CONSTRUCTS AND MODELS IN HEALTH PSYCHOLOGY - THE CASE OF PERSONALITY AND ILLNESS REPORTING IN DIABETES-MELLITUS

Objectives. Construct overlap among important variables in health psyc hology was studied. Competing theories related to the reporting of med ically unexplained symptoms (somatization) and illness severity were e xamined. These objectives were studied in the setting of insulin-depen dent diabetes mellitus (IDDM). Design. A correlational design was empl oyed. Sufficient numbers of participants were sought in order to lest theories of symptom reporting using structural equation modelling and other multivariate techniques. Methods. Two hundred and two people wit h IDDM attending a diabetes out-patient clinic acted as participants. Personality traits, alexithymia and illness-related coping factors wer e the key psychological constructs studied. Experience of medically un explained symptoms (somatization) and diabetes severity were measured. Results. Neuroticism, alexithymia and negative emotion coping were al l significantly related to somatization, and there was significant con struct overlap. Competing models of symptom reports-transactional vers us negative affectivity-were formulated and tested using structural eq uation modelling. However, the best fitting model was one that integra ted features of both. Constructs showed considerable overlap, with evi dence for a very general factor related to the reporting of negative a ffects and bodily disturbances. However, there is also evidence of uni que variance in constructs, especially that of negative emotion coping . Conclusions. Important constructs used in health psychology show sig nificant overlap, and this general source of variance warrants further study. Despite overlap, constructs contributed unique variance to hea lth outcomes. Transactional and negative affectivity models of symptom reporting may be reconciled in an integrated model. Generalization of the models tested here to other illness samples and healthy individua ls is recommended. Suggestions are made for further research to refine constructs in health psychology and to limit their present profusion.