Abstract:
Background: In many low-income countries, food insecurity is a pressing concern and thus ensuring food security for all segments of the population is a high priority. In a recent global analysis of 149 countries, the prevalence of any food insecurity ranged from 18.3% in the East Asian region to 76.1% in sub-Saharan Africa. In Ethiopia, both chronic and transitory (seasonal) food insecurity are persistent problems for a large segment of the population and in 2012, there was an 82.3% prevalence of household-level food insecurity in a zone in southern Ethiopia. Ending hunger and achieving food security is one of the United Nation’s Sustainable Developmental Goals (SDG) to be achieved by 2030. Mental illness is associated consistently with poverty, but its association with food insecurity in people with severe mental disorders (SMD; psychotic disorders and bipolar disorder) is not well understood. Primary care-based integrated mental health care supported by interventions at community and health system levels, has emerged as an important approach to address the large treatment gap for people with SMD. However, little is known about the impact of integrated mental health care on food insecurity status. Objectives The general objective of this study was to investigate the association between severe mental disorder and food insecurity in a rural Ethiopian district before and after improved access to mental health care. Methods Study Design: The study involved two designs: (1) Sub-study-1 was a cross-sectional community- based, comparative study which aimed to explore the association between SMD and food insecurity. Sub-study-2 was a cross-sectional, community-based study of factors associated with food insecurity and work impairment in people with SMD only. (2) Sub-study-3 was a before-after study which aimed to evaluate the impact of an integrated mental health care programme on food insecurity status in people with SMD when compared to the general population, over a period of 12 months. Participants: The participants of the study were residents in Sodo district, south Ethiopia. 1) persons with SMD identified by community key informants, referred to primary care, assessed by primary care health workers who had been trained to deliver frontline mental health care, and reassessed by a mental health specialist using a structured clinical interview (the Operational Criteria for Research interview guide) to confirm the diagnosis of psychosis or bipolar disorder, 2) respondents from households of persons with SMD, and 3) a comparison group of households which did not include a person with SMD. Assessments: At baseline (T1), when the mental health care programme was being implemented, and at a twelve month follow-up (T2) assessment, trained lay data collectors assessed food insecurity status using a validated version of the Household Food Insecurity Access Scale (HFIAS9). At T1 and T2, work impairment was assessed by trained psychiatric nurses using the Longitudinal Interval Follow-up Evaluation-Range of Impaired Functioning Tool. Disability was measured using the World Health Organisation Disability Assessment Schedule 2.0. A range of other demographic, socio-economic and psychosocial measures were also used. Statistical analysis: In sub-study-1, multiple logistic regression was conducted to test the hypothesis that the presence of a household member with SMD was associated with food insecurity in that household. Potential confounders identified a priori were included into the model. In substudy-2, multiple logistic regression was used to explore the factors associated with severe food insecurity and work-related impairment in persons with SMD. Variables included in the multivariable model were those anticipated to have associations with the outcome variables on the basis of existing literature. Potential effect modification by strata was explored using the MantelHaenszel test of homogeneity. For sub-study-3, a Poisson working model with sandwich estimators of the standard errors was used to estimate a risk ratio for change in food insecurity status in SMD cases and the comparison households between baseline and 12 months. Multiple linear regression was used to identify factors associated with change in food insecurity scores in the SMD group. To examine potential effect modification of disability between clinical attendance and food insecurity, an interaction term was included in the final multivariable model and a likelihood ratio test was used to investigate improvement in model fit. Path analysis was used to explore the possible mechanisms linking food insecurity and predictor variables. Results A total of 292 people with SMD, 292 respondents from households of people with SMD and 284 respondents from comparison households were included in the study at baseline. At follow-up, 239 people with SMD, 239 respondents from households of people with SMD and 273 respondents from comparison households were included in the final analysis. Participant Characteristics: Persons with SMD were more likely to be younger, had fewer children, to have attended formal education and be female, unemployed, unmarried and not the household head than respondents from comparison households without a person with SMD. Baseline (cross-sectional) studies: Sub-study-1: Severe household food insecurity was reported by 32.5% of people with SMD and 15.9% of respondents from comparison households: adjusted odds ratio 2.82 (95% confidence interval 1.62 to 4.91). Higher annual income was associated independently with lower odds of severe food insecurity. Sub-study-2: In the multivariable model in people with SMD, severe food insecurity was associated with poor social support, experience of negative discrimination, higher disability and lower household annual income, but not with symptom severity. Work impairment was associated significantly with symptom severity and disability, but not with discrimination. Follow-up study: (Sub-study-3): Improvement in food security status 12 months after engaging with care was observed in 43.5% of households of a person with SMD compared to 30.2% of control households (adjusted risk ratio 1.68; 95%CI 1.24, 2.26). The proportion of households in the “severe food insecurity” category declined from 71/237 (29.9%) at baseline to 37/237 (15.6%) at twelve months among the SMD group; whereas it declined from 37/273 (13.5%) at baseline to 26/273 (9.5%) at twelve months among the comparison group. In people with SMD, improvement in food security status was associated with being a household head at baseline assessment, lower baseline disability and physical impairment scores. In a path model, change in symptom severity was indirectly associated with follow-up food insecurity status via an impact on reducing work impairment and discrimination (p<0.001). Conclusions People with SMD living in a rural Ethiopian district experience relatively higher levels of severe food insecurity than the general population. Moreover, the findings from this study indicate that food insecurity and work impairment in people with SMD are not just a consequence of illness severity. Socioeconomic factors such as social support, discrimination and income emerged as important factors associated with food insecurity. Our study also indicates that improving access to mental health care can reduce food insecurity in households of people with SMD. Recommendations The inclusion and prioritization of people with SMD in food security programs and development opportunities, including income-generating opportunities and schemes via awareness-raising and tackling the stigma associated with mental illness should be ensured. Access to integrated mental health care should be expanded, with support for people with SMD to remain engaged in care to maximize the economic benefit. Provision of additional interventions to improve work functioning and tackle discrimination may further reduce food insecurity in this vulnerable group.