# Suicide ideation and/or attempt with substance use and associated factors among the youth in northwest Ethiopia, community-based – BMC Psychiatry

Jul 28, 2022

### Study area

Suicidal ideations and/or attempt with substance use were conducted in Central Gondar Zone Northwest Ethiopia. Central Gondar Zone is located in the Northwest of Ethiopia in Amhara regional which area covers 21,791.83 KM2. In central Gondar is a newly formed zone from the previous North Gondar zone. In the central Gondar zone, there are 16 districts (15 rural districts and Gondar special district) and 442 kebeles. According to the 2012 E. C population statistics the central Gondar zone population was estimated at 2,642,138. Of this, 575,656 young people (15 to 24 years old) made up 21.79% of the overall population. Males and females make up 286,385 and 289,271 of the youth, respectively.

### Source and study population

#### Sources population

All youth (15–24 years old) who were living in Central Gondar Zone.

#### Study population

All youth in the Central Gondar zone who were living in the selected Kebeles.

### Inclusion and exclusion criteria

#### Inclusion criteria

All youth aged (15–25 years old) were included in the study.

#### Exclusion criteria

Youth who were unable to communicate due to severe mental/ physical illness were excluded.

### Sample size and sampling technique

The sample size was determined by two stages, the first stage was calculated to determine the prevalence of substance use and the second stage was to determine the prevalence of substance-induced suicidal ideation and attempts among substance users as described below.

In the first stage, the required sample for this study was determined by using both single population proportion formulas and two population proportion formulas. A high value was taken from the single and two population proportion formula to get the maximum sample size. Epi-Info software was used to calculate and the following formula was also used to calculate manually.

$$frac{mathrm{n}={left({mathrm{Z}}_{upalpha /2}right)}^2mathrm{p}left(1-mathrm{p}right)}{{mathrm{d}}^2}$$

Where:p = estimated prevalence was taken from the previous study conducted in Woldiya town preparatory school among student on the current substance use (34.6%) [29].d = Margin of error (d) =3% = 0.04.Z𝘢/2 = Z value at (α = 0.05) = 1.96 corresponding to 95% confidence level.

$$frac{(1.96)^2mathrm X;0.346left(1-0.346right)}{(0.04)^2}=underline{598}$$

Since we have used the multistage sampling technique design effect of two is multiplied by the calculated sample which gives 1196. Then adding 10% (1196 × 0.1 = 119.6 ≈ 120) as a non-response rate the total sample size for prevalence is 1196 + 120 = 1316. Therefore, the minimum required to sample for this study is 1316 (Table 1).

Our study region is large, making a single stage challenging, hence the multistage sampling technique was adopted. To provide everyone an equal opportunity of being chosen, the zone was a geographic cluster based on each district’s woredas and then down to Kebele. The number of participants at each randomly selected kebele were allocated proportionally based on the size of the youth to be obtained from the central Gondar zone. Finally, from the randomly selected Kebeles, participants were recruited from Ketenas (Gott) inside Kebele until the proportional sample was filled with a cluster sampling technique which was surveyed randomly in each stage based on the rule of thumb (Fig. 1).

In the second stage, the participant were one or more substance users in the last 3 months from the whole sampled youth. From the first stage calculated sample (1316), the substance users considered whether they have suicidal ideation or attempt with substance use. Therefore, from the above (1316) recreated samples 370 were current substance users that are considered for the further analysis of substance-induced anxiety disorders.

### Operational definitions

Suicidal ideation: is the respondents, answer the question, have you seriously thought about killing self in the last one month? If yes, the respondent is considered as experiencing suicidal ideation.

Suicidal attempt: is defined as, if the respondents, answer the question have you attempted suicide in the last one month? If yes, the respondent is regarded as experiencing a suicidal attempt [30].

Youth: denote the late adolescent and young adult (15–25 years old) that developmental dramatic change takes place like rapid physical growth, cognitive and moral development as well as emotional development and change. This age category needs properly managed unless youth may prone to risk-taking behaviors, including substance use [31].

Ever substance use: based on ASSIST individuals who use at least one specific substance once in their life like alcohol, tobacco, and cigarette.

Current substance use: means based on ASSIST individuals who use at least one of the specified common substances in the last three months [32].

Anxiety: is defined based on the DASS-21 anxiety subscale, individuals who scored 15 or more were considered to have anxiety.

Stress: is defined based on the DASS-21 stress subscale, individuals who scored greater than or equal to 26 are considered as having stress [33].

