Interactions between physical activity and variants of the genes encoding uncoupling proteins −2 and −3 in relation to body weight changes during a 10-y follow-up

Abstract

OBJECTIVE:

To examine interactions between physical activity and possibly functional variants of the genes encoding uncoupling proteins −2 and −3 in relation to body weight change. We hypothesize that physical inactivity acts synergistically with a 45 bp insertion variant in the 3′untranslated region (3′UTR) of the UCP2-gene and with a t-allele of codon −55 in the promoter of the UCP3-gene in relation to subsequent weight change.

DESIGN:

Population-based longitudinal study of cohorts of juvenile obese and nonobese men, who were identified at the mandatory draft board examination in Copenhagen and adjacent regions at a median age of 19 y in 1943–77 and later examined at general health surveys in 1981–83 and 1991–93. The juvenile obese cohort included 568 men who at the draft board had a BMI ≥31 kg/m2 and the cohort of controls included 717 randomly selected draftees.

MEASUREMENTS:

Height and weight were measured, and information about physical activity was collected from a self-administered questionnaire. The genotyping of the polymorphisms was performed using RFLP techniques. The main outcome measure was change in BMI during the 10-y follow-up period. Additional outcome measures were obesity, waist circumference and body fat mass index measured at follow-up.

RESULTS:

Physical activity, the 3′UTR insertion polymorphism and the −55 c/t polymorphism were not consistently associated with changes in BMI, and there were no evidence for interactions between the UCP-variants and physical activity in relation to changes in BMI. No evidence for interaction between the UCP-variants and physical activity was found in relation to the additional obesity measures.

CONCLUSION:

This study does not support that interactions between physical activity and variants in the UCP2- or UCP3-gene are major determinants of subsequent weight changes in Danish Caucasian men.

Introduction

Family, adoption and twin studies have shown that obesity is under strong genetic influence.1 On the other hand, there is no doubt that environmental factors also influence the development obesity.2 Such findings suggest that obesity is the result of interaction between genetic predispositions and environmental influences, although the specific evidence for gene–environment interactions has been elusive.3, 4, 5

Physical inactivity is a behaviour, which is dependent on the environment, and is assumed to play an important role in the pathogenesis of obesity. Several cross-sectional studies have shown an inverse relationship between physical activity and obesity measures, but the results of prospective studies of the relationship are inconsistent.6, 7, 8, 9

Uncoupling proteins 2 and 3 uncouple the oxidative phosphorylation by increasing the proton conductance of the inner mitochondrial membrane, leading to heat production rather than energy storage.10, 11, 12 The genes encoding these proteins, UCP2 and UCP3, are therefore considered as candidate genes for obesity, because loss of function mutations in these genes are predicted to cause obesity.11 A 45 bp insertion polymorphism is present in the 3′untranslated region (3′UTR) of the UCP2-gene.11, 13 The insertion genotype has been associated with increased BMI in some,14, 15 but not in other populations.13, 16, 17 Some investigators have suggested a functional role of the insertion variant, because it decreases UCP2 mRNA expression,17 while others failed to show an altered function of the variant.14 Associations between a −55 c/t polymorphism in the UCP3 promoter and BMI have been reported,18, 19 while other studies find no association.14, 15, 20, 21, 22 A functional role of the variant has been suggested, because skeletal muscle UCP3 mRNA expression has been reported higher in males, who are hetero- and homozygous for the t-allele compared to males who are homozygous for the c-allele.20

A relation between physical activity and UCP-gene expression has been indicated.19, 23, 24 One study found that physical activity downregulates UCP expression, thereby making physically active people more energy efficient.23 This is in line with a Danish study, where subjects with a val/val variant in codon 55 of UCP2 had lower 24-h energy expenditure and higher energy efficiency during physical activity than subjects with the ala/val variant. In this study, it was hypothesized that a high-energy efficiency in val/val-genotypes during exercise could lead to a high level of activity, and thereby counteract their genetic susceptibility for obesity.24

Such counteracting effects could minimize the possibility of detecting associations between the UCP-variants and obesity in studies where the influence of physical activity is not considered. This view is supported by a French cross-sectional study, which found an interaction between self-reported physical activity and the −55 c/t polymorphism in the UCP3 promoter in relation to BMI. In this study, morbidly obese −55 c/c genotype carriers were the only subjects in whom BMI was inversely associated with physical activity.19

