Several studies have estimated effects of body mass index (BMI) on labour market outcomes, and these studies have mixed findings. A significant challenge has been to adequately control for omitted variables, selection, reverse causality, and measurement error. We examine the impact of BMI on income using genetic variants as instrumental variables for BMI. Individual-level pre-tax income from tax records was merged with health survey data containing measured height and weight, and data on genetic variants. The analyses were stratified by sex and a variety of methods were used to explore the sensitivity and validity of the instrumental variable (IV) strategy. For females we found that BMI had a negative effect on the logarithm of income. The effect estimated from the IV models (−0.02) was larger than the effect estimated from naïve ordinary least squares regressions (−0.01). For males, the coefficients for the effect of BMI on income were imprecise, and both positive and negative coefficients were estimated depending on the estimation method. Our results suggest that females are susceptible to reduced income levels following increased BMI.

中文翻译：

---

体重指数和收入之间的关系：使用 HUNT 的遗传变异作为工具变量

几项研究估计了体重指数 (BMI) 对劳动力市场结果的影响，这些研究的结果好坏参半。一个重大挑战是充分控制遗漏变量、选择、反向因果关系和测量误差。我们使用遗传变异作为 BMI 的工具变量来检查 BMI 对收入的影响。来自税收记录的个人税前收入与包含测量的身高和体重以及遗传变异数据的健康调查数据合并。分析按性别分层，并使用各种方法来探索工具变量 (IV) 策略的敏感性和有效性。对于女性，我们发现 BMI 对收入的对数有负面影响。从 IV 模型估计的效果 (-0. 02) 大于从朴素普通最小二乘回归估计的效果 (-0.01)。对于男性，BMI 对收入影响的系数不精确，根据估计方法估计正系数和负系数。我们的研究结果表明，随着 BMI 的增加，女性的收入水平容易降低。