Obesity Driven by Genes Less Harmful for the Heart?

It’s less nature and more about noshing when it comes to the adverse impact of obesity on the heart, a twin study from Sweden suggested.

In an analysis of nearly all twins in the country, obesity predicted cardiovascular disease (CVD) risk across the board in every genetic risk group, reported Ida Karlsson, PhD, of the Karolinska Institutet in Stockholm, and colleagues in eClinicalMedicine.

However, obesity in midlife despite having a low polygenic risk score for it — suggesting a greater role of environmental and lifestyle causes than genetics for that individual — was a more powerful predictor of CVD than obesity in conjunction with a high polygenic risk score (HR 2.08 vs 1.55).

“While a healthy lifestyle is always to strive for, findings from the current study and previous work indicate that obesity influenced by environmental factors may be more deleterious than obesity influenced by genetic factors,” the group concluded.

More fully controlling for confounding by looking just at identical twins within the cohort erased the difference in the link between obesity and polygenic risk score across the categories.

“Importantly, this indicates that the negative health effects of obesity may be influenced by other factors, rather than by the obesity in itself, as we would otherwise expect similar effects of obesity, regardless of if it is predicted by genetic predisposition or environmental factors,” Karlsson and colleagues suggested.

And it’s not an isolated finding, they noted, pointing to heterogeneity in genetically driven versus environmentally driven obesity seen for important outcomes, including dementia and mortality, in U.S. data.

The study encompassed data from sub-studies of the Swedish Twin Registry, a population-based register of virtually all twins born in the country. The analysis included 15,786 Swedish twins born before 1959 who had BMI measured at ages 40-64 and 5,488 with BMI measured at age 65 or later (3,286 were in both groups). CVD was determined from prospective data in linked nationwide healthcare and cause of death registries through 2016, for a follow-up averaging 18 years.

All the included registry sub-studies (conducted between 1984 and 2010) had genotyping available to generate individuals’ polygenic score for BMI. That risk score came from the most recent genome-wide association study for BMI, which identified 941 genetic variants significantly associated with BMI that together explained 6-14% of the trait variance.

Each standard deviation higher polygenic risk score was associated with a 1.23 (95% CI 1.07-1.18) units higher BMI at midlife and a 1.16 (95% CI 1.07-1.26) units higher BMI in late life, adjusted for sex and age when BMI was measured. Adjustment for smoking and education didn’t impact the estimates. The CVD risk was 14% lower among people with obesity when the polygenic risk score was elevated by one standard deviation versus average.

“Results were similar when obesity was measured in late life, but suffered from low power,” Karlsson and team noted, although they acknowledged that “the causes and consequences of overweight or obesity in late life are more complicated, with evidence of an inverse association with … mortality.”

To determine genetic factors not captured by those variants in the polygenic risk score, the researchers also looked at pairs of twins: 3,124 fraternal and 2,020 identical twins with BMI evaluated in midlife. Of them, 769 and 443 pairs, respectively, were discordant for CVD status.

“As twins share both DNA (to a varying extent), in utero environment, and early life environment, the co-twin control design elegantly controls for such confounding,” the researchers wrote. Prior twin studies have estimated the heritability of BMI at 45-85%.

In these analyses, the associations were blunted. The hazard ratios for the impact of obesity on CVD risk was 1.14 (95% CI 0.64-2.02) within identical twin pairs and 1.64 (95% CI 1.11-2.42) within fraternal twin pairs.

When the identical twin pairs were sorted into polygenic risk score groups, the difference disappeared in the CVD-obesity link between those genetically predicted to have a low versus high BMI (HR 1.21 and 1.29, respectively).

“If effect estimates remain stable within twin pairs (especially within monozygotic twin pairs, sharing identical DNA) it indicates, but is not proof of, a causal association,” Karlsson and team noted. “In contrast, if the effect estimate is close to zero within twin pairs, it is strong evidence against a causal association, as it indicates that the association is driven by genetic or other familial confounding.”

Limitations of the study included the reliance on some self-measured height and weight data in the analyses for calculating BMI, with the potential for some misclassification that would have biased the findings toward the null if non-differential in relation to the outcome. In addition, the registries didn’t have primary care data, which would have precluded finding associations with milder CVD.

“Even though we all know that it takes more than exercise and diet to combat obesity, there’s still a large stigma attached to it,” Karlsson said in a statement. “I think much could be gained by focusing on what has caused the obesity and what we can do to reduce the risk of comorbidities in each individual instead of mainly focusing on BMI.”