Therefore, we investigated if associations between the UCP2 and UCP3-genotypes and weight changes were dependent on the level of physical activity on the basis of the hypothesis that physical inactivity and a decreased function of the UCP-genes act synergistically in the pathogenesis of weight gain. We examined the interaction between leisure time physical activity and the 3′UTR insertion variant of the UCP2-gene and the −55 c/t variant of the UCP3-gene, respectively, in relation to changes in BMI in a cohort of juvenile obese men and a cohort of randomly sampled controls during a 10-y follow-up. In addition, we examined these interactions in relation to other obesity measures: BMI ≥30 (kg/m2), waist circumference (cm) and body fat mass index (BFMI) (kg fat/m2) at follow-up.

Methods

Study population

The study population was comprised of Danish men examined at the mandatory draft board examination in Copenhagen and surrounding counties from 1943 to 1977. Two groups were selected at the draft board examination: a group of men with juvenile onset obesity and a nonobese control group. All men with a BMI ≥31 kg/m2 (corresponding to 35% overweight according to a national standard in use when the obese sample was identified) at draft board were included in the juvenile obese group. The nonobese control group was selected as an approximately 1% random sample of all men at draft board examination in the same region and time period, from which those with BMI ≥31 kg/m2 were excluded. All men in the juvenile obese group and half of the control subjects, who were alive and still residing within the same region, were invited to participate in two general health surveys conducted by the second Copenhagen City Heart Study in 1981–83 (CCHS-81), and the third Copenhagen City Heart Study in 1991–93 (CCHS-91).

Follow-up studies (CCHS-81 and CCHS-91)

In CCHS-81, 964 men from the juvenile obese group and 1135 control subjects participated, and in CCHS-91, 791 from the juvenile obese group and 918 from the control group participated. The present study has baseline at CCHS-81 and follow-up at CCHS-91, and was performed separately in the juvenile obese and control group.

At both surveys, height and weight were measured, blood samples were drawn and information about physical activity was obtained from a self-administered questionnaire. At CCHS-91 bioimpedance and waist circumference were also measured. Details of the study design and methods have been published elsewhere.25

Height was measured without shoes, to the nearest half centimetre, and body weight was measured to the nearest decimal in kilograms in light indoor clothing and without shoes. BMI was calculated as the weight (kg) per height squared (m2), and change in BMI was calculated as (BMI at CCHS-91−BMI at CCHS-81) (kg/m2).

Obesity at follow-up was defined as BMI ≥30 kg/m2. Waist circumference (cm) was measured at the level of umbilicus with the subject standing without clothes and with relaxed breathing. Body fat mass (kg) was calculated from the bioimpedance measure by a sex-specific algorithm defined by Heitmann.26 A body fat mass index (BFMI) (kg/m2) was calculated as fat mass (kg) per height squared (m2).

Genomic DNA was obtained from blood samples drawn at CCHS-91. Genotyping of the 3′UTR insertion/deletion (ID) polymorphism in the UCP2-gene and genotyping of the −55 c/t polymorphism in the 5′region of the UCP3-gene were carried out using PCR-RFLP as described previously.13, 22 The UCP2- and UCP3-genotypes were shown to be in Hardy–Weinberg equilibrium.13, 22 Genotypes were categorized as: UCP2—(1) deletion/deletion (DD) genotypes, (2) ID genotypes, (3) insertion/insertion (II) genotypes; UCP3—(1) −55 c/c genotypes, (2) −55 c/t genotypes, (3) −55 t/t genotypes. The group of II genotypes and the group of −55 t/t genotypes were small, and were merged with the groups of ID genotypes and −55 c/t genotypes, respectively. Our main analyses are based on these merged groups, and we thereby assume that the c- and deletion-allele are recessive or at least codominant. However, this mode of inheritance is not confirmed in the literature.11 Therefore, a categorization with DD, ID and II genotypes kept separate and with −55 c/c, −55 c/t and −55 t/t genotypes kept separate, assuming no mode of inheritance, was used in additional analysis. Furthermore, a categorization with ID and DD genotypes merged, and with −55 c/t and −55 c/c genotypes merged, assuming that the c- and deletion-allele are dominant, was used in additional analysis.

