Prevalence and factors influencing low social support from family, friends, and significant others among people with mental illness attending psychiatric outpatient clinics in Gondar, Ethiopia

Abstract

Background

Evidence demonstrates a correlation between high social support and better health outcomes. However, people with mental illness are less likely to receive adequate social support to overcome mental health challenges when compared with the general population. The main objective of this study is to determine the prevalence and associated factors of low perceived social support from family, friends and significant others among people with mental illness attending a psychiatry outpatient clinic in Gondar, Ethiopia.

Methods

Data were collected from 636 participants attending a psychiatry outpatient clinic at the University of Gondar Comprehensive Specialized Hospital. Two individuals consented but did not complete the questionnaires from October 2022 to March 2023. A systematic random sampling technique was used to recruit participants. Perceived social support was measured with the Multidimensional Perceived Social Support scale (MPSS-12). Other questions assessed self-esteem, drug adherence, substance use and severity of illness, alongside sociodemographic factors. Variables were coded and entered into SPSS-28 software for further analysis. Bivariate and multivariate logistic regression analysis was conducted. Adjusted odds ratio (AOR) with 95% confidence interval (CI) and p-value less than 0.05 were considered significant.

Results

Prevalence of low perceived social support was 12.1% (N = 77). In the final regression analysis, several factors were associated with a greater likelihood of overall low perceived social support. These include living alone (OR = 2.40, CI = 1.24,4.63), having a relapse (OR = 2.13; CI = 1.13,4.02), family not participating in patient care (OR = 4.67; CI = 2.49,8.76), having moderate and severe objective severity (OR = 2.51; CI = 1.41,4.45), having low self-esteem (OR = 2.36; CI = 1.34,4.15) and having poor drug adherence (OR = 1.99; CI = 1.02,3.89).

Conclusions

Over 10% of people with a mental illness attending the outpatient psychiatry clinic reported low perceived social support. The study indicates that efforts to address low social support should focus specifically on patients, including those: living alone, having relapses, having families that do not participate in patient care, having moderate and severe objective severity of illness, having low self-esteem, and having poor drug adherence.

Clinical trial number

Not applicable.

Introduction

According to Lin (1979), perceived social support is “support provided to an individual through social relationships to other people, groups, and the community” [1]. Researchers have consistently found positive correlations between social support and psychological well-being [2, 3]. Perceived social support is an essential aspect of human relations vital to the human experience [4]. People who experience low perceived social support from their social network believe they are not receiving enough support, empathy, or encouragement [5]. In times of need, they may feel alone or detached and believe that they cannot depend on other people, resulting in them developing feelings of isolation and further stress [6]. This can contribute to mental health problems or make it more difficult for people to manage existing mental health issues [7, 8].

Social support comprises structural (social engagement frequency and connectivity) and functional (emotional and practical assistance) dimensions [9]. Interpersonal interactions and communication, monetary assistance, love and understanding, concern and care, affection and friendship, respect and acceptance are all examples of social support [10]. Different scholars suggest that the functional dimension of social support is a better indicator of good health than the structural dimension, even though both are essential [11]. The benefits of social support include improved quality of life (QoL) and protection from adverse life outcomes [12, 13]. Social variables significantly impact the recurrence and persistence of the disease, which can play a significant role in establishing and promoting health [14]. As such, focusing on social variables that affect mental health is crucial, and perceived social support is one of those factors [15].

Studies have identified that social support improves the QoL of people with a mental illness [16–20]. An effective social support from family and friends has been found to contribute to better QoL among schizophrenia patients [21], but those patients with poor social support have a poorer QoL [19]. Social support is also essential in assisting people with a mental illness in coping with the challenges of life and disease [22]. Social support has also been associated with satisfaction with services [23], Improved symptoms of illness, recovery, and social functioning [24] and psychological well-being [25, 26]. Evidence from several systematic reviews has also shown associations between social support and mental health more generally [27–31].

It is commonly recognised that people with severe mental illnesses engage in fewer interpersonal relationships, have reduced contact with individuals who are healthy [32] and have lower access to social support. For example, a study in Singapore found lower perceived social support among people with mental illnesses compared to those with no mental illnesses [33]. Additionally, a survey of Taiwanese adults with mental illnesses found that 72% of respondents had poor perceived social support [22]. Another study conducted in Malaysia with people affected by schizophrenia indicated that 72% had poor perceived social support, and support from significant others was at its lowest, followed by friends and family [21].

In Africa, limited studies have been conducted on perceived social support among people with mental illnesses. A study conducted in Egypt among inpatient and outpatient departments of a mental health hospital showed that two-thirds of participants reported low social support [19]. Another study in South Africa found an association between perceived social support and recovery for patients with schizophrenia [34]. In Ethiopia, our search found only one study, which was conducted among schizophrenia patients in a hospital, where social support was assessed, with findings indicating: low perceived social support, medium perceived social support, and high perceived social support to be 21.5%, 58.5%, and 20%, respectively [10]. In general, studies of patients with mental illnesses have revealed that many people do not have adequate social support [35–38].

