Skip to content
What Causes What
SMOKE+ INCREASES

Job strain → Depression

Higher Job strain increases Depression.

Madsen 2018's IPD-Work pool of 120,221 employees found a 27 percent higher rate of first hospital-treated depression after job strain (HR 1.27, 1.04 to 1.55). In Smith 2012, 3,735 Canadians free of depression who reported a rise in psychological demands had later depression odds of 2.36. Nobody assigned the job.

Meta-analysis · Low · N=120221

The association disappears when you account for other factors, like socioeconomic status or healthy-user bias.

Mechanism

Authors treat high demand with low control as a chronic stressor. They name HPA wear, broken sleep, isolation, and less leisure movement. The files score a questionnaire and later count a diagnosis or a PHQ-9. They do not isolate one path. Madsen, Nyberg, Magnusson Hanson and colleagues (2018) pooled six published cohorts (27,461 people, 914 incident clinical depressions) and 14 unpublished IPD-Work cohorts (120,221 employees, 982 first hospital-treated depressions). Job strain was high demand plus low control against every other combination. Published diagnostic interviews: OR 1.77 (1.47 to 2.13). IPD, age, sex, and cohabitation: HR 1.27 (1.04 to 1.55). A continuous symptom score killed the hospital link (HR 1.03). Burr, Müller, Rose and colleagues (2021) followed 2,212 German S-MGA employees, 2011/12 to 2017. COPSOQ items mapped to Karasek. PHQ-9 of 10 or more. Strain versus low strain: RR 1.51 (1.03 to 2.23). The joint effect was not larger than the sum (RERI -0.21). Kim, Shin, Oh and colleagues (2021) followed 63,959 Korean employees from a 2012 Kangbuk Samsung exam through 2017, 4,227 incident case-level depressive symptoms. The instrument is the Korean Occupational Stress Scale, not the Karasek quadrant. High job demand tracked new symptoms in men of every age and in women through midlife. Smith and Bielecky (2012) used 3,735 Canadians in the National Population Health Survey who were free of depression when job scores were taken. An abbreviated Job Content Questionnaire scored control, demands, support, and a demand-to-control ratio. A change counted only if it beat day-to-day noise. Depression two years later was CIDI-SFMD or a first medical diagnosis in the missed year. A rise in psychological demands: adjusted OR 2.36 (1.14 to 4.88). A rise in the strain ratio: 1.24 (0.57 to 2.68). Control and support changes did not. The authors say not to assume a demand-by-control interaction.

Caveats

The stamp stays smoke. Nobody randomized the job. These are cohorts and a meta-analysis of cohorts. Another cohort does not close the backdoor. More papers of the same design do not settle the claim. Baseline mood sits on the backdoor. People who already feel low report worse jobs and later meet a depression code. Madsen's hospital association vanished after a continuous symptom score. The same file also shows the reverse arrow: symptoms predict later strain. Madsen's IPD outcome is hospital-treated depression. Most episodes never reach a ward. The published OR 1.77 uses diagnostic interviews and may be inflated by publication bias. The IPD studies are Nordic plus Whitehall. Burr is German employees aged 31 to 60 on social insurance, not civil servants or the self-employed. Kim is one hospital screening program and a Korean stress battery, not Karasek's combination. Smith is a household survey, two demand items, and self-report. A change in the strain ratio itself did not move. The authors cannot tell a real job change from a change in how the job felt. Sample size: N from the papers. Burr found no extra risk from combining high demand and low control beyond adding the two. Madsen's continuous analysis found low control, not high demand, and no demand-by-control interaction. Passive jobs also sat higher in the IPD. Smith found the demand rise, not the control change. This page is not employment versus unemployment. Job loss is a different exposure. It is not long-term sickness absence. Jarroch 2024 asked whether strain modified the path from depression to absence; depression was not the outcome. It is not ovarian cancer. The 2017 work-characteristics file asked about ovarian risk and death. It is not job strain to cortisol. That pair already has its own page. This is a journal entry, not medical advice and not a reason to quit a job, to change a shift, or to treat a number.

Effect

qualitative. Madsen 2018 IPD-Work n=120221, HR 1.27 (1.04-1.55) hospital-treated. Published 6 cohorts n=27461, OR 1.77. Smith 2012 NPHS n=3735, demand rise OR 2.36. Observational. Design does not identify.

Nodes

Socioeconomic statusJob strainPsychology & behaviorDepression

Named confounders

Cite this page

What Causes What. “Job strain → Depression.” https://whatcauseswhat.org/edges/e-jobstrain-depression (atlas updated 2026-09-02).

https://whatcauseswhat.org/edges/e-jobstrain-depression

Papers