Physical activity in leisure time at CCHS-81 was graded in four levels based on a questionnaire developed by Saltin and Grimby:27 (1) inactivity: none or less than 2 h of exercise per week; (2) moderate activity: 2–4 h of light exercise per week, (3) high activity: 2–4 h of moderate exercise per week and (4) very high activity: more than 4 h of exercise per week. The utility of this questionnaire-based assessment of physical activity has recently been thoroughly discussed.8, 9 In this study, the group of level 4 was small, and it was therefore merged with the group of level 3 in our main analyses. Additional analyses were carried out with the four activity groups kept separate.

Age at CCHS-81 was included as a categorical variable: (1) 20–29 y, (2) 30–39 y and (3) 40–60 y.

We excluded all men for whom information on weight, height, age, genomic DNA or physical activity were not available. Complete information was then available in a sample of 717 controls and 568 juvenile obese subjects.

Statistical analysis

χ2 tests were used to analyse whether the level of physical activity and the UCP-genotypes differed between controls and juvenile obese subjects. A χ2 test was also used to analyse if the level of physical activity was associated with UCP2- or UCP3-genotypes.

Linear regression models were used to analyse, if interactions between the UCP-genotypes and physical activity in leisure time were associated with change in BMI from CCHS-81 to CCHS-91. All analyses were performed separately for the UCP2 and UCP3-genotypes, in the control and juvenile obese group, respectively. A first series of models included physical activity, UCP-genotype and age as categorical factors. A second series of models included physical activity, UCP-genotype and age as categorical factors and the interaction between physical activity and UCP-genotype. We checked the fit of the models to the data by examining the standardized residuals of each model. All residual plots showed no serious indication of systematic residual variation, and the probit-plots were close to straight lines. Log transformation of change in BMI improved all plots, but implied no changes in parameter estimates compared to the models on the original scale. We thus concluded that all models seemed to satisfy the model requirements.

In addition, linear regression analyses with waist circumference and BFMI at follow-up as outcome were performed. These linear regression models were similar to those with change in BMI as outcome, although BMI at CCHS-81 was included as a continuous factor in the analyses with BFMI and waist as outcome. Furthermore, logistic regression was used to analyse whether interactions between the UCP-genotypes and physical activity in leisure time were associated with obesity at follow-up, defined as BMI ≥30 kg/m2. These models were constructed similarly to those with waist circumference and BFMI as outcome.

We also analysed if interactions between the UCP-genotypes and physical activity in leisure time were associated with BMI in a cross-sectional design. These linear regression models were similar to those with change in BMI as outcome, although with BMI at CCHS-81 as outcome.

Finally, we compared the BMI distribution at CCHS-81 between those who remained in the study and those who dropped out in the follow-up period. This was carried out separately in the juvenile obese and control group.

All analyses were carried out using Statistical Package for Social Science (SPSS) for Windows version 11.0.

Results

Table 1 shows the distribution of age, BMI, physical activity in leisure time and the UCP-genotypes in the control and juvenile obese group, respectively. Few men were inactive in leisure time, although more juvenile obese than control subjects were inactive (Table 1). No differences in genotype frequencies were observed significant among the control and juvenile obese group (Table 1). No significant difference in the level of physical activity was found between the UCP2- or UCP3-genotypes in the juvenile obese or control group (Table 2 ). Furthermore, no significant difference in the level of physical activity was found between the three age groups, although the youngest group was slightly more active (data not shown).

Table 1 Distribution of age, BMI, physical activity and UCP2- and UCP3-genotypes in the control and juvenile obese subjects
Table 2 Difference in the level of physical activity between genotypes of the UCP2 3′UTR ID and genotypes of the UCP3 –55 c/t polymorphism in the control and juvenile obese subjects (number, %)

The linear regression analyses showed no significant difference between the UCP2-genotypes considering change in BMI from CCHS-81 to CCHS-91, when adjusted for age and physical activity. This was the case in both the control and juvenile obese group (Table 3 ). There was no consistent association across the level of physical activity in leisure time and changes in BMI, when adjusted for age and UCP2-genotypes, although the greatest increase in BMI was found in the group of men with the highest level of physical activity in both the control and juvenile obese group (Table 3). Furthermore, there was no significant difference between the three age groups considering change in BMI, although the greatest increase in BMI occurred among the youngest men in both the control and juvenile obese group (data not shown).

