Magnitude and associated factors of suicidal ideation and attempt among people with epilepsy attending outpatient treatment at primary public hospitals in northwest Ethiopia: a multicentre cross-sectional study.

OBJECTIVE: This study aimed to assess magnitude and associated factors of suicidal ideation and attempt among people with epilepsy attending outpatient treatment at primary public hospitals, northwest Ethiopia using suicide module of World Mental Health Survey initiative version of the WHO, composite international diagnostic interview.
DESIGN: Multicentre-based cross-sectional study was used.
SETTING: Data were collected using face to face interview from patients with epilepsy who attended outpatient treatment at primary public hospitals at northwest Ethiopia. PARTICIPANTS: Adult patients with epilepsy (n=563) who came to attend outpatient treatment during the study period were included in the study using systematic random sampling technique. OUTCOME MEASURES: Suicidal ideation, suicidal attempt and factors associated with suicidal ideation and attempt.
RESULTS: The overall magnitude of suicidal ideation and attempt was 26.5% and 12.6%, respectively. Being woman adjusted OR ((AOR)=1.68, 95% CI 1.09 to 3.23), living alone (AOR=2.4, 95% CI 1.47 to 3.92), divorced/widowed/(AOR=2.2, 95% CI 1.09 to 7.8), family history of suicidal attempt (AOR=2.53, 95% CI 1.34 to 4.79), depression (AOR=3.18, 95% CI 1.85 to 5.45), anxiety (AOR=2.92, 95% CI 1.68 to 5.09), comorbid medical illness (AOR=2.60, 95% CI 1.17 to 5.82) and poor social support (AOR=2.35, 95% CI 1.26 to 4.40) were statistically associated with suicidal ideation. Depression (AOR=4.87, 95% CI 2.56 to 9.28) living alone (AOR=2.66, 95% CI 1.62 to 5.41), family history of committed suicide (AOR=2.80, 95% CI 1.24 to 6.39), taking medication for mental illness (AOR=2.17, 95% CI 1.06 to 4.46), hazardous alcohol use (AOR=2.10, 95% CI 1.05 to 4.23) were statistically associated with suicidal attempts at a p value <0.05.
CONCLUSION: This study showed that the magnitude of suicidal ideation and attempt was high among people with epilepsy. Being woman, living alone, having depression and anxiety, single, divorced/widowed in marital status, family history of suicidal attempt and poor social support were statistically associated with suicidal ideation. Having depression, living alone, family history of suicide attempt, hazardous alcohol use and drug taking for mental illness were statistically associated with suicidal attempt. Based on the findings of this study early screening, detection and management of suicide were recommended in people with epilepsy. © Author(s) (or their employer(s)) 2021. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

This is a multicentre study to assess the magnitude of suicidal ideation and suicidal attempt among patients with epilepsy attending treatment at primary healthcare settings.

The study identified various factors associated with both suicidal ideation and suicidal attempt.

Study used standardised tools to assess both suicidal ideation, suicidal attempt and independent variables with both patient interview (primary data) and chart review (secondary data).

Recall bias could be a limitation to this study particularly for questions related to an outcome variable suicidal ideation.

Since the data were collected during a time when there was fear of COVID-19 pandemic, patients with epilepsy particularly with other chronic physical comorbid condition were thought to send their relatives to collect their monthly medications and this might have an effect on the current results.

Background

Suicide is fatal act that represents the person’s wish to die. There is complex process that involves a series of pathways and mechanisms that starts from initiation of suicidal ideation to planning for days, weeks or even years before acting, while others take their lives seemingly on impulse without premeditation.1 2 The Diagnostic and Statistical Manual of Mental Disorders defines suicidal ideation as a thinking about, considering or making plans for suicide and suicidal attempt a deliberate, self-destructive act with a clear expectation of death that is non-fatal.3

