Abstract
In this study, we aimed to investigate the causal effect of smoking on social isolation among older adults in England. Data from older adults of European ancestry who participated in 1 or more waves of the English Longitudinal Study of Ageing, from wave 1 (2002/2003) to wave 9 (2018/2019), were analyzed (n = 43,687 observations from 7,008 individuals; mean age = 68.50 years). The effect of current smoking on social isolation (ranging from 0 to 5) was estimated by 2-stage least squares regression using a polygenic score (PGS) for smoking cessation as the instrument. A low PGS for smoking cessation predicted current smoking (per 1-standard-deviation lower PGS, coefficient = 0.023, 95% confidence interval (CI): 0.015, 0.030; F = 36.420). The second-stage regression showed that current smoking increased social isolation by 1.205 points (95% CI: 0.308, 2.101). The association was larger for persons with higher socioeconomic backgrounds: 2.501 (95% CI: -0.024, 5.026) and 0.696 (95% CI: -0.294, 1.686) for those with higher and lower educational levels, respectively. This study showed that current smoking instrumented by a PGS for smoking cessation was associated with social isolation. Assuming that the PGS served as a valid instrument in this study, the findings support an effect of smoking on social isolation.