Table 3 Mean change in BMI (kg/m2) (95% confidence limits) from CCHS-81 to CCHS-91 in the juvenile obese and control subjects, when the UCP2 3′UTR ID polymorphism are crosstabulated by the level of physical activity

No interaction between the UCP2-genotype and physical activity in relation to change in BMI was found in the juvenile obese or control group (Table 3). Introducing the interaction in the linear regression model led to no significant improvement of the fit of the model. We thus concluded that there was no evidence of an interaction between the UCP2-genotype and physical activity in relation to change in BMI from CCHS-81 to CCHS-91 (Table 3).

The results from the linear regression analysis of the UCP3-genotype, physical activity, and the interaction between the UCP3-genotype and physical activity in relation to change in BMI from CCHS-81 to CCHS-91 were similar to the UCP2 results in both the control and obese group (Table 4 ).

Table 4 Mean change in BMI (kg/m2) (95% confidence limits) from CCHS-81 to CCHS-91 in the juvenile obese and control subjects, when the UCP3 –55 c/t polymorphism are crosstabulated by the level of physical activity

The analyses of the additional obesity measures showed no interactions between physical activity and UCP2- and UCP3-genotypes in relation to waist, BFMI or obesity at follow-up (BMI ≥30 kg/m2) (data not shown). Furthermore, the cross-sectional analysis showed no interactions between physical activity and UCP-genotypes in relation to BMI (data not shown).

In the presented results, II and ID UCP2-genotypes as well as −55 t/t and −55 c/t UCP3-genotypes were merged. However, analysis with II and ID and with −55 t/t and −55 c/t kept separate, as well as analysis with ID and DD and −55 c/t and −55 c/c genotypes merged, revealed results similar to those shown in Tables 2 and 3 (data not shown). Furthermore, analysis with active and very active subjects kept separate also revealed results similar to those shown in Tables 2 and 3 (data not shown).

The BMI at CCHS-81 was quite similar between controls who remained in the study (mean BMI 24.7 kg/m2) and controls who dropped out in the follow-up period (mean BMI 24.8 kg/m2). BMI at CCHS-81 was slightly higher in the juvenile obese who dropped out in the follow-up period (mean BMI 34.8 kg/m2) than in the juvenile obese who remained in the study (mean BMI 33.7 kg/m2).

Discussion

We have examined whether interactions between the UCP-genotypes and physical activity were associated with changes in BMI during a 10-y follow-up period in two samples of Danish men. Physical activity, UCP2- and UCP3-genotypes were not consistently associated with changes in BMI, and we found no evidence of an interaction between the UCP-genotypes and physical activity in relation to change in BMI. Furthermore, we found no evidence of interactions between the UCP-genotypes and physical activity in relation to BMI ≥30 (kg/m2), waist circumference (cm) and BFMI (kg/m2) measured at follow-up.

The literature of the relationship between physical activity and risk of later obesity has recently been reviewed,6, 7 as has the literature of genetic variation of uncoupling proteins in relation to diabetes and obesity.11 Our finding of no consistent association between baseline physical activity and change in BMI, and no association between the UCP2- and UCP3-genotypes and change in BMI, is compatible with the conclusions in these reviews.6, 7, 11 When neither physical activity nor UCP-genotypes were consistently associated with change in BMI, an interaction between physical activity and the UCP-genotypes seems unlikely. However, the focus of this study was that an interaction could be the reason for the absence of these associations, because of counteracting balanced effects between physical activity and the UCP-genes in relation to subsequent BMI changes. Such interacting effects could limit a detectable association to BMI change to a subset of particular susceptible individuals, but this was not the case in our data.

In contrast to our results, a cross-sectional study of a morbidly obese French population showed an interaction between physical activity and the −55 c/t variant of the UCP3-gene in relation to BMI. The study showed an inverse relation between physical activity and BMI for c/c genotypes, but no association between physical activity and BMI for t/c and t/t genotypes.19 The discrepancy between our study and the French results may in part be explained by the facts that the French study included mainly women who all were morbidly obese, and our study consisted exclusively of men with a broader range of obesity. An interaction between the UCP3-genotype and physical activity in relation to BMI could be gender specific, although the French study does not report an effect of gender. Furthermore, there might be different impacts of physical activity on obesity in a cross-sectional study compared to a prospective one. This interpretation is supported by findings of a clear inverse relationship between physical activity and obesity in cross-sectional studies, while prospective studies of the relation are inconsistent.6, 7, 8, 9 However, we found no interaction between physical activity and the investigated UCP-genotypes in relation to BMI when we examined the association cross-sectionally. Another reason for the discrepancy between our and the French results could be a type I error in the French study or a type II error in our study. With a conservative calculation, we expect that our data with a probability of 95% would capture an interaction resulting in a difference of 1.1 BMI change (kg/m2) and 3.1 BMI change (kg/m2) between the active −55 c/c genotypes and inactive −55 t/t and c/t genotypes, in the control and juvenile obese groups, respectively. The reported interaction in the French study resulted in a BMI difference of 4.5 kg/m2 between the inactive and active −55 c/c genotypes.

