Evaluation of the causal effects between subjective wellbeing and cardiometabolic health: mendelian randomisation study

Abstract Objectives To investigate whether the association between subjective wellbeing (subjective happiness and life satisfaction) and cardiometabolic health is causal. Design Two sample, bidirectional mendelian randomisation study. Setting Genetic data taken from various cohorts comprised of the general population (mostly individuals of European ancestry, plus a small proportion of other ancestries); follow-up analysis included individuals from the United Kingdom. Participants Summary data were used from previous genome wide association studies (number of participants ranging from 83 198 to 339 224), which investigated traits related to cardiovascular or metabolic health, had the largest sample sizes, and consisted of the most similar populations while minimising sample overlap. A follow-up analysis included 337 112 individuals from the UK Biobank (54% female (n=181 363), mean age 56.87 years (standard deviation 8.00) at recruitment). Main outcome measures Subjective wellbeing and 11 measures of cardiometabolic health (coronary artery disease; myocardial infarction; total, high density lipoprotein, and low density lipoprotein cholesterol; diastolic and systolic blood pressure; body fat; waist to hip ratio; waist circumference; and body mass index). Results Evidence of a causal effect of body mass index on subjective wellbeing was seen; each 1 kg/m2 increase in body mass index caused a −0.045 (95% confidence interval −0.084 to −0.006, P=0.02) standard deviation reduction in subjective wellbeing. Follow-up analysis of this association in an independent sample from the UK Biobank provided strong evidence of an effect of body mass index on satisfaction with health (β=−0.035 unit decrease in health satisfaction (95% confidence interval −0.043 to −0.027) per standard deviation increase in body mass index, P<0.001). No clear evidence of a causal effect was seen between subjective wellbeing and the other cardiometabolic health measures, in either direction. Conclusions These results suggest that a higher body mass index is associated with a lower subjective wellbeing. A follow-up analysis confirmed this finding, suggesting that the effect in middle aged people could be driven by satisfaction with health. Body mass index is a modifiable determinant, and therefore, this study provides further motivation to tackle the obesity epidemic because of the knock-on effects of higher body mass index on subjective wellbeing.


Index of Supplementary Figures
. Two-sample MR of subjective wellbeing on cardiometabolic health using genome-wide significant SNPs as the instrument Page 2 Figure S2. Funnel plot of individual SNP effects of BMI on wellbeing.
Page 4 Figure S3. Leave-one-out analysis Page 5 Figure S4. Bias plots of association with baseline confounders Page 6 Index of Supplementary Tables  Table S1. Percentage sample overlap between the SSGAC GWAS for subjective wellbeing and each of the GWAS for cardiometabolic health Page 8 Table S2. Descriptions of GWAS demographics Page 9 Table S3. Linear regression of the three genome-wide significant SNPs for subjective wellbeing predicting happiness in the UK Biobank independent sample (N= 242,219) Page 11  Figure S1. Two-sample MR analysis: the effect of subjective wellbeing on cardiometabolic health outcomes using 3 genome-wide significant variants as the instrument for subjective wellbeing.
The genetic instrument for the exposure, subjective wellbeing, was 3 genome-wide significant SNPs that each explain 0.1% of the variance, identified by the SSGAC [1]. Due to a restricted number of SNPs, MR-Egger, MBE and MR-PRESSO could not be conducted. One unit increase of subjective wellbeing is equivalent to one standard deviation increase of the subjective wellbeing composite continuous scale. The genetic instrument was the 3 genome-wide significant SNPs for subjective wellbeing from Okbay et al [1]. Suitable proxies were identified at an LD cut-off R 2 >0.8. Arrows on confidence intervals indicate they extend beyond the axis. Phenotype scores for all measures were standardised apart from for blood pressure which is represented on a different scale. There was no clear evidence to suggest a causal effect of subjective wellbeing on any of the health outcomes (see Figure 2). MR Egger and IV weighted estimates are represented with blue lines. On the x-axis, βIV represents the effect size of each SNP. On the y-axis, 1/SEIV represents the inverse standard error for each SNP effect. Figure S3. Leave-one-out analysis: each row represents a two-sample MR analysis of BMI on subjective wellbeing using all of the genome-wide significant SNPs available from Locke et al. [2] except for the SNP listed on the y-axis. The point represents the effect size with that SNP removed and the line represents the standard error.
Leave-one-out analysis was conducted using MR Base [3] to identify if any individual SNPs were driving the association between BMI and wellbeing. Results are shown in Figure S2. The SNP with the largest contribution to the effect is rs1421085 located on chromosome 16 in the second intron of the FTO (fat mass and obesity associated) gene. FTO has been repeatedly associated with obesity in different populations [4]. However, the biological consequences of intronic FTO SNPs are still unknown. They are currently thought to play a regulatory role in FTO gene expression in the hypothalamus [5].
Although research is not completely certain of the role of FTO, its large effect size and robust association with obesity suggest that this gene has the largest effect in the twosample MR because of its BMI effect size rather than because of pleiotropic effects. Figure S4. Bias plots of association with baseline confounders for BMI, comparing observational and MR analyses Shungin [9] 2015 GIANT 32% Waist Circumference Shungin [9] 2015 GIANT 32% Body Fat Lu [10] 2016 N/A 9% Any items of scales capturing positive affect or life satisfaction (e.g. "During the past week, I was happy?" and "How satisfied are you with your life as a whole?" respectively). The two were pooled and equally weighted.
sex, age, age 2 and cohort specific covariates e.g. batch effects At least 4 PCs 20.00

Coronary Artery Disease
Criteria for defining cases is given separately for each cohort in the cohort descriptions supplementary note [6]. Of the CAD cases, ~70% had a reported history of MI.
-Given for each cohort in the cohort descriptions supplementary note [6].
Given for each cohort in the cohort descriptions supplementary note [6].

59.97
Total Cholesterol Where possible, individuals on lipid lowering medication were excluded. The majority of studies measured lipids after >8 hours of fasting. 24% of studies directly measured cholesterol and the rest estimated it using the Friedewald formula. Total cholesterol is calculated from HDL, LDL and triglycerides.  [11]. A -0.002 (-0.010, 0.006) 0.657 Note. The first release of genetic data from the UK Biobank (~150,000) was part of the SSGAC discovery GWAS [1] therefore genetic data from the second release (~350,000) was used in this independent analysis.  [1] and have been extracted from the most recent GWAS of Major Depressive Disorder (MDD) [12]. None is associated with MDD in the current GWAS at the genome-wide level of significance. A1 = effect allele, A2 = non-effect allele, EAF = effect allele frequency. rG between subjective wellbeing and MDD was -0.65 (SE = 0.04) [12].   Regression dilution bias occurs when the SNPs are weakly associated with the exposure. This is also known as the `NO Measurement Error' (NOME) assumption. The I 2 statistic for the SNP-exposure (GX) effects is a measure of NOME violation. In order to conduct MR Egger regression I 2 should be greater than 0.9 or else simulation extrapolation (SIMEX) correction should be applied [13].   Figure 3. Only the exposures BMI, waist circumference and HDL cholesterol had significant global tests and outlier tests.