Long term sickness absence and diabetes: a Danish register-based longitudinal study with up to 17 years of followup

Publikation: KonferencebidragKonferenceabstrakt til konferenceForskningfagfællebedømt

Background and aims: The number of individuals with diabetes mellitus within the working age range is expected to rise. Type 1 and type 2 diabetes mellitus (diabetes) can have work related consequences such as decreased productivity, increased risks of sickness absence and early retirement. Sickness absence has been used as a central indicator of work disability, but has mainly been examined by short term spells of sickness absence in studies with limited follow-up time. This study aimed to identify the risk of long term sickness absence (LTSA) in individuals with diabetes within the first year of diagnosis and in subsequent years. Materials and methods: The study was based on registers with up to 17 years of follow-up. In a working population (n=102.746) individuals with the diagnoses of type 1 or type 2 diabetes (n=3.325: women, n=1.987, men=1338) and individuals without diabetes (n=99.421: women, n=73.332, men=26.089) were identified by Danish national registries of diagnosis (ICD-10 codes: E10.0-10.9, E11.0-E11.9, E12-E14) and prescribed medicine (ATC codes: A10A, A10B, A10BA02), in the period 1994 to 2011. We estimated the hazard ratios of transitions from work to LTSA (>3 consecutive weeks) in Cox proportional multi-state models with separate estimates for the risk within the first and subsequent years after diagnosis. Analyses were stratified by gender and controlled for immigrant status, highest attained education, marital status, calendar year, and job type. Results were reported in terms of incidence rates (LTSA events per 1.000 person years: PY), hazard rates (HR) with a 95% confidence interval (CI) and relative risks for LTSA in subsequent years compared to the first year after diagnosis. Results: Within the first year of diagnosis, the incidence rates for persons with diabetes were 126 in women (138 events/ 1.091 PY) and 100 in men (78 events / 780 PY) with significantly elevated risks of LTSA (HR=1.60, CI: 1.35-1.91 and HR=1.57, CI: 1.24-1.98). In the subsequent years after diagnosis the incidence rates for persons with diabetes were 137 in women (784 events / 5.718 PY) and 132 in men (506 events / 3.821 PY) with significantly elevated risks of LTSA (HR=1.49, CI: 1.36-1.63 in women and HR=1.82 CI: 1.60-2.06 in men). The relative risks of LTSA in subsequent years were 0.93 in women and 1.16 in men. Conclusions: Women and men with diabetes had elevated risk of LTSA, compared to those without diabetes. After the first year of diagnoses the risk decreased slightly in women and increased slightly in men. Studies of other chronic diseases have found substantial decrease in risk of LTSA after the first year of diagnosis – presumably due to a stabilization of disease with treatment. The absence of such an effect in diabetes may be due to diabetes-related comorbidities or poor treatment adherence. Future studies need to consider the impact of diabetes type, co-morbidities, and treatment adherence on LTSA.
StatusUdgivet - 2017

ID: 303678661