The burden of suicide constitutes a serious public health issue in the world.4 According to a global estimate from the WHO, around 800 000 people die due to suicide annually. This is corresponding to an age-standardised suicide rate—of around 11.5 per 100 000 people—a figure equivalent to someone dying in every 40 s.5 The worldwide burden of suicide is estimated to increase to 2.4% by 2020 year, and the rate of death due to suicide will be one person every 20 s.6

Suicidal acts result from a complex interaction of biological, psychological, sociological, cultural and environmental factors.7 From the biological and psychological conditions, people with mental and neurological illness such as epilepsy are vulnerable. Epilepsy is a chronic disorder of the brain and is one of the most common serious neurological disorders worldwide with no boundary to age, race, social class, nationality or geographical location.8 It affects more than 50 million people in the world.9 The prevalence of people with epilepsy (PWE) among the general population varies across the countries from 0.5% to 5%.10 The community-based epidemiological study in Ethiopia reported that PWE were 5.2/1000 population.11 Epilepsy affects relationship with family and friends, employment, school, leisure activities and results in social and economic consequences. Each of these effects may lead to high magnitude of comorbid psychiatric illness among PWE.12

PWE are at higher risk for suicide when compared with the general population. For example, the life-time prevalence of suicidal ideation among PWE that was reported 25% is higher than people without epilepsy 13%.13 The suicidal rate reported that people with temporal lobe epilepsy (TLE) 6%–25% times when compared with 1.4%–6.9% in the general population.14 This indicated that suicide is common comorbid psychiatric illness among PWE.15 Similarly, different study results showed that PWE are high risk for suicidal ideation and attempt.16 17 The estimated lifetime prevalence of suicidal attempt among PWE ranged from 3.3% to 14.3%,18 whereas about 38% of patients with epilepsy had suicidal ideation.17 The estimated prevalence of suicidal ideation and attempts was 45.2% and 28.6%, respectively.19 Around 11% of deaths among PWE are due to suicide and the suicide attempt increase the chance of later completing suicide by 38%.20 Results from meta-analysis indicated that the pooled prevalence for suicide ideation (24 studies) and suicide attempts (18 studies) was 23.2% and 7.4%, respectively.21 Centres for disease control and prevention report that the suicide rate among people with epilepsy is 22% higher than the general population with 16.89/100 000 people with epilepsy compare with 13.84/100 000 in the general population.22

The greatest risk factors of suicidal ideation and attempts among people with epilepsy is psychiatric disorders including depression and other mental disorders.23 Poor social support, drug treatment for mental illness, those who had psychiatric comorbidity such as comorbid depression, low economic status and unemployment were also reported risk factors from prior studies.24 Other factors included poor interpersonal relationship, stigma related to the epilepsy and pre-existing psychiatric illness.25–27 In addition, early age onset of epilepsy, antiepileptic drug polytherapy, high seizure frequency and duration of epilepsy were factors found to significantly associated with suicidal ideation and attempt.13 28

Despite this burden and consequences, there is limited study done on suicide ideation and attempt among people with epilepsy in eastern Africa as well as in Ethiopia. There was a single study conducted in Addis Ababa Ethiopia in 2014 in Saint Amanuel Specialized Mental Hospital the only specialised mental hospital in Ethiopia that provide. The suicidal ideation and attempt among people with epilepsy reported in the study was 29.8% and 14.1%, respectively, with various significantly associated risk factors.24 However, there are limited data in the country particularly suicide among patients with epilepsy in settings where psychiatry service limited and or not available including in patients attending primary hospitals. Therefore, this study aimed to assess the magnitude and associated factors of suicidal ideation and attempt among people with epilepsy. This would help for future integrated intervention and it would be an input of information for policymakers to institute intervention strategies.

Methods

Study design and setting

Institutional based cross-sectional study was conducted from 15 March to 1 May 2020, at central Gondar zone primary public hospitals in four primary hospitals named, ‘Wogera’, ‘Dembia’, ‘Delagi’ and ‘Aykel’ primary hospitals, which are located in the northwest part of Ethiopia.

