Fatigue in primary care: longitudinal associations with pain.

So far, most studies on the association between pain and fatigue have used cross-sectional data. We analysed the possibilities for a temporal relationship between pain and fatigue in a cohort study of patients presenting with fatigue in primary care. Of 856 recruited patients, 642 (75%) completed postal questionnaires after the consultation, and at one, four, eight and 12 months follow-up, with completion rates ranging from 82% to 88%. Pain was measured using the Short-Form health survey (SF-36) and fatigue using the severity scale of the Checklist Individual Strength (CIS). Longitudinal associations were analysed using generalised estimated equations (GEE). We used three different models assessing possible relationships between the symptoms in time, either in the same intervals or with a time-lag, suggesting either a synchronous or temporal association. The regression coefficients were strongest in the model assessing synchronous change, indicating that a one-point improvement in pain was associated with a 0.25 improvement in fatigue in the same time interval (adjusted for potential confounding). Baseline duration of fatigue and expectations of its future course significantly modified the association in this model, with stronger associations between changes in pain and fatigue found in patients with a shorter duration of fatigue or more positive expectations. The models using a time lag showed a significant but inverse association between changes in pain and subsequent changes in fatigue. The results indicate that changes in pain and fatigue are directly related in time, rather than showing temporal associations. Copyright (c) 2010 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.