An important question is if our failure to find interactions is due to a crude measurement of physical activity. However, the questionnaire-based assessment of physical activity correlates reasonably well with maximal oxygen uptake,28 and is inversely correlated with BMI in cross-sectional studies.8, 9 Furthermore, earlier studies using the same measurement of activity have shown a strong predictive value in relation to total mortality and cardiovascular disease,29, 30 but not to obesity.8, 9 However, a high BMI has a strong predictive value to later self-reported inactivity.8, 9 Therefore, we expect that the physical activity measure would capture interactions with UCP-genotypes and their relation to later obesity, if they had existed in our data.

Another important question is if our failure to find interactions is due to nonfunctionality of our chosen UCP2- and UCP3-gene variants. Some studies have shown an effect of the variants on mRNA expression,17, 20 while other investigators failed to show differences.14 However, before questions about functionality can be answered properly, more studies are needed to elucidate whether the variants affect mRNA expression and especially whether the variants affect specific protein content or activity.

A concern with the study is that we do not have dietary data available. However, if we had dietary data, we might still not be able to take the energy intake into account in a proper manner. It is suggested that the difference in caloric intake between a weight gaining and a weight stable phase is so small that it cannot be observed by measuring energy intake and energy output. We therefore stress that the focus on physical inactivity as a risk factor for obesity is also based on other regulatory metabolic effects of physical activity in relation to obesity than the energy balance equation alone as suggested in a recent review.7

Another concern with the study is that the follow-up design may have introduced selection bias in the results. However, the difference in BMI at CCHS-81 between those who dropped out in the follow-up period and those who remained in the study were modest, which reassures, but do not exclude bias.

In conclusion, interactions between physical activity in leisure time and variants in the UCP-2- or UCP-3-genes have no major influence on subsequent weight changes in Danish Caucasian men.

References

  1. 1

    Maes HHM, Neale MC, Eaves LJ . Genetic and environmental factors in relative body weight and human adiposity. Behav Genet 1997; 27: 325–351.

    CAS  Article  Google Scholar 

  2. 2

    French SA, Story M, Jeffery RW . Environmental influences on eating and physical activity. Annu Rev Public Health 2001; 22: 309–335.

    CAS  Article  Google Scholar 

  3. 3

    Sørensen TIA, Echwald SM . Obesity genes. Identifying single genes involved in polygenic inheritance is not easy. BMJ 2001; 322: 631–632.

    Article  Google Scholar 

  4. 4

    Heitmann BL, Kaprio J, Harris JR, Rissanen A, Korkeila M, Koskenvuo M . Are genetic determinants of weight gain modified by leisure-time physical activity? A prospective study of Finnish twins. Am J Clin Nutr 1997; 66: 672–678.

    CAS  Article  Google Scholar 

  5. 5

    Verdich C, Clement K, Sørensen TIA . Nutrient–gene interactions in the control of obesity. In: Remacle C, Reusens B (eds) Functional Foods, Ageing and Degenerative Disease. Chapter 10, Woodhead Publishing; 2004. pp 223–260.

    Google Scholar 

  6. 6

    Fogelholm M, Kukkonen-Harjula K . Does physical activity prevent weight gain—a systematic review. Obes Rev 2000; 24: 131–143.

    Google Scholar 

  7. 7

    Saris WHM, Blair SN, van Baak MA, Eaton SB, Davies PSW, Di Pietro L, Fogelholm M, Rissanen A, Schoeller D, Swinburn B, Tremblay A, Westerterp KR, Wyatt H . How much physical activity is enough to prevent unhealthy weight gain? Outcome of the IASO 1st Stick Conference and Consensus Statement. Obes Rev 2003; 4: 101–114.

