Case and Paxton discuss a phenomena where women are less likely to die at an earlier age than men even though they have both worse self-rated health and higher rates of hospitalization. They examine fourteen years of data from the U.S. National Health Interview Survey. As an explanation for the phenomena of self-rated health, they find the outcomes to be related to the various chronic health conditions faced by men and women. Men experience higher rates of hospitalization and mortality in relation to smoking related conditions (for example heart disease and lung disease). Case and Paxton mention that along with various chronic health conditions, the negative effects of these conditions can also be attributed to the explanation of higher rates of hospitalization and mortality. Unlike previous research, this analysis includes current and larger datasets.
Some explanations for the sex differences in self-rated health include certain conditions that are a result of various factors (i.e. biological, social, environmental, etc.) and the mortality outcomes of various conditions at a higher rate than others(i.e. migraines compared to heart disease). It is also considered that women tend to report worse health on surveys compared to actual health status. There is also a question of how health information is communicated to men and women and how responses differ as a result. Previous research has also questioned whether the significance of poorer self-rated health and rates of mortality are accurate. The idea that women have poorer self-rated health is predicated on the concept of women having worse health than men on every measure of health. Unlike previous research, Case and Paxson look at the differences in hospital stays which compare men and women due to the involvement of a physician which indicates a level of seriousness of presenting health condition. In previous research when adjustments were made for sociodemographic and health indicators, the relationship between self-rated health and mortality was in fact greater for men than women. Going one step further, when more health indicators were added to previous analysis, the gap between men and women decreased.
The data used in the analyses included fourteen years of data (1986-1994, 1997-2001) along with the Multiple Cause of Death file which provided the needed variables related to those individuals who were in the initial interviews from 1986-1994. The 1986-1994 sample included 237,140 men and women while the 1997-2001 sample included 146,966 men and women. The data was restricted to non-pregnant women, those who did not have a child one year or younger in the home, and non-pregnancy related hospital stays. The related mortality data was restricted to those with known vital status and aged 45-84. Regarding sex differences and the relationship to self-rated health and mortality, the difference lessens dramatically between the ages of 60-65 for women.
The analyses found that out of the six conditions that were related to poor self-rated health, stomach cancer was the largest effect related to poor self-rated health for both men and women. This was followed by respiratory cancer and diabetes, depression and CVD, and other pain. Skin cancer was found to not be significantly associated with poor health for men and women. There was found to be a significant difference between men and women in relation to the prevalence of gender reported condition. When looking at hospital stays, men and women who reported worse self-reported health also had an increased number of hospital stays on average. When looking at men and women who had the same health condition, men were found to have an increased number of hospital stays. The size of effect on mortality was similar to the size of effect on self-rated health for men and women.