Table 3.

Effect of job loss from a business closure on BMI with and without mPGS and vPGS interactions for workers aged 50–70 in the Health and Retirement Study (HRS).

	(1)	(2)	(3)	(4)	(5)	(6)
	Full sample	European ancestry sample	Genotyped European ancestry sample
			No PGS	With mPGS	With mPGS interaction	With mPGS & vPGS interaction
	Beta (SE)	Beta (SE)	Beta (SE)	Beta (SE)	Beta (SE)	Beta (SE)
	95% CI	95% CI	95% CI	95% CI	95% CI	95% CI
	p value	p value	p value	p value	p value	p value
Business closure (BC)	0.137 (0.088)	0.051 (0.103)	− 0.099 (0.105)	− 0.091 (0.104)	− 0.091 (0.103)	− 0.095 (0.102)
	[− 0.036, 0.310]	[− 0.151, 0.252]	[− 0.305, 0.106]	[− 0.294, 0.113]	[− 0.294, 0.112]	[− 0.296, 0.106]
	0.121	0.623	0.344	0.383	0.379	0.352
mPGS				0.175* (0.051)	0.170* (0.048)	0.163* (0.047)
				[0.075, 0.275]	[0.076, 0.263]	[0.071, 0.255]
				0.001	0.001	0.001
Business closure × mPGS					0.011 (0.091)	0.013 (0.090)
					[− 0.168, 0.189]	[− 0.164, 0.191]
					0.907	0.882
vPGS						− 0.056 (0.070)
						[− 0.193, 0.082]
						0.426
Business closure $\times$ vPGS						0.258* (0.092)
						[0.077, 0.439]
						0.001
BMI (t − 2)	0.918 (0.014)	0.929 (0.017)	0.932 (0.024)	0.921 (0.026)	0.921 (0.026)	0.923 (0.025)
	[0.890, 0.946]	[0.896, 0.962]	[0.884, 0.979]	[0.871, 0.971]	[0.871, 0.971]	[0.874, 0.973]
	< 0.0001	< 0.0001	< 0.0001	< 0.0001	< 0.0001	< 0.0001
R-squared	0.82	0.832	0.868	0.869	0.869	0.870
Observations	27,593	18,795	11,941	11,934	11,934	11,934
Treated observations	895	579	376	375	375	375
Control observations	26,698	18,216	11,565	11,559	11,559	11,559

*FDR corrected p value < 0.05. Abbreviations: SE, standard error; CI, confidence interval. Robust standard errors in parentheses. Each column in the table shows separate propensity score matched models for workers aged 50–70. All specifications adjust for BMI in the previous wave, or BMI (t − 2), and for the conditioning variables used in the propensity score matching that are reported in Table 1 and defined in detail in Supplementary Table S1. Columns 1–3 do not include mPGS or vPGS in the matching procedure. Column 1 includes all workers, regardless of ancestral background or the availability of genotype data and includes additional controls for race and Hispanic ethnicity in the matching procedure. Column 2 reports results for all European ancestry workers, regardless of the availability of genotype data. Column 3 reports results for the European ancestry sample with genotype data prior to matching on the mPGS and vPGS. Columns 4–5 include the mPGS in the matching procedure, and Column 6 includes the mPGS and the vPGS in the matching procedure. Individuals in the control group can have multiple observations. In the analytic sample, unique N(control) = 3564; unique N(treated) = 375.