Skip Navigation


American Journal of Epidemiology Advance Access originally published online on October 13, 2006
American Journal of Epidemiology 2007 165(1):101-108; doi:10.1093/aje/kwj342
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
165/1/101    most recent
kwj342v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (9)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Jacobson, P.
Right arrow Articles by Bouchard, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Jacobson, P.
Right arrow Articles by Bouchard, C.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

American Journal of Epidemiology Copyright © 2006 by the Johns Hopkins Bloomberg School of Public Health All rights reserved; printed in U.S.A.

ORIGINAL CONTRIBUTIONS

Spouse Resemblance in Body Mass Index: Effects on Adult Obesity Prevalence in the Offspring Generation

Peter Jacobson1, Jarl S. Torgerson1, Lars Sjöström1 and Claude Bouchard2

1 Department of Body Composition and Metabolism, Sahlgrenska University Hospital, Göteborg, Sweden
2 Human Genomics Laboratory, Pennington Biomedical Research Center, Baton Rouge, LA

Correspondence to Dr. Claude Bouchard, Pennington Biomedical Research Center, 6400 Perkins Road, Baton Rouge, LA 70808-4124 (e-mail: bouchac{at}pbrc.edu).

Received for publication December 2, 2005. Accepted for publication May 10, 2006.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
 References
 
Accruing evidence indicates that mate selection is promoted by similarity in body fatness. Assortative mating for obesity may contribute genetically to the obesity epidemic by increasing the risk in subsequent generations. To test this hypothesis, the authors analyzed measured and validated questionnaire data on family members, obtained between 1987 and 2000 from 7,834 obese probands and from 829 subjects randomly ascertained from the general Swedish population. Spouse correlations in body mass index were strongest among couples with the shortest duration of cohabitation. Obesity concordance in parents was associated with an obesity prevalence of 20.1% in adult offspring compared with 1.4% if parents were concordantly nonobese (odds ratio = 18.3, 95% confidence interval: 9.0, 37.4). The prevalence was 8.2% if parents were obesity discordant (odds ratio = 6.5, 95% confidence interval: 3.2, 13.2). No association was found between rearing parents' and nonbiologic offspring's body mass index. These results agree with the hypothesis that assortative mating for obesity confers a higher risk of obesity in the offspring generation and thus contributes to the obesity epidemic. Parental obesity concordance is a strong, easily identifiable genetic risk factor that should be considered in the complex network of risk factors for obesity in designing primary prevention programs.

body mass index; genetic screening; genetics, population; marriage; obesity; spouses

Abbreviations: BMI, body mass index; SOS, Swedish Obese Subjects; XENDOS, Xenical in the Prevention of Diabetes in Obese Subjects


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
 References
 
Over the last decades, obesity prevalence has increased significantly in developed countries (1, 2). Although body fatness is influenced by both environmental and genetic determinants (3), the very recent history of the obesity epidemic has naturally singled out lifestyles and environmental factors, associated with positive energy balance, as the immediate culprits. On the other hand, the effect of any lifestyle factor appears to be modulated by genes, often referred to as gene-by-environment interaction. Support for this notion as applied to obesity was given by controlled experiments on monozygotic twins (4, 5), showing that persons with identical genotypes respond similarly regarding body mass and body composition to changes in energy balance, whereas the variability is larger among those whose genotypes differ. Thus, the current obesity epidemic may be viewed as an environmentally and lifestyle-driven increase in the penetrance of obesity susceptibility genes.

The prevalence and heterozygosity of susceptibility genes are expected to remain constant across generations if (and only if) mate choice occurs randomly with respect to associated phenotypes. Data from several studies, however, suggest that a portion of all unions is characterized by positive assortment for various measures of body mass and fatness (6–14).

Recently, Hebebrand et al. (11) hypothesized that the current obesity epidemic has a genetic component mediated by increased rates of assortative mating for body fatness. Theoretically, the genetic consequences of assortative mating for complex traits, such as obesity, are expected to become more significant as the correlation between genotype and phenotype (penetrance) increases, even if rates of assortative mating remain constant across generations.

