Large socioeconomic differences exist in disease and mortality. This paper describes the distribution of specific medical reasons for sickness absence by grade of employment in the Whitehall II study and validates the medical reason by comparison with general practitioners' records. Analysis of sickness absence data on 5620 male and female civil servants aged 35-55 years. Data have been collected from 12 of the 20 London based civil service departments participating in the Whitehall II study, where medical reason for absence was available. Rates and distributions of reasons for absence for short spells (< or = 7 days) and long spells (> 7 days) were analysed. Respiratory disorders and gastroenteritis accounted for over half of all spells of absence, with headache and migraine, musculoskeletal disorders, injury, and neurosis accounting for a further 20%-30% of absences. There was an inverse association with employment grade, the lower the grade the higher the rate of absence for both short spells (< or = 7 days) and long spells (> 7 days). In general, women had higher rates of absence than men. Comparison of reason for very long spells of absence (> 21 days) showed moderate agreement between civil service and general practitioner. There is a lack of national comprehensive data on sickness absence and medical reason for absence, in particular for women and for spells of different duration. Data from the Whitehall II study show large employment grade and sex differences in the distribution of medical reasons for absence that are similar to socioeconomic differences in morbidity documented in other studies. Possible explanations include the subjective nature of illness and disease; the work/family interface; and the influence of the absence culture. Longer term follow up will provide information on whether sickness absence relates to serious morbidity and mortality.
