As longevity has increased over the last decades, the prevalence of chronic conditions, comorbidities, and disabilities has increased as well. Measures related to an individual's burden of multiple diseases and disabilities have relied on relatively simple measures, often using simple summary scores (e.g., disease counts) as descriptive and predictive measures. Such measures misrepresent the complexity of comorbidity by ignoring information related to the temporal order of combinations of conditions, and specific time intervals between diagnoses and disabilities. We propose to use existing longitudinal data from an unlike-sex twin study to analyze sequential patterns of the onset of diseases and disabilities within men and women, with a focus on gender differences. Here, we implement three specific strategies that will result in richer descriptive and predictive measures of disease and disability. By including information about temporal patterns of disease history, these measures will allow us to test hypotheses related to both within and between gender effects. The primary descriptive measures will make use of pattern variables, including numeric or alphanumeric strings. These techniques have been successfully used in computer applications related to subgroup creation and sorting, and to describe patterns of missing data. We will apply these methods to the description and analysis of the sequential patterns of comorbidity, including interval analysis, and assessment of disabilities. Person-level clustering methods will be used to identify subgroups of individuals who have specific constellations of comorbidities and disabilities. We will use the pattern measures and subgroup identifications to examine the relationship between disease and disability patterns within sibling pairs, and to test hypotheses regarding gender differences. Event history strategies will be used to predict mortality from disease and disability patterns. This study will provide important new information about gender differences relevant to aging and disability processes. Moreover, identifying patterns of co-morbidities, with the added temporal information on intervals between diagnoses has significant public health implications. Such information can contribute much to the development of timely gender-specific health promotion and disease prevention programs that could greatly reduce the risk of individuals with specific diseases from developing commonly occurring sequalae diseases and disabilities. ? ? ?
