Background: The validity of exposure data collected from proxy respondents of myocardial infarction patients has scarcely been studied. We assessed the level of disagreement between myocardial infarction patients and their spouses with respect to the reporting of the patient´s cardiovascular risk exposures.
Methods: Within the frame of the Stockholm Heart Epidemiology Program (SHEEP), a case-control study of risk factors of myocardial infarction performed in Stockholm county 1992-1994, a subset of 327 first time myocardial infarction cases aged 45-70 who survived >28 days after the event and who co-habited with a spouse or common-law spouse (proxy) were identified between 1993-04-05 and 1993-12-31. Among these, 243 cases participated along with their respective proxy in the present study. Control individuals, matched to cases by age, sex and residential area were also included (n = 243). Data were collected using questionnaires. Using conditional logistic regression we calculated for each of 82 exposures the odds ratio based on information collected from 1) myocardial infarction cases and controls [odds ratio A] and 2) proxies and the same set of controls [odds ratio B]. Disagreement was measured by calculating the ratio between odds ratio B and odds ratio A with 95% confidence intervals (CI) calculated using resampling bootstrap.
Results: For the vast majority of the exposures considered including diet, smoking, education, work-related stress, and family history of CVD, there was no statistically significant disagreement between myocardial infarction patients and proxies (n = 243 pairs). However, leisure time physical inactivity (proxy bias = 1.59, 95% CI 1.05-3.57) was overestimated by spouses compared to myocardial infarction patients. A few other exposures including some sleep-related problems and work-related issues also showed disagreement.
Conclusions: Myocardial infarction patients and their spouses similarly reported data on a wide range of exposures including the majority of the traditional cardiovascular risk factors, leisure time physical inactivity being an exception.