Abstract
BACKGROUND: Research shows that unemployment reduces access to health care and vaccines and increases financial difficulty, family conflict, and other sources of stress that are known to suppress immune function. In addition, seasonal unemployment rates parallel seasonal influenza activity. Following a theory that argues that macroeconomic conditions affect population health, this paper examines whether there is an association between monthly unemployment rates and influenza activity. METHODS: Data from influenza activity surveillance reports from the Centers for Disease Control and Prevention are combined with information from the Bureau of Labor Statistics on state-level unemployment rates in the U.S. for the flu seasons between 1999 and 2010. Pooled time-series cross-section logistic regression analyses are conducted to examine the effect of the unemployment rate on the likelihood of widespread and/or regional influenza activity in the 48 contiguous states throughout this period. A total of 3712 state-month observations are examined. RESULTS: Net of other factors included in the multivariate regression analysis, a one-percentage-point increment in the unemployment rate is associated with between a 7·1% and 37·0% increment in the odds of widespread influenza (Odds ratio = 1·21). Likewise, a one-percentage-point increment in the unemployment rate is associated with between a 17·1% and 44·7% increment in the odds of at least regional influenza (Odds ratio = 1·30). Results hold regardless of whether time-varying state-level characteristics are included. CONCLUSIONS: Higher state-level unemployment increases the likelihood of regional and widespread influenza activity.