Study participants

All adult patients (age ≥18 years), who had been clinically diagnosed with epilepsy and who were at outpatient treatment at Delagi, Wogera, Dembia and Aykel primary hospitals were study participants. Patients who were unable to communicate and seriously ill at the time of the data collection were not included.

Patient and public involvement

No patients were involved.

Sampling

The adequate number of samples required for this study were 569, determined by using single population proportion formula (ni = (Zα/2)2 p (1–p)/d2). By using a 14.1% prevalence of suicidal attempt24 and setting 3% tolerable margin of error, 95% CI and 10% non-response rate. A systematic sampling method was used to select study participants from selected hospitals during the study period using the number of patients following treatment per month as a sampling frame in each of the hospitals. Study participants were recruited based on proportional allocation from each primary hospital.

Data collection and measurements

Data were collected by face to face interviews using a structured questionnaire and through patients’ chart reviews. Suicidal ideation and attempt were assessed by items that were adopted from module of World Mental Health survey initiative version of the WHO composites international diagnostic interview in which suicide was studied and validated in Ethiopia both at clinical and community settings with internal consistence Cronbach’s alpha=0.97.29 30Its Cronbach’s alpha in current study was 0.85. Standardised and validated measurements in low-income countries including Ethiopia were used to assess other variables including Fast alcohol assesment tool (FAST) for hazardous alcohol use,31 32 Hospital Anxiety and Depression Scale for depression and anxiety33 34 and Kilifi stigma scale (KSS) for perceived stigma.35 Social support was collected by Oslo-3 item of Social Support Scale. It is 3-item questionnaires commonly used to assess social support and it had been used in several studies. The sum of score scale was ranging from 3 to 14, which had three categories: poor support 3–8, moderate support 9–11 and strong support 12–14.34 The internal consistence, Cronbach’s alpha of Olso-3 items in the current study was 0.81. Eight trained Bachelors of Science in psychiatric nurses collected the data using the Amharic version questionnaires.

Data processing and analysis

The data were checked for completeness and consistency and entered to Epi-Data V.4.6.0.2 and was exported to SPSS V.20 for analysis. The binary logistic regression model was used. Bivariate and multivariate logistic regression analysis was performed to identify associated factors to outcome variables. All variables with a p value less than 0.20 at bivariate analysis were entered into the multivariate logistic regression model. A p value of <0.05 was considered statistically significant, and the adjusted OR with 95% CI was calculated. Goodness of model fitness was checked by using Hosmer-Lemeshow test.

Results

Sociodemographic characteristics of respondents from a total of 569 samples, 563 participants were included in the study with response rate of 98.9%. The median age of respondents was 29 and IQR was 23–36 years. The majority of them 57.5 % (324) was men. Regarding living arrangement of participants, 63.7 % (359) were living with family, the majority of participants 65.2% (367) was orthodox by religion and 33.7% (190) was farmers as shown below in table 1.

Table 1

Sociodemographic characteristics of people with epilepsy of central Gondar zone primary public hospitals, northwest, Ethiopia, 2020 (n=563)

Variable	Category	Frequency (n=563)	Percentage
Sex	Male	324	57.5
	Female	239	42.5
Age	18–24	169	30.0
	25–31	197	35.0
	32–38	85	15.1
	39–45	58	10.3
	>45	54	9.6
Marital status	Single	208	36.9
	Married	283	50.3
	Divorced/widowed	72	12.8
Living arrangement	With family	359	63.8
	Alone	204	36.2
Religion	Orthodox	363	64.5
	Muslim	128	22.7
	Protestant/catholic	72	12.8
Occupation	Governmentworker	55	9.8
	Merchant	120	21.3
	Framer	190	33.7
	Student	62	11.0
	Unemployed	82	14.6
	Household worker	54	9.6
Education	No formal education	119	21.1
	Primary school(1–8)	242	43.0
	Secondary school(9–12)	137	24.3
	Diploma and above	65	11.5
Residence	Rural	350	62.2
	Urban	213	37.8