In the present study, we investigated a large Swedish cohort for evidence of assortative mating for body mass index (BMI; calculated as body weight in kilograms divided by body height in meters squared). Furthermore, we tested the data for consistency with the hypothesis that spouse resemblance in BMI affects the prevalence of adult obesity in the subsequent generation.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 APPENDIX
 References
 
Study participants
The study probands were either obese patients, screened for inclusion eligibility for either of two Swedish nationwide obesity intervention studies—the Swedish Obese Subjects (SOS registry, n = 6,328) and the Xenical in the Prevention of Diabetes in Obese Subjects (XENDOS, n = 4,423)—or participants selected randomly from the general population, in a cross-sectional study (SOS reference study, n = 1,135). Participants in the SOS registry and XENDOS studies were ascertained through advertisements in the media.

Details of the study cohorts have been described previously, as referenced below. Briefly, the SOS study (15) included primarily severely obese subjects living in 18 of the 24 counties of Sweden. Recruitment occurred over the 1987–2000 period. Among inclusion criteria were age 37–60 years and BMI above 34 kg/m2 for males and 38 kg/m2 for females. Measurements took place at 480 primary health care centers. Mean age was 47.4 (standard deviation, 6.0) years, and mean BMI was 39.7 (standard deviation, 4.5) kg/m2.

The SOS reference study (16) was completed during 1994–1999 and included persons from the Swedish cities of Mölndal and Örebro. Following random selection from population registries within the 37–61-year age span, individuals were approached with a request to participate. Mean age was 49.4 (standard deviation, 6.8) years, and mean BMI was 25.3 (standard deviation, 3.8) kg/m2.

Inclusion criteria in the XENDOS study (17) included age 30–60 years and a BMI of 30 kg/m2 or more, and recruitment took place during 1997 at 22 outpatient clinics. Mean age was 44.7 (standard deviation, 8.0) years, and mean BMI was 36.8 (standard deviation, 3.9) kg/m2.

Body mass and stature measured prior to any interventions were used in the present study. The demographics of the study subjects are summarized in table 1.


View this table:
[in this window]
[in a new window]

  		TABLE 1 Age and BMI* by generation and sex among families ascertained either through an obese proband (Swedish Obese Subjects intervention study, 1987–2000, and the XENDOS* study, 1997) or randomly from the general population (The SOS reference study, 1994–1999)

 
Data for spouses and adult offspring regarding current body mass, stature, age, sex, duration of cohabitation, as well as biologic relatedness to offspring were obtained from questionnaires completed by the probands. After exclusion of probands who had not reported data on spouses, data for a total of 8,663 spouse pairs were available for analysis. The research projects were approved by all ethics committees concerned, and all participants gave written informed consent.

Statistical methods
BMI was adjusted by using a multiple regression procedure in which age, age squared, and age cubed were forwardly entered and, if significant at the 5 percent level, were retained as covariates in the model. The model was applied separately in six groups: male spouses, female spouses, biologic daughters, biologic sons, foster daughters, and foster sons. The means and variances of the residuals, which constituted the adjusted phenotype, were standardized to the original scale for the purpose of graphic display.

Spouse resemblance was assessed by using Spearman rank correlation. This nonparametric method was chosen to avoid confounding by nonnormal trait distributions. Ninety-five percent confidence intervals for the correlation coefficients were computed by using Fisher's z transformation. Bivariate analyses of the SOS reference population were based on sex-specific quintile ranks obtained from the same population. A chi-squared test was used to investigate the magnitude and statistical significance of deviations from the null hypothesis of random mating. Because the SOS reference, the SOS registry, and the XENDOS studies used different BMI eligibility criteria, these studies were pooled to cover a wide range of BMI values. All analyses were performed by using SAS version 8.2 software (SAS/STAT Users' Manual; SAS Institute, Inc., Cary, North Carolina).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 APPENDIX
 References
 