    CAS  Article  Google Scholar 

  8. 8

    Petersen L, Schnohr P, Sørensen TIA . Longitudinal study of the relation between physical activity and obesity in adults. Int J Obes Relat Metab Disord 2004; 28: 105–112.

    CAS  Article  Google Scholar 

  9. 9

    Bak H, Petersen L, Sørensen TIA . Physical activity in relation to development and maintenance of obesity in men with and without juvenile onset obesity. Int J Obes Relat Metab Disord 2004; 28: 99–104.

    CAS  Article  Google Scholar 

  10. 10

    Krauss S, Zhang CY, Lowell BB . A significant portion of mitochondrial proton leak in intact thymocytes depends on expression of UCP2. Proc Natl Acad Sci USA 2002; 8: 118–122.

    Article  Google Scholar 

  11. 11

    Dalgaard LT, Pedersen O . Uncoupling proteins: functional characteristics and role in the pathogenesis of obesity and Type II diabetes. Diabetologia 2001; 44: 946–965.

    CAS  Article  Google Scholar 

  12. 12

    Urhammer SA, Dalgaard LT, Sørensen TIA, Tybjærg_Hansen A, Echwald SM, Andersen T, Clausen JO, Pedersen O . Organisation of the coding exons and mutational screening of the uncoupling protein 3 gene in subjects with juvenile-onset obesity. Diabetologia 1998; 41: 241–244.

    CAS  Article  Google Scholar 

  13. 13

    Dalgaard LT, Sørensen TIA, Andersen T, Hansen T, Pedersen O . An untranslated insertion variant in the uncoupling protein 2 gene is not related to BMI and changes in body weight during a 26-year follow-up in Danish Caucasian men. Diabetologia 1999; 42: 1413–1416.

    CAS  Article  Google Scholar 

  14. 14

    Walder K, Norman RA, Hanson RL, Schrauwen P, Neverova M, Jenkinson CP, Easlick J, Warden CH, Pecqueur C, Raimbault S, Richquier D, Harper M, Silver K, Shuldiner AR, Solanes G, Lowell BB, Chung WK, Leibel RL, Pratley R, Ravussin E . Association between uncoupling protein polymorphism (UCP2–UCP3) and energy metabolism/obesity in Pima Indians. Hum mol Genet 1998; 7: 1431–1435.

    CAS  Article  Google Scholar 

  15. 15

    Cassel PG, Neverova M, Jammohamed S, Uwakwe N, Qureshi A, McCarthy MI, Saker PJ, Albon L, Kopelman P, Noonan K, Easlick J, Ramachandran A, Snehalatha C, Pecqueur C, Ricquier D, Warden C, Hitman GA . An uncoupling protein 2 gene variant is associated with a raised BMI but not type II diabetes. Diabetologia 1999; 42: 688–692.

    Article  Google Scholar 

  16. 16

    Otabe S, Clement K, Rich N, Warden C, Pecqueur C, Neverova M, Raimbault S, Guy-Grand B, Basdevant A, Ricquier D, Froguel P, Vasseur F . Mutation screening of the human UCP2 gene in normoglycemic and NIDDM morbidly obese patients: lack of association between new UCP2 polymorphisms and obesity in French Caucasians. Diabetes 1998; 47: 840–842.

    CAS  Article  Google Scholar 

  17. 17

    Esterbauer H, Schneitler C, Oberkofler H, Ebenbichler C, Paulweber B, Sandhofer F, Ladurner G, Hell E, Strosberg AD, Patsch JR, Krempler F, Patsch W . A common polymorphism in the promoter of UCP2 is associated with decreased risk of obesity in middle-aged humans. Nat Genet 2001; 28: 178–183.

    CAS  Article  Google Scholar 

  18. 18

    Halsall DJ, Luan J, Saker P, Huxtable S, Farooqi IS, Keogh J, Wareham NJ, O’Rahilly S . Uncoupling protein 3 genetic variants in human obesity: The c-55t promoter polymorphism is negatively correlated with BMI in a UK Caucasian population. Int J Obes Relat Metab Disord 2001; 25: 472–477.

    CAS  Article  Google Scholar 

  19. 19

    Otabe S, Clement K, Dina C, Pelloux V, Guy-Grand B, Froguel P, Vasseur . A genetic variation in the 5′ flanking region of the UCP3 gene is associated with BMI in humans in interaction with physical activity. Diabetologia 2000; 43: 245–249.

