van Hedel Karen, Martikainen Pekka, Moustgaard Heta, Myrskylä Mikko
Max Planck Institute for Demographic Research, Rostock, Germany.
Population Research Unit, University of Helsinki, Helsinki, Finland.
SSM Popul Health. 2018 Feb 2;4:244-253. doi: 10.1016/j.ssmph.2018.01.001. eCollection 2018 Apr.
Marriage is associated with better mental health. While research on the mental health of cohabiting individuals has increased in recent years, it has yielded mixed results thus far. We assessed whether the mental health of cohabiters is comparable to that of married individuals or those living alone using longitudinal data on psychotropic medication purchases. Panel data from an 11% random sample of the population residing in Finland for the years 1995 to 2007, with annual measurements of all covariates, were used. Ordinary least squares (OLS) models were applied to disentangle the relation between cohabitation and psychotropic medication purchases while controlling for relevant time-varying factors (age, education, economic activity, and number of children), and individual fixed effects (FE) models to further account for unobserved time-invariant individual factors. Our sample consisted of 63,077 men and 61,101 women aged 25 to 39 years in 1995. Descriptive results and the OLS model indicated that the likelihood of purchasing psychotropic medication was lowest for married individuals, higher for cohabiters, and highest for individuals living alone. This difference between cohabiting and married individuals disappeared after controlling for time-varying covariates (percent difference [% diff] for men: 0.3, 95% confidence interval [CI]: -0.0, 0.6; % diff for women: -0.2, 95% CI: -0.6, 0.2). Further controlling for unobserved confounders in the FE models did not change this non-significant difference between cohabiting and married individuals. The excess purchases of psychotropic medication among individuals living alone compared to those cohabiting decreased to 1.2 (95% CI: 1.0, 1.4) and 1.4 (95% CI: 1.1, 1.6) percentage-points in the fully-adjusted FE model for men and women, respectively. Similar results were found for all subcategories of psychotropic medication. In summary, these findings suggested that the mental health difference between cohabiting and married individuals, but not the difference between cohabiting individuals and those living alone, was largely due to selection.
婚姻与更好的心理健康状况相关。虽然近年来关于同居者心理健康的研究有所增加，但迄今为止结果不一。我们利用精神类药物购买的纵向数据，评估同居者的心理健康状况是否与已婚者或独居者相当。使用了1995年至2007年芬兰11%随机人口样本的面板数据，对所有协变量进行年度测量。应用普通最小二乘法（OLS）模型来厘清同居与精神类药物购买之间的关系，同时控制相关的随时间变化的因素（年龄、教育程度、经济活动和子女数量），并使用个体固定效应（FE）模型进一步考虑未观察到的随时间不变的个体因素。我们的样本包括1995年年龄在25至39岁之间的63077名男性和61101名女性。描述性结果和OLS模型表明，购买精神类药物的可能性已婚者最低，同居者较高，独居者最高。在控制了随时间变化的协变量后，同居者和已婚者之间的这种差异消失了（男性的百分比差异[% diff]：0.3，95%置信区间[CI]：-0.0，0.6；女性的% diff：-0.2，95% CI：-0.6，0.2）。在FE模型中进一步控制未观察到的混杂因素并没有改变同居者和已婚者之间这种不显著的差异。在完全调整的FE模型中，独居者相对于同居者精神类药物购买的额外增加在男性和女性中分别降至1.
