical activity.

The genetic effects on high and low physical activity were assessed by separate analyses of the OR of high and low physical activity. Male MZ twins were 14 times more likely to be very active themselves if their cotwin was very active compared with if the cotwin was moderately active. The corresponding figure in DZ twins was about five. For low physical activity, male MZ twins with a cotwin with low physical activity were five times more likely to be inactive than if the cotwin was moderately active, and this increase was about 1.8 in DZ twins. Thus, in males, the ratio of MZ/DZ OR was about 3 for both high and low physical activities, indicating genetic effects at both ends of the physical activity spectrum (Table 3). In females, twins with very active cotwins were 33.6 (MZ) and 10.7 (DZ) times more likely to be very active themselves than if the cotwin was moderately active. For low physical activity, the corresponding estimates were 5.4 (MZ) and 2.3 (DZ), respectively. When these results were stratified by age (Table 4), similar results were seen in younger (14-28 yr) as well as older subjects (29-46 yr).

Back to Top | Article Outline


Our study demonstrates that MZ twin pairs are more similar in their leisure-time physical activity than DZ twins. This indicates that genetic factors influence the level of physical activity in which a person engages. The genetic component was estimated to be 57 and 50% in males and females, respectively. These values are higher than what was reported in most previous studies (5) but consistent with results reported in the Finnish Twin cohort (6). This is the first twin study to have investigated hereditary effects on physical activity in males and females separately. According to our results, the magnitude of these effects were equally strong in both genders

A number of studies have shown that physical performance of an individual measured as aerobic performance, muscle strength, muscle endurance, and anaerobic performance is influenced by genetic factors (21). Phenotypes involved in physical functioning such as muscle mass, grip, and leg extensor strength as well as bone density and bone size also display a moderate to strong genetic component (1,14,16,18). In a recent Swedish study, it was also shown that the individual response to a certain level of physical activity is heritable (19). Relatively few studies have addressed the genetic effects on the behavioral aspect of physical activity measured as sports participation or leisure-time physical activity. Previous studies, including the Danish Twin study (5) and a study of 1587 Dutch adolescent twin pairs (8), found that hereditary factors explained 45-49% of the variation in sports participation. For overall leisure-time physical activity, genetic effects explained 38% in males in the Vietnam Era Twin Study (9) and 29% in males and females in the Canadian Family study (15). The strongest genetic effect (by magnitude of heritability coefficients) on physical activity measured as an activity scored based on amount, intensity, and duration was 62% in males reported in the Finnish Twin Cohort (6). One possible explanation for this large effect is that genetic factors are more important for physical activity in men than in women. Such gender differences have been reported in adolescents (2,3,11). However, our findings do not support a gender difference because the genetic component was similar in men and women.

Another potential explanation for the discrepancies between study results is that the genetic effect on physical activity is dependent on age. In the present study, a larger part of the variation in physical activity was explained by hereditary factors in adolescents and young adults (aged 14-28 yr) than in middle-aged subjects (aged 29-46 yr) (approximately 60 and 40%, respectively). On the other hand, Kaprio et al. (6) investigated twins from 18 to older than 60 yr of age and showed that the heritability coefficients did not change systematically with age. Correspondingly, Fredriksen and Christensen (5) indicated heritability of 50% in twins aged 45-68 yr, which is very similar to our findings in younger adults and adolescents. Hence, whether genetic effects differ across the age spectrum remains unclear. It should be noted that some difference in heritability coefficients between studies is to be expected because these studies differ in the methods used to capture physical activity and in the type of activities in which people engage. Further, there may also be genetic differences between the populations investigated.

Few studies have investigated genetic effects at different levels of physical activity. Kaprio et al. (6) suggested that genes may be more important for high physical activity. Similarly, data from the Vietnam Era Twin study indicated higher heritability estimates for participation in vigorous compared with moderate activity (8). In the present study, the results seem to indicate that both high physical activity and a sedentary lifestyle are, in part, heritable, although the MZ/DZ OR were more pronounced for high than for low physical activity for females. Thus, genetic effects are important for all types of physical activity, but it is still not clear whether the genetic effects are less important for low physical activity.

An important finding of the present study was that the influence of common environmental factors on the distribution of physical activity was small compared with the influence of hereditary and environmental factors unique to the individual. These are factors shared by the family unit such as values, dietary habits, leisure-time activities, socioeconomic situation, and neighborhood. Similar findings have been reported in previous studies of adult twins (5,9). A small influence of common environmental factors indicates that besides heritage, environmental factors unique to the individual influence physical activity in adults. In adolescents, common environmental factors seem to be very important for physical activity (2,18). Stubbe et al. (18) show that the genetic effect on physical activity starts to appear in late adolescence and that there is a shift in the determinants of sports participation from adolescence to adulthood in that common environmental factors cease to be of importance in favor of genetic factors and individual environmental factors. Our analyses in subjects aged 14-28 yr indicate that genetic factors influenced physical activity, although not as much as in 29- to 46-yr-old subjects. Common environmental factors did not play a significant role except in females aged 14-28 yr. The difference in our findings compared with previous work may be due to our small number of adolescents.

A limitation of the present study was that we investigated leisure-time physical activity measured in 1970. The physical activity pattern 36 yr ago probably differs from the pattern today. One assumption is that everyday activities such as walking and bicycling, in the absence of other modes of transportation, were probably more common in the 1970s. Today, organized sports almost certainly make up a larger proportion of overall activities. For this reason, the relationship between genetic effects and physical activity may be slightly different today. It seems possible that because participation in organized sports reflects an active choice, the genetic component may be more important for such activities than for overall physical activity. If this is true, then the genetic effects on physical activity may be even stronger today. Another limitation was that physical activity was measured with a single question on leisure-time activities. If MZ and DZ twins differ with regard to reliability of self-reported physical activity, the results may be confounded. However, we have no reason to believe that this is the case. It is possible that imprecise information on physical activity may have led us to underestimate the genetic effects.

