Disability pension due to common mental disorders : subsequent psychiatric morbidity and suicidal behaviour

Abstract

Background: Mental diagnoses have become the most common disability pension (DP) diagnoses in OECD countries. Up to half of DP due to mental diagnoses are due to common mental disorders (CMDs). So far, research has mainly focused on risk factors of DP rather than on the future life situation of people with DP. Nevertheless, the latter types of studies suggest higher mortality, including suicide among DP recipients. Therefore, knowledge on morbidity and suicidal behaviour following DP is needed. Investigations of the associations between DP and adverse mental health outcomes should be multifaceted, considering socio-demographics, co-morbidity, existing regulations about DP benefit, medical factors, i.e., healthcare use and prescribed medication, etc. The thesis aimed to gain knowledge about the association of DP due to CMD with subsequent psychiatric morbidity and suicidal behaviour in the general population of Sweden.

Methods: In study I and II all individuals who were on DP due to CMD throughout 2005 were included. Cox regression analyses calculating hazard ratios (HRs) were conducted for measuring the associations between socio-demographics, specialized healthcare use, medication, and measures of DP with suicidal behaviour during a 5-year follow-up. In study II, all analyses were stratified by sex and age. In study III, Generalized Estimating Equations (GEE) were applied to identify the trajectories of specialized healthcare use before and after granted DP during a 7-year period in cohorts with DP due to CMD, granted either in 2005-06 or in 2009-10. The latter cohort also comprised the study population for study IV. In study IV, a group-based trajectory method was used to identify different trajectory groups according to the amount of defined daily doses of prescribed antidepressants (AD) during 6 years of observation. Information on socio-demographics, DP, specialized healthcare use, medication, and death was derived from five nationwide registers. Data were linked at individual level by unique identification numbers.

Results: In the disability pensioners, sex, age, education level, family situation, previous in- or specialized outpatient care due to mental diagnoses or suicide attempt, and medication were strongly associated with subsequent suicidal behaviour (HR range: 1.15-3.89) (study I). Among the DP measures, stress-related mental disorders as main DP diagnosis, compared to depressive disorders were associated with a lower risk for subsequent suicidal behaviour (HR range: 0.4-0.7). Among other measures, psychiatric comorbidity, and full-time DP were associated with a higher risk of subsequent suicidal behaviour (HR range: 1.3-3.3) compared to no comorbidity and part-time DP, respectively (study II). Moreover, healthcare use due to mental diagnoses increased until the year preceding DP and declined thereafter (study III). After the introduction of stricter DP granting criteria, people had higher levels of such healthcare preceding DP, nevertheless, still less than half of the individuals received specialized mental healthcare a year prior to granted DP. Analyses also showed a steeper decline in healthcare use due to mental diagnoses immediately following DP among individuals granted DP with stricter criteria (study III). Among the disability pensioners, five trajectory groups were identified according to the annual amount of prescribed ADs. The groups differed particularly regarding age, main DP diagnosis, and previous healthcare use. Moreover, many had a very low dose of AD before granted DP, and for most of them, there was hardly any decline in the AD trajectories following granted DP (study IV).

Conclusion: Important aspects to consider for subsequent psychiatric morbidity and suicidal behaviour in individuals after granted DP due to CMD should include, socio-demographics, the main and secondary DP diagnoses, grade of DP, and diagnosis specific previous specialized healthcare use. The frequency of healthcare use was lower immediately following DP, but the amount of prescribed ADs did not alter for most of the disability pensioners in that period. Both DP granting and subsequent mental health appears to be associated with the DP granting regulations. Importantly, the low frequency of specialized mental healthcare use and amount of prescribed ADs during the pre-DP period may indicate a sub-optimal treatment before the premature labour market exit among individuals granted DP due to CMD.