The study was approved by the Research and Ethics Committee of Bishop Stuart University. All individuals provided written informed consent.
The study was a cross-sectional survey that employed exploratory and analytical data analysis. This design was deemed appropriate because it is flexible and can systematically capture the information necessary for determining the prevalence of depression and associated factors among adults in the study area.
The study was conducted in three sub-counties in Mbarara district, namely, Nyamitanga, Rugando and Rwanyamahembe. A large fraction of the population in these three sub-counties is socially and economically disadvantaged and lives in the less urbanised part of Mbarara district. The district is bordered by Ibanda district in the north, Ntungamo district in the west, Isingiro district in the south and Kiruhura district in the east. The district headquarters formerly at the newly created Mbarara City, the largest urban center in the sub-region, are located approximately 290 kilometres (180 miles) southwest of Kampala, Uganda's capital city and largest metropolitan area. Study participants were deemed eligible if they were 18 years and older, resided in the study area and wilfully accepted to give written consent. Those below 18 years and those who appeared too weak to participate in the study were excluded.
The researchers used multi-stage sampling technique to select sub-counties. This technique was applied because the study involved different stages of sampling before selecting the actual study participants. In the first stage, cluster sampling was used to select parishes. In the second stage, simple random sampling, particularly lottery system, was used to select villages from each parish mainly because village leaders did not have comprehensive lists of people living in their communities. Villages in each parish were used as a sampling frame and households were randomly selected from each village. Participants were randomly selected and only eligible ones who provided consent were interviewed and the number of participants determined based on the size of a village.
A semi-structured questionnaire with both closed and open-ended questions in English and Runyankore was employed to obtain data on factors associated with depression and measures for reducing depression risks. Trained interviewers who spoke the same language as respondents were deployed to help with translating and administering the questionnaire. Section A consisted of participants’ demographic information including their gender, age, marital status and level of education. Sections B, C and D comprised various questions regarding medical history and behaviour of each participant and possible measures for reducing depression risks in the study area. A 7-item version of Hopkins Symptom Checklist for Depression (HSCL-D) was used to assess depression [4,12]. The checklist enabled participants, a representative sample from the general population, to rate how often they have experienced different symptoms associated with depression in the last 7 days. Scores ranged from 1 ‘‘not at all’’ through 4 ‘‘very much’’ and were based on a Likert-type scale. Level of depressive symptoms was then calculated as the mean score for each participant, with mean scores exceeding 1.75 indicating probable depression.
Data screening and statistical analyses
The data collected was entered in Microsoft Excel and exported to Minitab (version 14) for analysis. Scores from a 7-item version of Hopkins Symptom Checklist for Depression (HSCL-D) were averaged and the probable depression determined for each participant using a cut-off of 1.75 [4,12]. Caseness is dichotomous. It takes only two values, 1 if the score exceeds the cut-off and 0 otherwise. For objective 1, the prevalence of depression was determined by calculating the percentage of participants who met caseness for depression. For objective 2, bivariate and multiple logistic regressions were used to investigate associations between depression outcomes and socio-demographic, behavioural and medical history variables. Variables with a p-value less than 0.05 in the bivariate analysis were included in the multivariate model to identify factors associated with depression after adjusting for potential confounders. Results from statistical tests were considered statistically significant if the p-value was less than 0.05. Finally, for objective 3, we analysed the measures for reducing depression risks by calculating the frequency and percentage of each measure mentioned by the participants.
Data presentation, analysis and interpretation
A total of 383 adults aged 18 and above from the sub-counties of Nyamitanga, Rugando and Rwanyamahembe participated in the study. Out of 383 participants, 212 (55.4%) were female and 171 (44.6%) were male. The majority 229 (60.4%) had attained secondary education and above while 24 (6.2%) had never had formal education. Their mean age was 35.9 years, with the youngest aged 18 and the oldest 85 years. Households consisted of 5.5 members on average (range: 1–20). More details of the socio-demographic characteristics and behavioural and medical history of the participants are shown in Table 1.
Prevalence of depression
The frequencies of each symptom endorsed are listed in Table 2, with the most common being sadness and tension. Feeling hopeless about the future and feeling of worthlessness were the least experienced symptoms of depression (Table 2). These scores were obtained using a 7-item version of Hopkins Symptom Checklist for Depression. They were averaged for each participant and the probable depression determined using a cut-off of 1.75. From this analysis, it was revealed that the prevalence of depression in the study sample stands at 27.7% (Figure 1).
Factors associated with depression
At a bivariate level, logistic regression was used to determine the single factors associated with depression. The analysis revealed that problem alcohol use (OR = 1.9, 95% CI = 1.16–3.13, P = 0.011), drug use (OR = 4.92, 95% CI = 1.88–12.88, P = 0.001), discrimination (OR = 3.34, 95% CI = 1.93–5.79, P = 0.001), being HIV positive (OR = 2.01, 95% CI = 1.13–3.6, P = 0.018) and taking medications routinely (OR = 2.14, 95% CI = 1.31–3.48, P = 0.002) are associated with a higher likelihood of getting depression. On the other hand, the analysis showed that having fewer members in a household (OR = 0.56, 95% CI = 0.35–0.9, P = 0.017), being educated (primary education: OR = 0.44, 95% CI = 0.21–0.95, P = 0.036; secondary education: OR = 0.35, 95% CI = 0.16–0.75, P = 0.007; tertiary education: OR = 0.34, 95% CI = 0.16–0.73, P = 0.006) and having family support (OR = 0.47, 95% CI = 0.28–0.78, P = 0.004) are associated with a lower risk of depression (Table 3).
At a multivariate level, all factors that had a p-value less than 0.05 in the bivariate analysis were included in the model. From this analysis, the results reveal that participants who had had drug-abuse related problems were 4.4 times more likely to be depressed than those who had not had such problems (Table 3; AOR = 4.44, 95% CI = 1.34–14.74, P = 0.015). In addition, participants who had experienced discrimination on account of their health status were 2.3 times more likely to be depressed (AOR = 2.33, 95% CI = 1.22–4.45, P = 0.010), as compared to the participants who had not experienced discrimination. Participants who hailed from households with five or fewer members had 46% reduction in risk of being depressed (AOR = 0.54, 95% CI = 0.32–0.93, P = 0.026), those with primary education had 64% reduction in risk (AOR = 0.36, 95% CI = 0.15–0.88, P = 0.024), those with secondary education had 70% reduction in risk (AOR = 0.30, 95% CI = 0.12–0.76, P = 0.011) while those with tertiary education had 63% reduction in risk of suffering from depression (AOR = 0.37, 95% CI = 0.15–0.92, P = 0.032).
Measures for reducing depression risks
Using the questionnaire, I also sought views of study participants regarding measures that can be implemented to reduce depression risks among adults in the study area (Table 4).
Most of the participants (55.9%) called for more guidance and counselling services which can be achieved by establishing counselling centres and employing professional counsellors, 43.9% suggested the need for awareness campaigns and self-help projects while 29.2% pointed out the need to improve access and quality of medical services (Table 4). Other measures suggested include stress management and spiritual care services which may include prayers, retreats and fellowships, as reported by 11.0% of the participants, socialisation and problem sharing (11.0%), establishing specialised clinics and rehabilitation centres for people with depression (5.0%), timely screening and adherence to medication (4.2%), promoting nutritional management and physical exercise (3.7%), treatment of depression patients with respect (3.1%) and helping adults improve self-esteem and avoid negative influence through patient empowerment programs (2.9%).