One issue is the generalizability of our results. Evans and Martin (4) argue that twins are representative of the general population with respect to most health and behavioral outcomes, and we have no reason to believe that this is not true also for physical activity.

Back to Top | Article Outline


In summary, this study confirms that genetic effects play an important role in explaining individual differences in physical activity in adults. The genetic component seems to be present and substantial in males and females. In addition, our results suggest that both active and inactive lifestyles are, in part, inherited. The genetic effects on physical activity may be one reason for the difficulties in convincing people to adopt an active lifestyle. Still, this study shows that there is a substantial influence on physical activity from unique environmental factors.

The Swedish Twin Registry is funded by a grant from the Department of Higher Education, the Swedish Scientific Council, and AstraZeneca. Research resources provided by the Stockholm County Council and Stockholm Center for Public Health have supported this work as well as the Swedish Council for Working Life and Social Research.

Back to Top | Article Outline


1. Arden, N. K., and T. D. Spector. Genetic influences on muscle strength, lean body mass, and bone mineral density: a twin study. J. Bone Miner. Res. 12:2076-2081, 1997.
2. Beunen, G., and M. Thomis. Genetic determinants of sports participation and daily physical activity. Int. J. Obes. Relat. Metab. Disord. 23 (Suppl 3):S55-S63, 1999.
3. Boomsma, D. I., M. B. van den Bree, J. F. Orlebeke, and P. C. Molenaar. Resemblances of parents and twins in sports participation and heart rate. Behav. Genet. 19:123-141, 1989.
4. Evans, D. M., and N. G. Martin. The validity of twin studies. GeneScreen 1:77-79, 2000.
5. Frederiksen, H., and K. Christensen. The influence of genetic factors on physical functioning and exercise in second half of life. Scand. J. Med. Sci. Sports 3:9-18, 2003.
6. Kaprio, J., M. Koskenvuo, and S. Sarna. Cigarette smoking, use of alcohol, and leisure-time physical activity among same-sexed adult male twins. Prog. Clin. Biol. Res. 69:37-46, 1981.
7. King, A. C., S. N. Blair, D. E. Bild, et al. Determinants of physical activity and interventions in adults. Med. Sci. Sports Exerc. 24(6 Suppl):S221-S236, 1992.
8. Koopmans, J. P., L. J. P. van Doornen, and D. I. Boomsma. Smoking and sports participation. In: Genetic Factors in Coronary Heart Disease. U. Goldbourrt, U. de Faire, and K. Berg, (Eds.). Dordrecht: Kluwer Academic Publishers, pp. 217-235, 1994.
9. Lauderdale, D. S., R. Fabsitz, J. M. Meyer, P. Sholinsky, V. Ramakrishnan, and J. Goldberg. Familial determinants of moderate and intense physical activity: a twin study. Med. Sci. Sports Exerc. 29:1062-1068, 1997.
10. Lichtenstein, P., U. De Faire, B. Floderus, M. Svartengren, P. Svedberg, and N. L. Pedersen. The Swedish Twin Registry: a unique resource for clinical, epidemiological and genetic studies. J. Intern. Med. 252:184-205, 2002.
11. Maia, J. A., M. Thomis, and G. Beunen. Genetic factors in physical activity levels: a twin study. Am. J. Prev. Med. 23(2 Suppl):87-91, 2002.
12. Medlund, P., R. Cederlöf, P. Medlund, et al. A New Swedish Twin Registry. Acta Med. Scand. 600:1-111, 1977.
13. Neale, M. C. MX: Statistical Modeling. 5th ed., Richmond, VA: Virginia Commonwealth University, 2002, p. 112.
14. Peacock, M., C. H. Turner, M. J. Econs, and T. Foroud. Genetics of osteoporosis. Endocr. Rev. 23:303-326, 2002.
15. Perusse, 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.
16. Reed, T., R. R. Fabsitz, J. V. Selby, and D. Carmelli. Genetic influences and grip strength norms in the NHLBI twin study males aged 59-69. Ann. Hum. Biol. 18:425-432, 1991.
17. Sallis, J. F., M. F. Hovell, and C. R. Hofstetter. Predictors of adoption and maintenance of vigorous physical activity in men and women. Prev. Med. 21:237-251, 1992.
18. Stubbe, J. H., D. I. Boomsma, and E. J. De Geus. Sports participation during adolescence: a shift from environmental to genetic factors. Med. Sci. Sports Exerc. 37:563-570, 2005.
19. Timmons, J. A., O. Larsson, E. Jansson, et al. Human muscle gene expression responses to endurance training provide a novel perspective on Duchenne muscular dystrophy. FASEB J. 19:750-760, 2005.
20. Varo, J. J., M. A. Martinez-Gonzalez, J. De Irala-Estevez, J. Kearney, M. Gibney, and J. A. Martinez. Distribution and determinants of sedentary lifestyles in the European Union. Int. J. Epidemiol. 32:138-146, 2003.
21. Wolfarth, B., M. S. Bray, J. M. Hagberg, et al. The human gene map for performance and health-related fitness phenotypes: the 2004 update. Med. Sci. Sports Exerc. 37:881-903, 2005.


©2006The American College of Sports Medicine