The study protocol was approved by the Frantz Fanon University Institutional Ethics Review Committee (Reference: FFU-00000002), and all methods were performed in accordance with the principles of the Helsinki Declaration. The data collection was initiated after a letter of permission was obtained from each district and sub-district administrator. Informed consent was obtained from each participant.
Study design, setting, and population
A community-based cross-sectional study was conducted in August 2023 in Hargeisa, the capital city of the Republic of Somaliland, a self-declared country in the Horn of Africa. The city is divided into eight districts. Each district are divided into administrative units called sub-districts. In 2023, an estimated 1.9 million people were residing in the city 30, of which adult population constituted 44% 31.
All Hargeisa City residents (for a minimum of 6 months prior to the survey) were the source population. Residents of Hargeisa City with an age of ≥ 18 years were the study population. Residents living in randomly selected households as heads or any other household members greater than or equal to 18 years of age and residents living more than 6 months and available during data collection were included in the study. Subjects who left the house for some reason and who were seriously ill and who were unable to communicate were excluded from the study.
Sample size determination
The sample size was determined using a single population proportion formula as denoted blow
$$\text{N}= \frac{\left.{Z}_{{\alpha /2}^{2}}\times p(1-p\right)\times \text{D}}{{d}^{2}}=\frac{\left.{1.96}^{2}\times 0.149\times (0.851\right)\times 2}{{0.05}^{2}}=390$$
Where
N = minimum sample size required for the study
Z = Standard normal distribution with confidence interval of 95%, Z = 1.96
P = Proportion of the prevalence of CMD conducted in Eastern Ethiopia in 2016 which was 14.9% 11; hence P = 14.9% (0.149) was used.
d = Absolute precision or tolerable margin of error (d) = 5% = 0.05
D = design effect (D = 2) was used, because of multistage sampling technique 4.
Then adding the 10% (390\(\times\) 0.1 = 39) of non-respondents, the total sample size for this study was 429.
Sampling technique and procedure
Multi-stage sampling was used to draw study respondents; at the first stage, two districts were selected randomly using the lottery method, followed by selection of one sub-district from each. The number of households in each sub-district was allocated proportionately during the second stage. Then, each household was chosen using a systematic sampling approach. To determine the sampling interval, which was three, the total number of households in each sub-district was gathered from each administrative sub-district. After the first household was randomly selected, households at every third interval were approached. If a household had two or more eligible adults, only one of them was selected randomly by the lottery method. The next household was approached if the chosen household remained closed even after a revisit or if adult was ineligible.
Data collection tools and procedures
Face-to-face interviews were conducted using hand-held mobile devices loaded with the KoBo Toolbox, which contained structured questionnaires addressing socio-demographic and behavioral factors and the Kessler Psychological Distress Scale (K10). The outcome variable (common mental disorder) was assessed using K10 questionnaire of 10-item. It has 10 questions that are answered by “none of the time," “a little of the time," “some of the time," “most of the time," or “all of the time," with codes from 1 to 5. The internal consistency, measured by Cronbach’s alpha, was good to excellent in previous validation studies 32–34. The K10 in this study showed excellent internal consistency reliability coefficient (Cronbach’s alpha = 0.93).
Data quality control
To ensure the quality of data, the K10 items were adopted accordingly. The questionnaire was translated into Af Somali, by an expert and then translated back to the English language by another person to check for consistency. Data were collected by a public health professional who was masters holder supervised by a psychiatrist who was a PhD holder. A pretest was conducted one week prior to the actual data collection on a population of 5% of the determined sample size who were residents of other districts that were not included in the main survey to see the applicability of the instruments, and feedback was incorporated into the final tool to improve the quality. The collected data was checked daily for completeness, and the K-10 questionnaire had excellent internal consistency.
Study variables and measures
The dependent variable was “suffering from common mental disorder (CMD)” with the values Yes or No. CMD was measured using 10 items (k10) with a cut-off point of > 21.85, which was the overall K10 mean score. Subjects who scored less than 21.85 were considered to have no CMD, and those who scored greater than or equal to 21.85 were considered to have a CMD. Furthermore, individuals with a K10 score from 10 to 19, 20 to 24, 25 to 29, and 30 to 50 were defined as normal, having mild mental disorder, moderate mental disorder, and severe mental disorder, respectively, based on the Clinical Research Unit for Anxiety and Depression (CRUfAD) 35.
Covariates were obtained from the baseline survey data, including gender (1 = man, 2 = woman), age, marital status (0 = single, 1 = married, 2 = divorced), highest educational level, employment status (0 = unemployed, 1 = self-employed, 2 = government employee, 3 = private employee), monthly household income, smoking status, physical activity, habit of khat chewing, frequency of khat chewing (1 = ≤ 4 days per week, 2 = > 4 days per week), and duration of khat chewing (1 = ≤ 2 years, 2 = > 2 years). For age, participants were divided into three groups (1 = 18–39 years, 2 = 40–59 years, and 3 = ≥ 60 years). The highest educational level was categorised as “no formal," "primary," "secondary," and "university." Monthly household income consisted of “≤$100," "$101-$300," “$301-$500," and "≥$500." Smoking status was classified into three groups (0 = never smoking, 1 = former smoking, and 2 = current smoking). The habit of khat chewing was a dichotomous variable between “no” and "yes." Likewise, physical activity was categorised as "no physical activity," "some physical activity," and "regular activity."
Data management and analysis
We exported and downloaded the KoBo Toolbox data in XLS forms and cleaned and coded it in Excel before importing it into SPSS.
The internal consistency, measured by Cronbach’s alpha, was good in previous validation studies 32, 33. The K10 in this study showed excellent internal consistency reliability coefficient (Cronbach’s alpha = 0.93).
We used the Kolmogorov-Smirnov test to assess the normality of the scale variables. Continuous variables with normal distribution were presented as mean (SD). Non-normally distributed variables were shown as median and inter-quartile range. Categorical variables were presented as numbers and percentages.
We examined the difference in baseline characteristics and CMD status (yes or no) by using chi-square test for categorical variables. The binary logistic regression analysis was conducted to examine the relationship between khat and CMD status, controlling for confounding factors. The estimates represent the odds ratio (OR) with A 95% confidence interval (CI). Statistical analyses were performed with IBM SPSS v. 20, and the level of significance was set to a two-tail p-value < 0.05.