Dissertation Title： "Gender and Health: The Influence of Psychosocial Factors on Health"
The medical literature presents many examples of differences in health care use by men and women. The roots of these differences include biological, social, and psychological factors. This dissertation thesis consists of three papers that investigate the dynamic confluences that affect health care use.
Paper 1 explores predictors of patient preferences for coronary artery bypass graft surgery (CABG), hypothesizing that men will report a greater preference for CABG. I find evidence that this is true, but age interacts with gender, such that older women are less likely to express a preference for CABG. Further, I find that preference for CABG predicts subsequent invasive heart surgery. This suggests that older women may arrive in the physician's office with a bias against CABG. Physicians may need to present information differently to older women if CABG is an appropriate treatment modality.
Paper 2 considers whether somatosensory amplification, sensitivity to minor somatic sensations, is associated with changes in health status or health services use over a ten-year period. Somatosensory amplification is thought to reflect a propensity to report, and is more common in women. I find evidence that women are more likely to amplify than men, but women's scores decrease with age. Respondents who develop a serious chronic illness report higher amplification scores over time, which may reflect more careful bodily monitoring. Although postmenopausal women may report unexplained medical symptoms to their clinicians, it does not appear that somatosensory amplification is a factor in these reports.
Paper 3 examines somatosensory amplification as a measure to improve Rose's angina and dyspnea score reports. These scales predict clinically observable ischemic heart disease and death in men, though not in women. I hypothesize that women amplify more, and thus report more cardiac symptoms, which influences the ability of these measures to predict mortality. Overall, controlling for amplification has little effect on the scores. When I stratify by gender and adjust for somatosensory amplification, both scores remain significant predictors of mortality for men, but not for women. This supports the observations that these scores do not calibrate well for women, perhaps due to reporting styles that women develop.