Body mass index is a highly heritable trait, but heritability estimates of BMI are lower in childhood because of the influence of shared environmental factors, in old-age because of unique environmental factors

Obesity and eating behavior from the perspective of twin and genetic research. Karri Silventoinen, Hanna Konttinen. Neuroscience & Biobehavioral Reviews, Volume 109, February 2020, Pages 150-165. https://doi.org/10.1016/j.neubiorev.2019.12.012

Highlights
• Body mass index (BMI, kg/m2) is a highly heritable and polygenic trait.
• Heritability increases after early childhood and is highest in early adulthood.
• Obesogenic micro- and macro-environments reinforce genetic variation.
• Candidate genes of BMI express in brain tissue, suggesting the importance of behavior.
• Emerging evidence suggests that genes can affect BMI through eating behavior traits.

Abstract: Obesity has dramatically increased during the last decades and is currently one of the most serious global health problems. We present a hypothesis that obesity is a neuro-behavioral disease having a strong genetic background mediated largely by eating behavior and is sensitive to the macro-environment; we study this hypothesis from the perspective of genetic research. Genetic family and genome-wide-association studies have shown well that body mass index (BMI, kg/m2) is a highly heritable and polygenic trait. New genetic variation of BMI emerges after early childhood. Candidate genes of BMI notably express in brain tissue, supporting that this new variation is related to behavior. Obesogenic environments at both childhood family and societal levels reinforce the genetic susceptibility to obesity. Genetic factors have a clear influence on macro-nutrient intake and appetite-related eating behavior traits. Results on the gene-by-diet interactions in obesity are mixed, but emerging evidence suggests that eating behavior traits partly mediate the effect of genes on BMI. However, more rigorous prospective study designs controlling for measurement bias are still needed.

Keywords: TwinsGeneticsObesityBMIEating behavior

7. Conclusions

A century of genetic family studies and a decade of GWA studies have dramatically increased our understanding on the genetic architecture of common obesity, eating behavior and their mutual associations. However, this increasing knowledge has also clearly demonstrated the challenges, especially when trying to understand the mechanisms of how genes affect BMI and other obesity indicators. BMI has been shown to be a highly heritable trait, but the heritability changes over the life course. The heritability estimates of BMI are lower in childhood and in old age as compared to early adulthood and middle-age. In childhood, the lower heritability is because of the influence of environmental factors shared by co-twins and in old-age because of environmental factors unique to each twin. The similar pattern of increasing influence of genetic factors and diminishing effect of the shared environment during late childhood and adolescence has been reported for many psychological traits, such as intelligence (Plomin and Deary, 2015), and probably reflects the changing dynamics of the interplay between genes and the environment. During adolescence, dependence on parents decreases, social networks widen, influence from peers become stronger and sensation-seeking increases (Ahmed et al., 2015; Kilford et al., 2016). This probably leads to the possibility to more freely create one’s own environment, including the environment influencing BMI, which is partly affected by genetically influenced preferences. There is a lot of evidence for this so-called active gene–environment correlation for psychiatric traits (Jaffee and Price, 2007), and genetic factors have been found to influence life events, also demonstrating the dependence of genes and environment (Kendler and Baker, 2007). However, for BMI the direct evidence on gene–environment correlations is still suggestive. Studies on the heritability of macro-nutrient intake and eating patterns suggest that shared environmental factors have effect on eating behavior in childhood and adolescence, and this influence disappears until adulthood. Twin and molecular genetic studies have shown that after early childhood new genetic variance emerges. It is very possible that this genetic variance is related to eating behavior when children can more independently regulate their own eating, but direct evidence is still lacking. There is some evidence that eating behaviors can modify the genetic effects of obesity, but most of these studies are based on cross-sectional data and the results are somewhat mixed. Thus, more studies on how the interplay between genes and the environment modifies the genetic architecture of BMI during the formative years of childhood and adolescence are still needed. The strong effect of genetic factors on BMI does not mean, however, that the family environment does not have effect on BMI. Adoption studies have clearly shown that the adoptive family also has an effect on BMI. A likely explanation for these results is that the family environment affects BMI by reinforcing the effect of genes affecting BMI. There is direct evidence on this based mainly on twin studies since both the micro-level environment (e.g., parental education) and the macro-level environment (measured as the level of obesity between countries and measurement years) affect the genetic variation of BMI. Thus, those children having a genetic susceptibility to obesity gain more weight in family environments or societies predisposing to obesity. These results underline the importance of community food environments, since they can suppress or reinforce the effects of genetic variants associated with obesity. There has been a lot of discussion on which specific community-level factors are behind the obesogenic environments, but there is no clear consensus (Kirk et al., 2009). The associations are also likely to be very complex, as found in a previous study demonstrating that the community food environment can modify how health counseling affects eating behavior (Lorts et al., 2019). There is a lack on studies whether the micro- and macro-environment can modify the genetic variation of macro-nutrient intake in a similar way as they affect the genetic variation of BMI. Thus, more research is needed to specify which community-level factors reinforce the genetic variation of BMI and analyze the role of eating behavior behind these associations. GWA studies have clearly shown that BMI is a highly polygenic trait and thus confirms the basic principle of genetic family studies. The mechanisms of how genes affect BMI are still poorly understood, but the expression of the candidate genes of BMI in the brain tissue suggests that they affect BMI through behavioral factors. There is also evidence based on both twin and GWA studies that genetic factors affect macronutrient intake and appetite-related eating behavior traits. However, to date, there is only limited direct evidence on the overlap of genes affecting BMI and eating behavior which would suggest that the genes affect BMI through eating behavior. Some studies have shown this mediation effect, but they can explain only a fraction of the association between genetic factors and BMI. This area is, however, very challenging because of the well-known difficulties to measure dietary intake and reliance on self-report scales to assess eating behavior traits. Some sex differences in the genetic architecture of obesity indicators were identified. In BMI the proportion of genetic variation was roughly similar in males and females from infancy to old age, but especially after puberty, somewhat different sets of genes started to affect BMI in males and females and this difference increased during adulthood. It is likely that this reflects differences in body composition since somewhat different sets of genes affect muscle and fat body tissues. Accordingly, the SNPs associated with WHR adjusted for BMI showed different effect sizes in males and females. Very little is still known on sex differences in the genetic architecture of eating behavior. Thus, it is too early to argue whether genetic factors affect obesity traits in males and females differently through eating behavior or whether the found differences reflect only endocrinological differences between the sexes. At the beginning of this review we presented the hypothesis: Obesity is a neuro-behavioral disease having a strong genetic background mediated largely by eating behavior and being sensitive to the macroenvironment. There is strong evidence for this hypothesis based on previous genetic research, but the evidence that the genes affect especially through eating behavior is still emerging and mainly indirect at the moment. More rigorous prospective study designs controlling the well-known biases of measuring food intake would be necessary to prove this part of the hypothesis or to show that other behavioral mechanisms are also important when explaining the effect of genes on BMI.