Clinical, psychosocial and substance-related factors of respondents

Majority, (68.2% (384)), of participants reported the age onset of epilepsy when they were 18 years and above. Nearly half, (44.0% (248)), of the respondents had 1–6-year duration of treatment while 59.9% (337) had up to 5-year duration of illness. More than three-forth, 79.6% (448), of participants were taking one antiepileptic medication. Out of the total study participants, 21.1% (119) and 29.8% (168), were found to have anxiety and depression. respectively. whereas 8.75 %36 had comorbid medical illness. Family history of suicide indicated that about 12.4% (70) had family history of suicide attempt and 7.3%37 had family history of committed suicide. Regarding social support, 42.6% (240) had moderate social support. Out of the total participant. 66.4% (374) had not perceived stigma and 27.7% (156) of participants were current substance use as shown below in table 2.

Table 2

Description of clinical, psychosocial, substance use characteristics of people with epilepsy of central Gondar zone primary public hospitals, northwest, Ethiopia, 2020

Variable	Category	Frequency (n=563)	Percentage
Age onset of epilepsy	Under 18	179	31.8
	18 and above	384	68.2
Duration of treatment (years)	Up to 1	163	29.0
	1–6	248	44.0
	7–12	115	20.4
	>12	37	6.6
Duration of illness (years)	Up to 5	337	59.9
	6–10	140	24.9
	11–15	67	11.9
	16 and above	19	3.4
Number medication	One	448	79.6
	Two and above	115	20.4
Types of medication	Phenobarbital	322	57.2
	Phenytoin	133	23.6
	Na valproate	85	15.1
	Carbamazepine	23	4.1
Frequency of seizure	Very frequent	86	15.3
	Frequent	138	24.5
	Occasional	251	44.6
	Rare	88	15.6
Depressive	Yes	168	29.8
	No	395	70.2
Anxiety	Yes	119	21.1
	No	444	78.9
Comorbid medical illness	Yes	49	8.7
	No	514	91.3
Family history of suicidal attempt	Yes	70	12.4
	No	493	87.6
Family history of committed suicide	Yes	41	7.3
	No	522	92.7
Family history of mental illness	Yes	57	10.1
	No	506	89.9
Family history of epilepsy	Yes	58	10.3
	No	505	89.7
Medication taking for mental illness	Yes	88	15.6
	No	475	84.4
Social support	Poor	196	34.8
	Moderate	240	42.6
	Strong	127	22.6
Perceived stigma	Yes	374	66.4
	No	189	33.6
Ever substance use	Yes	231	41.0
	No	332	59.0
Current substance use	Yes	156	27.7
	No	407	72.3
Hazardous alcohol use	Yes	78	13.9
	No	485	86.1

Magnitude of suicidal ideation and attempt among people with epilepsy

The lifetime magnitude of suicidal ideation and attempt in this study was 26.5% (149) with (95% CI 23.1 to 30.4) and 12.6% (71), with (95% CI 9.9 to 15.5), respectively. Among the respondents who had a suicidal ideation, 59.1% (88) had suicidal ideation within the last 12 months and 15.8% (89) had planned to commit suicide in a lifetime (table 3).

Table 3

Distribution of suicidal and attempt among of people with epilepsy of central Gondar zone primary public hospitals, northwest, Ethiopia, 2020 (n=563)

Variable	Category	Frequency (n=563)	Percentage
Ever suicidal ideation	Yes	149	26.5
	No	414	73.5
Duration of ever suicidal ideation	≤12	88	59.1
	>12	61	40.9
Suicidal thought in last on month	Yes	36	6.4
	No	527	93.6
Ever plan of suicide	Yes	89	15.8
	No	474	84.2
Duration of ever plan	≤12	63	71.6
	>12	25	28.4
Suicidal attempt	Yes	71	12.6
	No	492	87.4
Duration of suicidal attempt	≤12	47	66.2
	>12	24	33.8
Suicidal attempt in last 1 month	Yes	19	3.4
	No	544	96.6
Frequency of suicide attempt	Once	47	66.2
	Twice	19	28.8
	More than twice	5	7.0
Reason for suicidal attempt	Family conflict	13	18.3
	Death of family	17	23.9
	Financial loss	11	15.5
	Related to current illness	14	19.7
	Physical illness	9	12.7
	Others	7	9.9
Methods of attempt	Hanging	26	36.6
	Poisoning	25	35.2
	Sharp tools	7	9.9
	Drug overdose	10	14.1
	Jumping from high place	3	4.2

