This was the first study of attitudes toward mental illness to be conducted in Malawi. An over whelming majority of participants recruited from mental health and non-mental health clinics at a large general hospital attributed mental illness to substance misuse and spiritual causes such as spirit possession and God’s punishment [10
]. Our findings are broadly consistent with those from a large and robust community survey using the same rating scale conducted in Nigeria. That work found illicit drugs and alcohol (80.8%), spirit possession (30.2%), psychological trauma (29.9%) and genetic explanations (26.5%) to be the most common attributions for mental illness [10
]. That such similar results should be found is perhaps surprising considering the different study designs and samples. The Nigerian study involved multistage cluster sampling of households across three separate states whilst our work focused on consecutive attendees at hospital clinics. Though more work is required across Sub-Saharan Africa, these findings would seem to suggest that views on causation of mental illness maybe common across the region.
The fact that most of our sample attributed mental illness most strongly to drugs and alcohol might be considered a positive finding on one level, as it suggests potential treatability. However only a few mental disorders are known to be aetiologically attributable to alcohol or illicit drug use, therefore the majority view found in our sample is not factually correct. In most Sub-Saharan African societies alcohol and illicit drug use are viewed negatively and are considered due to moral failings on the part of the user. This may also be why many participants attributed mental illness to spiritual processes.
More of our sample attributed mental illness to Gods punishment compared to participants in West Africa (21.9% in Malawi compared to 9.3% in Nigeria). In Nigeria, religious-magical views of causation have been found to be more associated with negative and stigmatizing attitudes to the mentally ill compared with biological explanations [20
]. Spiritual explanations have also been found for mental states due to physical illness such as delirium in this region [21
]. These beliefs may also explain why many cases of mental illness in Sub-Saharan Africa are treated punitively or outside of the Western Health care systems, for example, via traditional or faith healers. Our findings regarding drugs and alcohol and spiritual matters as the most popular causes of mental illness are therefore a concern, as they reflect a potential for discrimination and non-medical treatment or at its worst, maltreatment.
One notable difference in our results from other work in West Africa is the more frequent attribution of mental illness to brain disease (92.8% in Malawi versus 9.2% in Nigeria). Once again the differences observed between Nigeria and Malawi may reflect the different populations sampled. Whilst we are not aware of any direct health promotion strategies regarding mental health and mental illness in the Malawian clinics, it seems reasonable that a population attending these services will be more attuned to a medical model of causation. The strong attribution of mental illness to brain disorders does however appear contradictory considered alongside the equally strong spiritual attributions given by participants for mental illness (such as spirit possession and mental illness being a punishment from god). While in Western traditions, the mind and body are traditionally considered as distinct entities, this may not be true in Malawi. It is possible that spiritual possession is believed to influence the brain directly. Further qualitative research is needed to better understand the culturally specific inter relationships between these explanations for mental illness.
Malawi is currently ranked 153 of 169 on the latest UN Development Index and Category E (very high mortality) on the WHO mortality register. It is therefore perhaps surprising that only around half of our sample (43.3%) endorsed poverty as a cause of mental illness. Furthermore those defined as existing in poverty did not more readily attribute mental illness to poverty than their more affluent counterparts. The reasons for this may be two fold. If most of a population exists in relative poverty, then it may be hard for those within the society to consider poverty as a contributing factor for an illness which only affects a minority. In developed economies, though a diagnosis of mental illness can carry stigma, it can also entail sympathy (from some quarters), treatment from established health services and support from the welfare state. These factors are far from the established norm in Malawi where destitution may await those with mental illness. It is possible that those in a state of poverty are reluctant to consider that they may be at risk of an even worse fate.
Only a quarter of our respondents believed mental illness could be treated outside of the hospital setting. On one level this is to be expected as around half our sample consisted of patients and their carers attending established mental health clinics at a major teaching hospital in an urban centre. The population in our sample may therefore have been self selected to have a more positive or biased view towards hospital treatment. However our finding may reflect the reality that there is very little community mental health care in Malawi. Respondents may simply not have been aware of any alternative to treatment in a hospital. They may also have had concerns about the reality of treatment in a less specialised centre.
With regards to social distance and mental illness, very few of our respondents would have been ashamed if someone in their family experienced mental illness and most were prepared to maintain a friendship with someone who had been mentally ill. However, our respondents seemed to be less prepared to consent to increasing social intimacy with someone who had experienced mental illness. Less than half were prepared to share a room with someone who had experienced mental illness and only approximately one in five was prepared to consider marriage. Since genetic factors were believed by half of our participants to be a cause of mental illness, fears about mental illness being passed on to future offspring may have influenced these findings.
Our study found less stigmatising beliefs in terms of social distance compared with Nigerian samples (8.1% of our sample compared with 82.9% in Nigeria). This may be explained by the fact that half of our sample had either personally experienced mental illness or were related to someone who had. Promoting direct personal contact between individuals experiencing mental illness and the general public has been shown to reduce stigma [22
]. Stigma / increased social distance have also been found to be correlated with a lack of personal contact with mental illness in three Nigerian studies [25
]. It is possible that many in Gureje’s Nigerian community sample had not had this personal experience compared to our population. This does not however explain why we found no difference in stigma scores between psychiatric and non psychiatric clinic attendees. According to the available evidence it would be expected that those attending psychiatric clinics would have had more personal experience of living with mental illness and so express less stigmatising beliefs than those participants attending medical and surgical clinics. In Nigeria psychiatric patients have been found to experience high self stigma rates of up to 21.6% and this phenomenon may explain the lack of difference in stigma scores found between mental health and non mental health clinic attendees in our results [28
Our study has some limitations. It is possible that our sample population, consisting of literate persons attending an urban centre teaching hospital may not be representative of Malawi as a society and so limits the generalisability of our results. The role of demographic variables in stigma (and therefore any potential bias in our sample) is far from clear from the existing research that has been performed in Africa. Studies from Nigeria, Ghana and Ethiopia suggest that urban dwelling and higher education correlates with biological/ psychological attributions for mental illness and lower stigma scores [25
]. However, the most robust study in West Africa by Gureje et al found no correlation between any demographic variable, including urban, semi rural and rural dwelling [10
]. Any associations between stigmatising beliefs and demographic groupings found in our study therefore need to be interpreted with extreme caution. Though it is possible that older participants were indeed more conservative and so were more likely to consider those with mental illness a nuisance, and that younger participants might have had more experience of alcohol and illicit drugs as a direct cause of mental illness (as was shown in our results), it needs to be borne in mind that we found no consistent differences in stigmatising beliefs between demographic groups, only single question associations. Ultimately we can only speculate as to why participants held the particular views or attitudes regarding mental illness they expressed in our study. This is due to the limitations of a quantitative study design. Whilst our study is an important first step in clarifying what patients and their carers think about mental illness in Malawi further qualitative work is required to deepen our understanding of this important issue.