    CAS  Article  Google Scholar 

  20. 20

    Schrauwen P, Xia J, Walder K, Snitker S, Ravussin E . A novel polymorphism in the proximal UCP3 promoter region: effect on skeletal muscle UCP3 mRNA expression and obesity in male non-diabetic Pima Indians. Int J Obes Relat Metab Disord 1999; 23: 1242–1245.

    CAS  Article  Google Scholar 

  21. 21

    Meirhaeghe A, Amouyel P, Helbecque N, Cottel D, Otabe S, Froquel P, Vasseur F . An uncoupling protein 3 gene polymorphism associated with a lower risk of developing Type II diabetes and with atherogenic lipid profile in a French cohort. Diabetologia 2000; 43: 1424–1428.

    CAS  Article  Google Scholar 

  22. 22

    Dalgaard LT, Sørensen TIA, Drivsholm T, Borch-Johnsen K, Andersen T, Hansen T, Pedersen O . A prevalent polymorphism in the promoter of the UCP3 gene and its relationship to BMI and long term body weight change in the Danish population. J Clin Endocrinol Metab 2001; 86: 1398–1402.

    CAS  PubMed  Google Scholar 

  23. 23

    Schrauwen P, Troost FJ, Xia J, Ravussin E, Saris WHM . Skeletal muscle UCP2 and UCP3 expression in trained and untrained male subjects. Int J Obes Relat Metab Disord 1999; 23: 966–972.

    CAS  Article  Google Scholar 

  24. 24

    Buemann B, Schierning B, Toubro S, Bibby BM, Sørensen TIA, Dalgaard L, Pedersen O, Astrup A . The association between the val/ala-55 polymorphism of the uncoupling protein 2 gene and exercise efficiency. Int J Obes Relat Metab Disord 2001; 25: 467–471.

    CAS  Article  Google Scholar 

  25. 25

    Sonne-Holm S, Sørensen TIA, Jensen G, Schnohr P . Independent effects of weight change and attained body weight on prevalence of arterial hypertension in obese and non-obese men. BMJ 1989; 299: 767–799.

    CAS  Article  Google Scholar 

  26. 26

    Heitmann BL . Evaluation of body fat estimated from BMI, skinfolds and impedance. A comparative study. Eur J Clin Nutr 1990; 44: 831–837.

    CAS  Google Scholar 

  27. 27

    Saltin B, Grimby G . Physiological analysis of middle-aged and old former athletes. Comparison with still active athletes of same ages. Circulation 1968; 38: 1104–1115.

    CAS  Article  Google Scholar 

  28. 28

    Saltin B . Physiological effects of physical conditioning. In: Hansen AT, Schnohr P, Rose G (eds) Ischaemic Heart Disease. The Strategy of Postponement. Year Book Medical Publishers: Chicago and FADL's forlag: Copenhagen; 1977. pp 104–115.

    Google Scholar 

  29. 29

    Andersen LB, Schnohr P, Schroll M, Hein HO . All-cause mortality associated with physical activity during leisure time, work, sports, and cycling to work. Arch Intern Med 2000; 160: 1621–1628.

    CAS  Article  Google Scholar 

  30. 30

    Hippe M, Vestbo J, Hein HO, Borch-Johnsen K, Jensen G, Sørensen TIA . Familial predisposition and susceptibility to the effect of other risk factors for myocardial infarction. J Epidemiol Commun Health 1999; 53: 269–276.

    CAS  Article  Google Scholar 

Download references

Acknowledgements

The Danish Medical Research Council, The Danish Heart Foundation and The Danish National Research Foundation supported the study. We also thank the staff of the Copenhagen City Heart Study for their invaluable assistance in the two examinations.

Author information

Affiliations

Authors

Corresponding author

Correspondence to T I A Sørensen.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Berentzen, T., Dalgaard, L., Petersen, L. et al. Interactions between physical activity and variants of the genes encoding uncoupling proteins −2 and −3 in relation to body weight changes during a 10-y follow-up. Int J Obes 29, 93–99 (2005). https://doi.org/10.1038/sj.ijo.0802841

Download citation

Keywords

  • longitudinal study
  • UCP2-gene variant
  • UCP3-gene variant
  • physical activity

Further reading

Search