Validity of reported anthropometric data
The validity of BMI, calculated from body mass and stature in spouses and adult offspring as reported by probands, was assessed in an ongoing family ancillary study within the SOS project. Subjects completed questionnaires identical to those used in the current study. The BMI of 135 spouses and 146 biologic adult offspring measured 1–30 days following questionnaire completion was compared with reported data. Results are shown in figure 1. Average difference between reported and measured BMI in spouses and offspring was –0.2 and –0.8 kg/m2, respectively. The negative sign indicates a tendency toward underestimation of BMI, as has been seen in previous validation studies of data from self-reports (18–23) and reports for relatives (24).


Figure 1
View larger version (10K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIGURE 1 Bland-Altman (27) plots showing agreement between measured body mass index (BMI) and BMI as calculated from height and weight reported by probands for their spouse (panel A) and biologic offspring (panel B) in a Swedish cohort of obesity-discordant sibpairs and their biologic parents, 1997–2003. The broken horizontal lines define the two-standard-deviation range about the mean (solid line) of the difference between measured and reported data.

 
Assortative mating
Spouse correlation in current age-adjusted BMI, based on all 8,663 pairs, was 0.18 (95 percent confidence interval: 0.16, 0.20). Spouse similarity does not necessarily indicate assortative mating because it might also be the effect of cohabitation. In contrast to assortative mating, a cohabitation effect should make spouses more similar over time. To investigate the nature of the spouse resemblance, spouse correlations were analyzed within five duration-of-cohabitation classes. As shown in figure 2, an assortative mating effect was strongly supported; the correlation was highest among pairs who had lived together 5 years or less, whereas correlation point estimates were lower for pairs with a longer duration of cohabitation. This decline in correlation was not related to interclass differences in age, BMI or year during which the proband was included in the study.


Figure 2
View larger version (6K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIGURE 2 Spearman rank correlations for current body mass index among 8,663 spouse pairs, categorized by duration of cohabitation. Pooled data from the Swedish Obese Subjects intervention and reference studies, 1987–2000, and the XENDOS study, 1997. n, number of pairs in each category. Body mass index was adjusted for age. Error bars represent 95% confidence intervals.

 
To investigate whether spouse resemblance was uniform across the entire BMI range, the 829 available pairs from the SOS reference study were analyzed categorically. (This analysis required that only the randomly ascertained pairs be used, since pairs from the obesity studies always included one obese proband, by definition.) Sex-specific BMI quintiles were computed, thus yielding equally large sex-by-quintile subgroups, with 4 percent frequency per pair-type expected under random mating. As illustrated in figure 3, deviations from this frequency were positive among quintile-concordant pairs and negative among discordant pairs (chi-square = 43.5, df = 16, p = 0.0002). Such deviations were evident across the whole range, but most prominently within the top and bottom quintiles. Thus, assortative mating appears to occur most frequently within the extremes of the BMI distribution in the general population.


Figure 3
View larger version (7K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIGURE 3 Associations between quintiles of body mass index (BMI) among 829 randomly ascertained spouse pairs. Positive (negative) deviations from 4% (the expected frequency given random mating) indicate positive (negative) nonrandom assortment. The Swedish Obese Subjects reference study, 1994–1999.

