Job Loss Has Same MI Risk as Smoking/Diabetes

November 21, 2012

DURHAM, North Carolina — Loss of employment carries a similar risk of MI as hypertension, diabetes, or smoking, a new study suggests [1]. In addition, there appears to be a clear dose-related effect, with the risk of MI increasing with each additional job loss.

The study, published online on November 19, 2012 in the Archives of Internal Medicine, was conducted by Dr Matthew Dupre (Duke University, Durham, NC).

Dupre told heartwire that this study was more detailed than previous research on unemployment and heart disease risk, and to his knowledge, it is the first to examine the cumulative effect of multiple dimensions of unemployment on the risk of MI.

Interestingly, the effect of unemployment on MI risk was most marked in the first year after the job loss occurred, suggesting that the risk is tied to the actual event, rather than long-term factors.

Mediated by Stress

Dupre commented to heartwire : "It seems that the transition itself is the most dangerous. We think the MI risk is probably caused by the stress of becoming unemployed. But many other factors may come into play, such as changes in diet and sleep, increased smoking, and loss of control of other risk factors, such as diabetes and hypertension, as people are thrown into a chaotic state. We would therefore urge physicians to be more vigilant in terms of health awareness in patients who have recently become unemployed."

The researchers analyzed data from the Health and Retirement Study (HRS), a nationally representative sample of the older US adult population (over 50 years). The sample for the current study included 13 451 people aged 51 to 75 years at baseline who reported ever having worked. They underwent interviews every two years, which included detailed questions about employment history, as well as health and socioeconomic information, and were followed prospectively over an 18-year period. Patients who had had an MI before baseline were excluded.

Results showed that there were 1061 AMI events (7.9%) during the follow-up period. Among the participants, 14% were unemployed at baseline, 70% had had one or more job losses, and 35% had spent time unemployed.

Multivariate models showed that MI risk was significantly higher among the unemployed and that risk increased with each job loss.

Risk of MI According to Unemployment Status and Each Incremental Job Loss

  HR (95% CI)
Unemployment status 1.35 (1.10–1.66)
One job loss 1.22 (1.04–1.42)
Two job losses 1.27 (1.05–1.54)
Three job losses 1.52 (1.22–1.90)
Four job losses 1.63 (1.29–2.07)


Although the risk of AMI was greatest in the first year after job loss, unemployment status, cumulative number of job losses, and cumulative time unemployed were each independently associated with increased risk of MI, even after adjusting for sociodemographic, socioeconomic, behavioral, psychological, and clinical risk factors.

The type of work had no effect on the findings, and voluntary loss of employment, such as retirement, was not associated with increased MI risk.

Exact Mechanism Unclear

Noting that the exact mechanism behind the link is not clear--although it does appear to be stress related--the researchers suggest that future studies should consider whether other job-related factors, such as seasonal employment, underemployment, multiple jobs, or family demands, may be sources of employment instability, stress, and increased cardiovascular events.

"As rates of job instability continue to increase and unemployment reaches 30-year highs . . . the cardiovascular costs of repeated job losses in younger cohorts are yet unknown," they add.

In an accompanying editorial [2], Dr William Gallo (City University of New York, NY), points out that the stressors involved in job loss "are too many and too entangled to enumerate and describe," which makes identifying the mechanisms linking loss of employment to health problems extremely difficult.

He concludes that: "The next generation of studies should identify reasonable pathways from job separation to illness so that nonoccupational interventions may be developed and targeted to the most vulnerable individuals."

The authors report no conflicts of interest.