Factors associated with suicidal ideation among people with epilepsy

In bivariate logistic regression analysis, variables like being woman, living alone, single, divorced/widowed, family history of suicidal attempt, family history of committed suicide, depression, anxiety, ever substance use, taking medication for mental illness, comorbid medical illness, perceived stigma, hazardous alcohol use and poor social support were significantly associated with suicidal ideation. In multivariate logistic regression analysis, variables like being woman, living alone, family history of a suicidal attempt, anxiety, depression, being single, divorced/widowed, comorbid medical illness and poor social support were statically significant for suicidal ideation with p value less than 0.05. To identify the effect of each variable, a multicollinearity statistics test was checked and the variance inflation factors of each variable were a range of 1.03–2.54 (table 4).

Table 4

Bivariate and multivariate logistic regression analysis between some selected factors and suicidal ideation among people with epilepsy at enteral Gondar zone primary public hospitals, northwest, Ethiopia, 2020

Explanatory variables	Suicide ideation		COR (95% CI)	AOR (95% CI)
	Yes (N)	No
Sex
Male	70	254	1	1
Female	79	160	1.79 (1.23 to 2.61)	1.68 (1.09 to 3.23) *
Marital status
Married	48	235	1	1
Single	72	136	2.59 (1.70 to 3.95)	1.98 (1.70 to 3.95) *
Divorced/widowed	29	43	3.3 (1.87 to 5.80)	2.20 (1.09 to 7.8) *
Living arrangement
With family	69	290	1	1
Alone	80	124	2.72 (1.85 to 3.98)	2.40 (1.47 to 3.92) **
Family history of suicidal attempt
Yes	35	36	3.32 (1.99 to 5.55)	2.53 (1.34 to 4.79) *
No	114	379	1	1
Family history of committed suicide
Yes	19	22	2.6 (1.36 to 4.96)	1.31 (0.59 to 2.89)
No	130	392	1	1
Anxiety
Yes	62	58	4.25 (2.77 to 6.53)	2.92 (1.68 to 5.09) **
No	88	356	1	1
Depression
Yes	91	77	6.87 (4.54 to 10.36)	3.18 (1.85 to 5.45) **
No	58	337	1	1
Ever substance use
Yes	87	144	2.63 (1.79 to 3.86)	1.56 (0.89 to 2.53)
No	130	376	1	1
Medication for mental illness
Yes	61	27	1.75 (1.08 to 2.83)	1.32 (0.59 to 2.89)
No	62	270	1	1
Comorbid medical illness
Yes	22	27	2.48 (1.37 to 4.51)	2.60 (1.17 to 5.82) *
No	127	387	1	1
Perceived stigma Yes	84	105	3.80 (2.57 to 5.63)	1.50 (0.89 to 2.72)
No	65	309	1	1
Social support
Poor	92	104	3.60 (2.15 to 6.07)	2.35 (1.26 to 4.40) *
Moderate	32	208	0.63 (0.35 to 1.16)	0.49 (0.25 to 1.98)
Strong	25	102	1	1
Hazardous use
Yes	33	45	3.4 (2.08 to 5.57)	0.51 (0.26 to 1.02)
No	116	369	1	1

χ2=10.07, df=8, Hosmer-Lemshow test =0.23.

**= p<0.05, and **= p<0.001.

AOR, adjusted OR; COR, crude OR.