 
Adult obesity in the second generation
The relation between current spouse resemblance in obesity status and age-adjusted BMI in adult offspring, conditional on parent-offspring biologic relatedness, was investigated in families of the 4,118 probands who had reported weight and height of spouses and adult offspring. Offspring were classified as 1) reared by two biologic parents, 2) reared by one biologic and one foster parent, or 3) reared by two foster parents. Parental pairs were classified as 1) obesity concordant, 2) concordantly nonobese, or 3) obesity discordant. The interaction between the biologic relatedness vector and parental obesity status vector yielded nine offspring classes. Moreover, an additional offspring class was defined because cases for whom rearing parents were discordant for both obesity and biologic relatedness to the offspring were dichotomized depending on whether it was the biologic or nonbiologic parent who was obese. Figure 4 shows means of age-adjusted BMI among the 10 classes of offspring. The effect of parental obesity on offspring BMI appeared to totally depend on whether there was biologic relatedness to the parents. In offspring reared by their biologic parents, obesity prevalence was 20.1 percent (227/1,131) when both parents were obese (odds ratio = 18.3, 95 percent confidence interval: 9.0, 37.4) and 8.2 percent (413/50,41) when pairs were of the obese/nonobese type (odds ratio = 6.5, 95 percent confidence interval: 3.2, 13.2). The prevalence was 1.4 percent (8/592) when both biologic parents were nonobese. Although representing 16 percent of parental pairs, obese/obese pairs contributed 35 percent of all obese adult offspring.


Figure 4
View larger version (11K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIGURE 4 Body mass index (BMI) of 7,824 adults given obesity status in rearing parents and parent-offspring biologic relatedness. Offspring whose parents are discordant for both obesity and biologic relatedness are represented by paired bars: obese biologic parent (right), obese foster parent (left). Pooled data from the Swedish Obese Subjects intervention and reference studies, 1987–2000, and the XENDOS study, 1997. n, total number of offspring in each category along the horizontal axes. BMI was adjusted for age. Error bars represent 95% confidence intervals.

 
Simulation studies
Although the notion of a genetic component in the obesity epidemic was supported by the observation that assortative mating was most prominent in the upper BMI distribution and by the substantially higher obesity risk for biologic offspring from such matings, it is not evident that the observed deviations from random mating could affect the genotype at associated loci. We investigated this issue by simulating a normally distributed trait under additive influence of a biallelic quantitative trait locus in Hardy-Weinberg equilibrium, with the risk allele prevalence set to 30 percent. Trait decile-specific genotype probabilities were paired at frequencies identical to the decile mating-type frequencies observed in the population-based sample, and a new generation of trait values was iteratively simulated based on the updated set of genotype probablities. A detailed account of the simulation procedure is given in the Appendix. The effect over multiple generations on the prevalence of trait values exceeding 1, 2, and 3 standard deviations above the population mean is shown in figure 5. The slopes of the regression lines suggest that the highest per-generation incidence occurs at the uppermost extreme of the distribution.


Figure 5
View larger version (20K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIGURE 5 Results from simulations showing generation-wise increases in prevalence of trait values 1, 2, and 3 standard deviations (SD) above the population mean as predicted from the trait decile mating pattern observed in a population-based sample (The Swedish Obese Subjects reference study, 1994–1999). Regression lines for each category are shown. The model assumes an additive biallelic quantitative trait locus with the three genotypes in Hardy-Weinberg equilibrium at baseline. The trait was set to be normally distributed about each genotype-specific mean. At generation 1, the trait means in homozygotes for the major allele (49% of the observations), heterozygotes (42%), and homozygotes for the minor allele (9%) are –1, 0, and 1, respectively, all with SD = 1. The increases are relative to the pooled baseline trait distribution. A total of 106 observations per generation were simulated. Refer to the Appendix for further details.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
APPENDIX
 References
 
In the present study, we tested a large body of measured or validated questionnaire data for consistency with the hypothesis that assortative mating for obesity occurs and contributes to the obesity epidemic.

Ideally, assessment of mate selection should be based on data obtained when selection took place. Because such information was unavailable, we used current spouse BMI resemblance by duration of cohabitation as a proxy. The finding that BMI correlation was strongest among the couples with the shortest duration of cohabitation provides evidence that spouse resemblance was the consequence of assortative mating rather than the sharing of a common environment.

