Associations of non-employment with common mental disorder subcomponents among working age population: Analysis of national data from 1993, 2000, 2007 and 2014

Background: Associations between employment status and mental health are well-recognised and such associations may have multiple modifying factors which may also contribute to variations in results. Aims: We aimed to investigate associations between non-employment and CMD subcomponents and the extent of their variation across age groups using nationally representative data in Britain. Method: We used a series of national mental health surveys of adults living in private households: the British National Surveys of Psychiatric Morbidity of 1993, 2000, 2007 and 2014. Employment status was the primary exposure of interest. Presence or absence of each fourteen symptoms of common mental disorder (CMD), as the primary outcome, was ascertained identically in all surveys from the revised Clinical Interview Schedule (CIS-R). Odds ratio for the association between exposure and outcome and population attributional fractions (PAFs) for each association was calculated. Results: Within the highest-risk 45 to 54 years age group, all odds ratios were statistically significant and strongest associations were observed with panic symptoms (OR = 2.33), followed by depressive symptoms (1.90), worry about physical health (1.84), depression (1.82), forgetfulness (1.82) and somatic symptoms (1.70). In the 55 to 64 years age group, highest population attributable fractions were observed for non-employment as a hypothetical risk factor for panic symptoms (51.7%), phobias (44.2%), forgetfulness (39.5%), depressive symptoms (38.5%), worries about physical health (37.9%) and somatic symptoms (36.0%). Conclusions: The particularly high impact in middle-aged, pre-retirement groups of non-employment on CMD suggests a policy focus on alleviating stressors and providing support for those made redundant and/or compelled to take unwanted early retirement.