This cross sectional study uses data from the Health Survey for England to explore the relationship between poor mental health and low income.

The research team from Exeter University included 94,870 adults in the study (1997-2006) which used the 12-item General Health Questionnaire (GHQ-12) to assess happiness, symptoms of anxiety or depression and sleep disturbance in the previous 4 weeks.

Logistical regression was used to calculate the relative odds of mental health problems in men and women, within 10-year age groups and by age-specific household income. Analyses were adjusted for age, gender, education, occupation and social class, race/ethnicity, smoking and number of comorbidities in addition to self-reported mental health problems.

The outcomes measured were:

• Psychological distress (GHQ-12 score ≥4)
• Self-reported longstanding problems with mental illness, anxiety or depression
• Self-reported use of antidepressants, anxiolytics or antipsychotics

Here’s what they found:

• 2.7% of men and 3.4% of women said they had mental health problems
• 3% of men and 5.3% of women said they had taken medication for a mental health condition
• Psychological distress (high GHQ-12 score) was reported by 12.2% of men and 16.8% of women
• Self-reported mental illness, psychiatric medication use, and psychological distress peaked in the 45–54 age group in both men and women in the lowest income quintile; this peak was not evident in the other income groups
• After adjustment, the greatest increase in the odds of reporting mental illness in the lowest income group compared with the highest income group was seen in men aged 35–44 years (adjusted OR 16.8, 95% CI 9.2 to 30.6)
• Among women, the greatest increase in the odds of reporting mental illness in the lowest income group compared with the highest income group was seen in the 45–54 age group (OR 10.3, 95% CI 6.2 to 17.1).

The authors concluded:

An increased prevalence of psychological distress, common mental disorder diagnoses and treatment in midlife is not a universal phenomenon but is found only in those in low-income households. This implies the phenomenon is not inevitable but is potentially manageable or preventable.

This kind of cross sectional study does not prove that low income is the cause of mental illness. We will need to see further longitudinal studies to be able to clarify this relationship, as well as whether the results truly indicate a midlife peak in mental health problems, rather than a period or cohort effect.

Lang IA, Llewellyn DJ, Hubbard RE, Langa KM, Melzer D. Income and the midlife peak in common mental disorder prevalence. Psychol Med. 2011 Jul;41(7):1365-72. Epub 2010 Dec 10. [PubMed abstract]