Although assortative mating for BMI occurred across the whole range, it was more pronounced at the extremes. This finding is in agreement with published data from the Canada Fitness Survey (13). Whereas an excess of obese-obese matings is consistent with the observed increase in obesity prevalence, the expected increase in leanness due to an excess of lean-lean matings has not been reported before to our knowledge. This finding might be due to heterogeneity in the etiology of leanness, that is, the lean population constitutes a mix of yet-unexposed and exposed, but genetically obesity-resistant people, thus potentially diminishing the genetic effects of assortative mating for leanness to the point that it is not always detectable, particularly in smaller studies. While it is obvious that a lean phenotype does not absolutely require a particular genotype, it is also evident (from this study, among others) that an obesogenic environment does not inevitably lead to obesity. Consequently, one could hypothesize that assortative mating for obesity may have a greater genetic component than would assortative mating for leanness.

Direct evidence to the effect that assortative mating for obesity contributes genetically to the obesity epidemic will ultimately be obtained by molecular genetic studies over multiple generations, in which secular changes in allelic distributions at associated loci can be observed. The simulations, albeit simplified, indicated that modest deviations from random mating, as observed in this study, are sufficient to bring about appreciable changes at an associated locus. Indirect evidence to this effect was also provided; obese-obese pairs were more common than expected, and offspring reared by them had the highest mean BMI and obesity prevalence, but only if there was biologic kinship to the offspring. Since this analysis was based on current (rather than premarital) spouse resemblance, and since only adult offspring BMI was available, the minimal spouse duration of cohabitation was 17 years. Consequently, parental trait similarity did not necessarily equal premarital similarity. Given the tendency of BMI to increase with age, the most likely bias was introduced by the unrealistic assumption that all currently obese parents were already obese at the time of mate selection. However, it is unlikely that the results were driven by this bias, since that would imply that children whose parents became obese later in life have a higher obesity risk than children of parents with earlier-onset obesity. Conversely, inclusion of parents with late-onset obesity more likely resulted in underestimation of the obesity risk for offspring, because late-onset obesity appears to be less heritable than early-onset obesity (25).

Our results agree with previously reported evidence of assortative mating as a risk factor for obesity in offspring. In a study on German children hospitalized because of extreme obesity, it was observed that their parents' BMI, whether contemporaneously measured or recalled at age 20 and 30 years, tended to cluster in the upper distribution (11). A population-based study in Canada showed stronger correlation in BMI among parents of children in the upper or lower extremes of the distribution (13). Recently, an English study found that parental obesity was the strongest of eight risk factors for obesity at age 7 years (26). None of those studies provided any data to indicate that assortative mating influences obesity prevalence, for example, that parental obesity concordance was more common than expected. Furthermore, we investigated the effects of parental obesity on an offspring's risk in adulthood. Finally, our results highlight the fact that biologic inheritance is a requirement for the transfer of obesity risk across generations; that is, exposure to an obesity-promoting environment alone is not always sufficient to cause obesity.

In conclusion, these data support the hypothesis that positive assortative mating for obesity includes a genetic component that may partly explain the current obesity epidemic. The observation that spouse concordance for obesity was associated with a 20-fold higher obesity risk for biologic offspring compared with children of concordantly nonobese parents may provide a simple, yet powerful screening tool. Therefore, we propose that biologic mothers' and fathers' BMI be considered in early identification of persons who would benefit the most from targeted primary and secondary obesity prevention programs.


   APPENDIX
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
References
 
Simulation of Genetic Nonrandom Mating Effects on a Quantitative Trait Locus
Consider an additive quantitative trait locus with alleles A and a with probabilities p and q, respectively. Let the genotypes AA, aA, and aa be in Hardy-Weinberg equilibrium (p2 + 2pq + q2 = 1), with corresponding trait means –1, 0, and 1, and each with a standard deviation of 1. Rank the pooled simulated trait values into b quantiles, and estimate for each quantile the genotype probabilities of AA, aA, and aa. Let the vector

Formula
denote genotype probabilities for males in the ith quantile and

Formula
for females in the jth quantile. The probability matrix of genotype unions from the ijth mating type, Gij, is defined by the outer product

Formula
For offspring of the ijth mating type, the probability of AA is given by

Formula
whereas the offspring probability of aa is given by

Formula
(Note that Gij·A represents element-wise multiplication.) To obtain the grand genotype probabilities (kAA, kaA, kaa) in offspring, let pAAij form the elements of a quadratic i by j matrix

Formula
and use for each element a weight proportional to the mating probability qij (as estimated from the observed mating frequencies in the population under study)

Formula
(Similarly to above, PAA·N is element-wise multiplication.) The corresponding probabilities of aa and aA are given by

Formula
Based on the new genotype probability estimates, a new data set is simulated, and the procedure is iterated by using the N matrix.