Factors associated with suicidal attempt among people with epilepsy

In bivariate logistic analysis, variables like living alone, being divorce/widowed, family history of committed suicide, family history of mental illness, depression, taking medication for mental illness, comorbid medical illness, perceived stigma and hazardous alcohol use were statically significant to suicidal ideation at p value less than 0.2. In multivariate logistic analysis, variables like living alone, family history of committed suicide, depression, taking medication for mental illness and hazardous alcohol use were statically significant at p value less than 0.05 (table 5).

Table 5

Bivariate and multivariate logistic regression analysis showing an association between factors and suicidal attempt among people with epilepsy at central Gondar zone primary public hospitals, northwest, Ethiopia, 2020

Explanatory variables	Suicidal attempt		COR (95% CI)	AOR (95% CI)
	Yes (N)	No
Living arrangement
With family	30	329	1	1
Alone	41	163	2.76 (1.66 to 4.58)	2.66 (1.62 to 5.41) **
Marital status
Married	26	257	1	1
Single	27	181	1.47 (0.83 to 2.61)	1.02 (0.52 to 1.99)
Divorced/widowed	18	54	3.29 (1.69 to 6.43)	1.92 (0.87 to 4.24)
Family history of committed suicide
Yes	14	27	4.23 (2.09 to 8.53)	2.80 (1.24 to 6.39) *
No	57	465	1	1
Family history of mental illness
Yes	12	45	2.02 (1.01 to 4.04)	1.62 (0.69 to 3.82)
No	59	447	1	1
Depression
Yes	50	118	7.55 (4.35 to 13.08)	4.87 (2.56 to 9.28) **
No	21	374	1	1
Medication for mental illness
Yes	20	68	2.45 (1.37 to 4.35)	2.17 (1.06 to 4.46) *
No	51	424	1	1
Comorbid medical illness
Yes	38	51	2.83 (1.42 to 5.66)	1.71 (0.69 to 4.23)
No	133	341	1	1
Perceived stigmaYes	38	151	2.6 (1.57 to 4.30)	0.69 (0.51 to 1.84)
No	33	341	1	1
Hazardous alcohol use
Yes	23	55	3.81 (2.15 to 6.73)	2.10 (1.05 to 4.23) *
No	48	437	1	1

*p<0.05, and **=p<0.001, χ2=7.04, df=8, Hosmer-Lemshow test=0.53.

AOR, adjusted OR; COR, crude OR.

Discussion

The current study showed that the lifetime magnitude of suicidal ideation was 26.5% with (95% CI 23.1 to 30.4). Regarding magnitude, the finding of current study was in line with study carried out in Brazil, which reported a prevalence of 28.9%,38 a study from Egypt that showed 23.5% of suicidal ideation39 and in Addis Ababa Ethiopia where the prevalence of suicidal ideation was 29.8%.24 It was also similar with a pooled prevalence result of a meta-analysis that reported 23.2% of suicidal ideation.21

However, the result of this study was lower than study conducted in Bosnia and Herzegovina, where suicidal ideation was indicated 38%.17 A study of suicide among patients with epilepsy in Havana city of Cuba has shown in 45.2% of suicidal ideation19 whereas in Malaysian study, it was reported 33.75%40 with study result from Brazil indicating 36.7%.41 The possible reason for the discrepancy might be the study participants in which the current study used only outpatients, but the study done in Bosnia and Herzegovina and Cuba used both inpatient and outpatients. The other reason might be the sample size where 5 013 180 153 samples were used in Bosnia and Herzegovina, Havana city of Cuba, Malaysia and Brazil, respectively, which were smaller compared with the current study participants (n=563). The difference in study design used might also be another reason for this variation since this study used an institutional based cross-sectional study but a prospective cohort study was used in the Havana city of Cuba, case–control study in Malaysia and community-based case–control study in Brazil.

