How income changes affect mental health for working-age adults: a systematic review and meta-analysis

• Published in: The Lancet (British edition) 2021-11, Vol.398, p.S4-S4
• Main Authors: Thomson, Rachel M, Igelström, Erik, Purba, Amrit Kaur, Shimonovich, Michal, Pearce, Anna, Leyland, Alastair, Thomson, Hilary, Katikireddi, S Vittal
• Format: Article
• Language: English
• Subjects: Adults; Bias; Health services; Heterogeneity; Income; Medical research; Mental health; Meta-analysis; Natural disasters; Personal income; Policies; Poverty; Psychological stress; Quality of life; Risk analysis; Systematic review; Unemployment; Well being
• ISSN: 0140-6736; 1474-547X
• DOI: 10.1016/S0140-6736(21)02547-2

Income is thought to be a major determinant of mental health, but we believe this literature has not been quantitatively synthesised in its entirety. We systematically reviewed studies assessing the effect of income changes on mental health in working-age adults (16–64 years). We searched seven databases (MEDLINE, Embase, Web of Science, PsycINFO, ASSIA, EconLit, and RePEc) for randomised controlled trials and quantitative non-randomised studies with no date limits on Feb 5, 2020 (PROSPERO CRD42020168379). We included English-language studies measuring effects of individual or household income change on any mental health or wellbeing outcome (search terms included mental health, depression, anxiety, wellbeing, quality of life, life satisfaction, psychological distress AND income [or synonym] NEAR change [or synonym]). Risk of bias was assessed using Cochrane methods. We did three-level, random-effects, meta-analyses and explored potential effect modification using meta-regression and stratified analyses. Critical risk of bias studies were excluded from primary analyses. Of 16 521 hits screened, 136 studies (12·5% randomised controlled trials) were narratively synthesised and 86 meta-analysed. Risk of bias was high, with 41 studies (30·1%) rated critical and 64 (47·1%) serious or high. The source of income change was generally not known (n=63, 46·3% of studies), with welfare policies (n=38, 28·0%), natural disasters (n=10, 7·4%), and lottery wins (n=7, 5·1%) being the most commonly described sources. A binary income increase was associated with 0·08 SD improvement in mental health (95% CI 0·038 to 0·13; n=258 637) and an income decrease with 0·21 SD worsening (–0·30 to –0·13; n=227 804). For continuous log(income) exposures, a 10% income increase was associated with 0·003 SD improvement in mental health (0·000 to 0·005; n=1 510 221). For wellbeing, binary increases were associated with 0·27 SD improvement (0·14 to 0·41; n=163 969) and a 10% income increase with 0·003 SD improvement (0·002 to 0·005; n=105 326). Heterogeneity was high (I2=79–87%). Effect sizes were larger in low-income and middle-income settings, for people of lower socioeconomic position, and in studies with higher risk of bias. Effects are potentially larger for wellbeing outcomes, for income losses, and in the most socioeconomically disadvantaged. To support mental health, welfare policies need to reach the most disadvantaged, and consider wider factors such as financial insecurity and emp