Perceived social support: as MSPSS-12 items, any means scale score ranging from 1 to 2.9 could be considered low support; a score of 3 to 5 could be considered moderate support; a score from 5.1 to 7 could be considered high support [34].

### Study variables

#### Dependent variable

Suicidal ideation and/or attempt with substance users.

#### Independent variables

Personal characteristics: Gender, age, religion, marital status, educational status, occupation, residency, and current living arrangement. Family-related variables: family size, father’s/mother’s level of education, occupational status of father/ mother, biological parents alive, having a family member using a substance, the recent loss of loved ones, and having a friend using a substance. Psycho-social variables: Perceived social support, level of psychological stress. Substance use: such as alcohol and chat, and cigarettes.

### Data collection procedure and tool

The first part of the data was a socio-demographic questionnaire that include basic demographic and family-related characteristics of participants.

The second part of the data was the Alcohol, Smoking, and Substance Involvement Screening Test (ASSIST) questionnaire that was used to assess risky substance use. The ASSIST consists of eight questions covering tobacco, alcohol, cannabis, cocaine, amphetamine-type stimulants, inhalants, sedatives, hallucinogens, opioids, and other drugs. The WHO recommended ASSIST cut-off scores for conventional risk levels (low, moderate, high) are as follows: for alcoholic beverages: low risk (0–10), moderate [11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26], and high risk (27+).

Substances other than alcohol are categorized based on the following scores low (0–30, moderate [4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26], and high risk (27+).

The third part Depression, Anxiety and Stress Scale-21 Items (DASS-21) were used to assess the presence of depression, anxiety, and stress. Each of the DASS-21 scales contains 7 items, divided into subscales with similar content. The responses for each statement scored on a Likert scale from 0 to 3 which indicates how much the statement applied to you over the last week. DASS-21 needed to be multiplied by 2 to calculate the final score.

The recommended cut-off scores for conventional severity labels (normal, moderate, severe) are as follows:

Anxiety: 0–7 normal; 8–9 mild; 10–14 moderate and 15 and above severe.

Stress: 0–14 normal, 15–18 mild, 18–25 moderate, and score greater than 26 severe.

Depression: 0–9 rated as normal; 10–13 mild, 14–20 moderate; 21 and above severe.

The fourth part of the questionnaire was the Multidimensional Scale of Perceived Social Support (MSPSS) contains a 12-item scale designed to measure perceived social support from three sources: Family, Friends, and a Significant Other. The scale is comprised of a total of 12 items, with 4 items for each subscale.

Significant Other Subscale: Sum across items 1, 2, 5, & 10, and then divide by 4. Family Subscale: Sum across items 3, 4, 8, & 11, and then divide by 4. Friends Subscale: Sum across items 6, 7, 9, & 12, and then divide by 4. Total Scale: Sum across all 12 items, then divide by 12. In this approach any mean scale score ranging from 1 to 2.9 could be considered low support; a score of 3 to 5 could be considered moderate support; a score from 5.1 to 7 could be considered high support.

### Data quality assurance

The questionnaire was translated by bilingual experts after first being prepared in the English language, then back-translated Amharic language. 15 data collectors (BSc in Psychiatry) and 5 supervisors (mental health professionals) were selected, and training was provided for a 2 half days duration about data collection tools, collection techniques, and ethical issues during the selection and collection of the data. The supervisors have assessed the consistency and completeness of data on daily basis. A pre-test of all structured questionnaires was checked on 5% of young people before the main data collection. A pre-test was conducted in the North Gondar which is out of the study area among youth.

### Data analysis procedure

The data was entered into the appropriate statistical software program (Epi-Data 4.6 version) and then exported into SPSS version 20 for further analysis. Descriptive statistics were used to measure the summarize the statical data distribution. Bi-variable logistic regression was conducted to check the associated factors with suicidal ideation and/or attempt with a p-value of < 0.2 to be reanalysis in multi-variable regression. Multi-level binary logistic regression analysis was performed to assess the significant associated factors of suicide ideation and/or attempt. Adjusted odds ratio with 95% confidence interval was used to declare statistically significant variables based on p < 0.05 in the multivariable logistic regression model. The basic assumptions and model fitness of the statistical analysis model used in this study was 75.30 on Hosmer and Lemeshow test. The absence of collinearity among independent variables was determined by VIF, which means there is no collinearity since it is greater than ten. Therefore, the outcome variable was distributed normally with the determinant factors.