   ACKNOWLEDGMENTS
 
Supported by a grant (05239) from the Swedish Medical Research Council and by grants from Hoffman-La Roche, Basel, Switzerland (to Dr. Sjöström), and by an unrestricted grant from Bristol-Myers Squibb and the George A. Bray Chair in Nutrition (to Dr. Bouchard).

Conflict of interest: none declared.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
 References
 

  1. Flegal KM, Carroll MD, Kuczmarski RJ, et al. (1998) Overweight and obesity in the United States: prevalence and trends, 1960 –1994. Int J Obes Relat Metab Disord 22:39–47.[CrossRef][Web of Science][Medline]
  2. Seidell JC. (2002) Prevalence and time trends of obesity in Europe. J Endocrinol Invest 25:816–22.[Web of Science][Medline]
  3. Bouchard C. (1991) Current understanding of the etiology of obesity: genetic and nongenetic factors. Am J Clin Nutr 53:1561S–1565S.[Abstract/Free Full Text]
  4. Bouchard C, Tremblay A, Després JP, et al. (1990) The response to long-term overfeeding in identical twins. N Engl J Med 322:1477–82.[Abstract]
  5. Bouchard C, Tremblay A, Després JP, et al. (1994) The response to exercise with constant energy intake in identical twins. Obes Res 2:400–10.[Medline]
  6. Spuhler JN. (1982) Assortative mating with respect to physical characteristics. Soc Biol 29:53–66.[Web of Science][Medline]
  7. Tambs K, Moum T, Eaves LJ, et al. (1992) Genetic and environmental contributions to the variance of body height in a sample of first and second degree relatives. Am J Phys Anthropol 88:285–94.[CrossRef][Web of Science][Medline]
  8. Knuiman MW, Divitini ML, Bartholomew HC, et al. (1996) Spouse correlations in cardiovascular risk factors and the effect of marriage duration. Am J Epidemiol 143:48–53.[Abstract/Free Full Text]
  9. Allison DB, Neale MC, Kezis MI, et al. (1996) Assortative mating for relative weight: genetic implications. Behav Genet 26:103–11.[CrossRef][Web of Science][Medline]
  10. Katzmarzyk PT, Pérusse L, Rao DC, et al. (1999) Spousal resemblance and risk of 7-year increases in obesity and central adiposity in the Canadian population. Obes Res 7:545–51.[Web of Science][Medline]
  11. Hebebrand J, Wulftange H, Goerg T, et al. (2000) Epidemic obesity: are genetic factors involved via increased rates of assortative mating? Int J Obes Relat Metab Disord 24:345–53.[CrossRef][Web of Science][Medline]
  12. Magnusson PK and Rasmussen F. (2002) Familial resemblance of body mass index and familial risk of high and low body mass index. A study of young men in Sweden. Int J Obes Relat Metab Disord 26:1225–31.[CrossRef][Web of Science][Medline]
  13. Katzmarzyk PT, Hebebrand J, Bouchard C. (2002) Spousal resemblance in the Canadian population: implications for the obesity epidemic. Int J Obes Relat Metab Disord 26:241–6.[CrossRef][Web of Science][Medline]
  14. Silventoinen K, Kaprio J, Lahelma E, et al. (2003) Assortative mating by body height and BMI: Finnish twins and their spouses. Am J Hum Biol 15:620–7.[CrossRef][Web of Science][Medline]
  15. Sjöström L, Lindroos AK, Peltonen M, et al. (2004) Lifestyle, diabetes, and cardiovascular risk factors 10 years after bariatric surgery. N Engl J Med 351:2683–93.[Abstract/Free Full Text]
  16. Larsson I, Berteus Forslund H, Lindroos AK, et al. (2004) Body composition in the SOS (Swedish Obese Subjects) reference study. Int J Obes Relat Metab Disord 28:1317–24.[CrossRef][Web of Science][Medline]
