The rationale and use of ridit analysis in epidemiologic studies of blood pressure

Alameda County redux: replication in Michigan

This article describes an attempt to reproduce, in Michigan, cross-sectional findings concerning the relationship between health habits and health status previously obtained in Alameda County, California by Belloc and Breslow. Data for this study were gathered by a telephone interview of a state-wide sample of 3,259 adult Michigan residents. The ridit (relative to an identified distribution) analyses presented include a comparison of findings from the Michigan and Alameda County studies. In Michigan, as in Alameda County, health status was found to be associated with various health practices, both individually and in combination. Consistent relationships were found between physical health status and individual health practices regarding hours of sleep, eating breakfast, eating between meals, cigarette smoking, weight for height, and physical activity. Physical health status was also linked to the overall number of health practices individuals engaged in. When the confidence interval for each ridit value was taken into account, however, only some of the findings proved statistically reliable.

Physical health status as a consequence of health practices

The relationship between health practices and physical health status was examined in a sample of 542 adults. A sub-sample of these individuals were also studied prospectively. Strong relationships were found in the cross-sectional data. "Good" health practices were correlated with higher health status. The analysis of the prospective data collected 18 months after the first interview also revealed a strong relationship between previous health practices and subsequent physical health status. The relationship still obtains when previous health status is controlled.