On the other hand, the finding of current study was higher than studies done in the USA, which reported a suicidal ideation prevalence ranging from 8% to 11.9%37 42 and a Canadian study that reported a 12.7% suicidal ideation.43 Similar studies conducted in different countries including Poland,44 Rio de Janeiro, Brazil,45 Republic of china46 and Nigeria47 reported 10%, 13.3%, 14.3% and 20.0% suicidal ideation, respectively, which were lower than the current study. The possible reasons for the difference might be variation in study design and the difference in assessment tools used. In the USA, study suicidal ideation was assessed by using the ninth items of beck depression inventory, the same suicidal assessment tool was used in the Poland and Do Rio de Janerio, Brazil studies. The time frame in which the suicidal ideation occurred might be another reason for the difference in the prevalence and many of the previous studies used the current suicidal ideation but the current study considered the lifetime suicidal ideation. The difference in sample size could also be another possible reason for the discrepancy in which the sample sizes ranged from 25 to 251 in the previous studies compared with the 563 participants of the current study.

The magnitude of the suicidal attempt in this study was 12.6% with (95% CI 9.9 to 15.5). This study result was in line with study results from Croatia,48 Brazil,41 Egypt39 and in Addis Ababa Ethiopia24 with a reported prevalence of 14.1%, 12.1%, 11.5% and 14.1%, respectively.

However, the finding of this study was lower than studies conducted in Havana city of Cuba 28.6 %,19 Bosnia and Herzegovina 18 %17 and in Brazil 21.2%.38 The possible reasons for the discrepancy might be the difference in study participants. For example, only patients with TLE were the study participants in the Brazil study, whereas both inpatient and outpatients were used in studies conducted in the Bosnia-Herzegovina and Cuba. The difference in study design could be also another reason. A prospective cohort study was used in Bosnia and Herzegovina, whereas the current was a cross-sectional study.

On other hand, the finding of this study was higher than the study conducted in Iran13 and Asia21 which was 5.5% and 7.4% attempted suicide, respectively. This discrepancy might be due to sample size differences where only 200 individuals participated in the Iran study. It was also lower than a meta-analysis study result (7.4%), which was based on pooled prevalence from 18 different studies21

Multivariate logistic regression models showed that being woman, living alone, single and divorced/widowed in marital status, depression, anxiety, family history of suicidal attempt, comorbid medical illness and poor social support were significantly associated with suicidal ideation.

In this study, women had about 1.68 times higher suicidal ideation than men. Similar result was reported from studies conducted in the USA42 and Poland.44 The possible justification for this association might be due to culture influence in which women may not discuss their problems with others as mwn, and it might be due to women having greater vulnerability to other psychosocial stress. The other possible justification might be related to depression in which women are two times more likely to have depression as compared with men.1 This is evidenced by a result from meta-analysis study that consists of studies from 30 countries indicating that point prevalence of depression was significantly higher in women.49

Participants who lived alone were 2.4 times more likely to have suicidal ideation than those who lived with family. This finding is supported by studies conducted in Washington37 and a pervious study done in Addis Ababa Ethiopia.24 This might be due to the fact that those who lived alone had no a nearby individual (family) that they could communicate and share their problems and that could lead to increased hopelessness and suicidal ideation. This had also been strengthen by the result regarding marital status in which being single and divorced and or widowed were found to be more vulnerable to suicidal ideation though it does not mean these group of participants live alone. This result was supported by a study conducted in Poland.44 This might be because of lack of social support and significant other to share emotional and other psychosocial burdens due to various environment stresses that predispose to depression with manifestations including hopelessness and suicidal behaviours.

Regarding to family history of a suicidal attempt, participants who had family history of an attempt were 2.53 times more likely to have suicidal ideation than participants who had no family history of a suicidal attempt. This result was consistent with studies conducted in Croatia,48 Cuba,19 and previous study in Addis Ababa Ethiopia.24 The possible justifications might be from a non-genetic perspective like having a common life style, learnt behaviour, faced with similar family stress and environmental related factors and may also be due to biological predisposition.