  17. Torgerson JS, Arlinger K, Käppi M, et al. (2001) Principles for enhanced recruitment of subjects in a large clinical trial. The XENDOS (XENical in the prevention of Diabetes in Obese Subjects) study experience. Control Clin Trials 22:515–25.[CrossRef][Web of Science][Medline]
  18. Boström G and Diderichsen F. (1997) Socioeconomic differentials in misclassification of height, weight and body mass index based on questionnaire data. Int J Epidemiol 26:860–6.[Abstract/Free Full Text]
  19. Giacchi M, Mattei R, Rossi S. (1998) Correction of the self-reported BMI in a teenage population. Int J Obes Relat Metab Disord 22:673–7.[CrossRef][Web of Science][Medline]
  20. Niedhammer I, Bugel I, Bonenfant S, et al. (2000) Validity of self-reported weight and height in the French GAZEL cohort. Int J Obes Relat Metab Disord 24:1111–18.[CrossRef][Web of Science][Medline]
  21. Gunnell D, Berney L, Holland P, et al. (2000) How accurately are height, weight and leg length reported by the elderly, and how closely are they related to measurements recorded in childhood? Int J Epidemiol 29:456–64.[Abstract/Free Full Text]
  22. Spencer EA, Appleby PN, Davey GK, et al. (2002) Validity of self-reported height and weight in 4808 EPIC-Oxford participants. Public Health Nutr 5:561–5.[CrossRef][Web of Science][Medline]
  23. Wada K, Tamakoshi K, Tsunekawa T, et al. (2005) Validity of self-reported height and weight in a Japanese workplace population. Int J Obes Relat Metab Disord 29:1093–9.[CrossRef][Web of Science][Medline]
  24. Sørensen TI, Stunkard AJ, Teasdale TW, et al. (1983) The accuracy of reports of weight: children's recall of their parents' weights 15 years earlier. Int J Obes 7:115–22.[Web of Science][Medline]
  25. Stunkard AJ, Sørensen TIA, Hanis C, et al. (1986) An adoption study of human obesity. N Engl J Med 314:193–8.[Abstract]
  26. Reilly JJ, Armstrong J, Dorosty AR, et al. (2005) Early life risk factors for obesity in childhood: cohort study. BMJ 330:1357.[Abstract/Free Full Text]
  27. Bland JM and Altman DG. (1986) Statistical method for assessing agreement between two methods of clinical measurement. Lancet i:307–10.

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Am. J. Clin. Nutr.Home page
C. Bouchard
Childhood obesity: are genetic differences involved?
Am. J. Clinical Nutrition, May 1, 2009; 89(5): 1494S - 1501S.
[Abstract] [Full Text] [PDF]

Home page
 Am J EpidemiolHome page
A. Di Castelnuovo, G. Quacquaruccio, M. B. Donati, G. de Gaetano, and L. Iacoviello
Spousal Concordance for Major Coronary Risk Factors: A Systematic Review and Meta-Analysis
Am. J. Epidemiol., January 1, 2009; 169(1): 1 - 8.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Clin. Nutr.Home page
J. R Speakman, K. Djafarian, J. Stewart, and D. M Jackson
Assortative mating for obesity
Am. J. Clinical Nutrition, August 1, 2007; 86(2): 316 - 323.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
165/1/101    most recent
kwj342v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (9)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Jacobson, P.
Right arrow Articles by Bouchard, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Jacobson, P.
Right arrow Articles by Bouchard, C.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?