Excess salmonellosis in women in the United States: 1968-2000.

We describe recent epidemiological changes in salmonellosis. Linking 1968-2000 National Salmonella Surveillance System to census data, we calculated population-based age- and sex-stratified rates of non-urinary salmonellosis for the top 30 non-typhoidal serotypes. Using 1996-1997, 1998-1999, and 2000-2001 population-based FoodNet surveys, we compared reported diarrhoea, medical visits, and stool cultures. Despite an overall female-to-male incidence rate ratio (FMRR) of 0.99, the sex-specific burden of salmonellosis varied by age (<5 years FMRR 0.92; 5-19 years 0.85; 20-39 years 1.09; 40-59 years 1.23, and 60 years 1.08) and serotype (FMRR range 0.87 for Mississippi to 1.25 for Senftenberg). Serotype-specific FMRRs and median age (range 2 years for Derby to 29 years for Senftenberg) were related (correlation 0.76, P<0.0001). Recently, the relative burden of salmonellosis in women has increased. FoodNet data suggest that this change is real rather than due to differential reporting. Excess salmonellosis in women may reflect differences in exposure or biological susceptibility.