Respondents who had depression were 3.18 times more likely to have suicidal ideation than respondents who had no depression. This result was supported by study done in Washington,37 Canada,43 Bosnia and Herzegovina,17 Poland,44 Brazil,38 Malaysia,40 Korea,36 Nigeria47 and Addis Ababa Ethiopia.24 Many of this previous studies justified that the presence of depression is highly associated with suicidal ideation. The possible reason might be due to the direct effect of depression, which makes individuals feel hopeless and worthless. It also indicated that the decreased level of serotonin neurotransmitter in the brain of a depressed individual was found to be associated with increased suicidal behaviour.50

Regarding anxiety, participants who had anxiety were 2.9 times higher likely to have suicidal ideation than participants who had no anxiety. This was similar with the study done in Canada,43 Brazil,41 Korea36 and Nigeria.47 The possible reason for this may be fear of adapting to and anxious feeling about the illness and drug, fear of a seizure attack, increasing psychological stress and pressure related to epilepsy.

Another factor that was associated with suicidal ideation was poor social support. Respondents with poor social support were 2.35 times higher likely to have suicidal ideation as compared with strong social support. It was similar with other studies done in Havana city of Cuba,19 China46 and Addis Ababa Ethiopia.24 The WHO report revealed that weak social ties and low support from relatives have been significantly associated with suicidal ideation.51

With respect to a suicidal attempt, respondents who had depression were 4.87 times more likely to have a suicidal attempt than those who had no depression. This is consistent with study conducted in Canada,43 Washington,37 Poland,44 Brazil,38 Malaysia,40 Korea,36 Nigeria47 and Addis Ababa Ethiopia.24 The possible justification for this might be that depression will decrease the neurotransmitter serotonin, as studies have shown an association between decreased level of serotonin and its metabolite and being suicidal.50 On the other hand, depression can be a direct effect on the feeling of hopeless and worthless.

Those study respondents who were taking medication for mental illness in addition to antiepileptic medication were highly risk for suicidal attempt. This was similar to study conducted in Washington,37 Croatia,48 Korea36 and Addis Ababa Ethiopia.24 Having two chronic illnesses and pill burden to treat the illnesses took longer time to recover and make negative feeling, which could lead feeling hopeless and worthless and finally suicidal attempts.

Those study participants who had a family history of committed suicide were 2.8 times more likely to have a suicidal attempt as compared with those who had no family history of a committed suicide. The result is supported by the previous study done in Addis Ababa Ethiopia.24 The possible reason might be due to biological perspective and might be due to non-biological factors like sharing similar family stress and environmental factors.

Regarding alcohol use, participants who had hazardous alcohol use were 2.1 times higher to a suicidal attempt as compared with those who had no hazardous alcohol use, this was supported by a study conducted in Croatia.48 This might be due to participants who drink alcohol were exposed to increased impulsiveness, poor judgement and a weakening of normal restraints against dangerous behaviour. And they might also consider suicidal ideation in withdrawal state due to dysphonic feeling related with alcohol withdrawal.

Limitation of study

The use of retrospective items in the questionnaire may have incurred recall bias like duration of illness and duration of treatment. The cross-sectional nature of the study was its main limitation. In addition, the ever suicidal ideation and attempt were used to do the analysis as there were very few responses for the current suicidal ideation and attempt. This could be another limitation as it was difficult to determine the temporal relation between suicidal ideation and attempt and associated factors. It is difficult to address types of epilepsy since it was difficult to get specific diagnosis from the patient chart. Though there are evidences from previous study results that indicated the role of personality disorders in suicide, the current study did not assess personality disorders among patients with epilepsy.52

Conclusion

This study showed that suicidal ideation and attempt were common among people with epilepsy. Being woman, living alone, single, divorced/widowed in marital status, family history of suicidal attempt, anxiety, depression, comorbid medical illness and poor social support were significantly associated with suicidal ideation. Family history of committing suicide, living alone, depression, taking medication for mental illness and hazardous alcohol use were significantly associated with suicidal attempt. Therefore, it is better to give due emphasis and screen patients with epilepsy to suicidal ideation and ask history of suicide attempt including depression and anxiety as routine clinical evaluation. Providing intervention targeting risk factors to these groups of people is also vital.