Different factors, such as marital status, the onset of the disease, and prior hospitalisation, impact the level of social support for people with mental illnesses [39], the type of mental illness [33], poor drug adherence, hospital admission, level of education [40], substance use [41] have been linked to levels of social support, self-stigma [42] and relapse [43] may also contribute to poor social support, where patients think that people do not want to help them or that they do not deserve support.

Perceived social support from family, friends, and significant others varies depending on several factors, such as the nature of the relationships, the level of emotional closeness, and the type of support provided. Family provides unconditional and long-term support, while friends offer companionship and emotional backing. Significant others, including partners and spouses, contribute emotional intimacy, companionship, and sometimes financial assistance. This high prevalence of low perceived social support is more common from friends and significant others. people with mental illness often report receiving less social support from friends and significant others than from family members. This difference can be attributed to several factors, including misconceptions about mental illness that may cause friends and significant others to distance themselves. Furthermore, those facing mental health challenges often lean more on their families, who typically feel a sense of responsibility that friends and significant others may not share.

However, there is limited evidence of prevalence and associated factors of low perceived social support among people with mental illnesses in African contexts, and the only existing study from Ethiopia focused on prevalence but not the association of low perceived social support. Therefore, the current study aims to fill this research gap by assessing the prevalence and association of low perceived social support during outpatient follow-up among patients with mental illnesses in Ethiopia.

Methods and materials

Study design and procedure

This hospital-based cross-sectional study was conducted from October 2022 to March 2023 among people with mental illnesses at an outpatient psychiatry clinic of the University of Gondar Comprehensive Specialized Hospital in Gondar Town. Gondar town is situated in the Amhara National Regional State, Northwest Ethiopia. The university serves as the catchment area for more than ten million people and is a referral facility for people in Northwest Ethiopia. A total of 2400 patients attended the clinic during the current study period.

The current study was part of a large study that examined the QoL of people with mental illnesses. The sample size for the QoL study was calculated using a single population proportion formula, considering several assumptions. From a previous study conducted in Ethiopia, a 41% prevalence of poor QoL, a 95% confidence, and a 4% margin of error [44]. The following formula was applied: n = (Zα/2)² * P (1-P)/d2, where n is the minimum sample size needed, Z is a standard normal distribution (Z = 1.96) with a confidence interval of 95% CI and ⍺ = 0.05, d is the absolute precision or tolerable margin of error (4%), and P = estimated proportion is assumed to be 41% (0.41). Then, n= (1.96)² *(0.41) *(0.59)/ (0.04)²= 580 and 10% nonresponse rates (580 *10/100) = 58, with a 10% non-response rate of 580 + 58 = 638.

A systematic random sampling technique was employed to determine the sample size from a total of 2400 patients, indicating that a sample of 638 patients would be sufficient. After applying the inclusion criteria, which required patients to have been followed up for at least three months and be 18 years or older, 638 patients were included in the study. However, two individuals consented but did not complete the questionnaires, resulting in their exclusion from the analysis.

The study obtained ethical clearance from the Flinders University Human Research Ethics Committee (Project No: 5416), and the permission to conduct the survey in Ethiopia was obtained from the Institutional Review Board of the University of Gondar. Before their participation, written informed consent was diligently obtained from all study participants. Participants were fully briefed on the study’s objectives and informed of their freedom to withdraw without repercussions. Furthermore, to safeguard the confidentiality and privacy of the participants, the study employed code numbers instead of personal identification, ensuring that their personal information remained secure and undisclosed. These ethical measures underscore the commitment to the well-being and rights of the study participants and are based on established ethical principles in research.

Participants

All patients over 18 years old who visited the outpatient clinic of the hospital during the study period and had received outpatient care for at least three months for their mental disorder were potential study participants. Patients with acute mental illness who could not communicate because of severe physical or mental illness were excluded. The data were collected using an interviewer-administered standardised questionnaire during a face to face interview at the outpatient psychiatry clinic by nurses. The questionnaire was prepared in English and translated into the local language, Amharic. Five psychiatry nurses and two MSc psychiatry supervisors participated in the data collection.

Measures and operational definitions

Validated measures were used for several variables, with the remaining measured using single item measures.

Perceived social support

The multidimensional scale of perceived social support (MSPSS) was used to assess perceived social support. This tool has been adapted and validated in various settings worldwide, including in African countries such as Malawi and Uganda [45–48]. A pilot study was conducted to evaluate the tool’s performance, and it demonstrated excellent internal consistency, with a Cronbach’s alpha of 0.92. The instrument assesses perceived social support sufficiency through three subscales: family, friends, and significant others [45]. The MSPSS has 12 items (each subscale has 4 items) with 7-point Likert scale response options from 1 (very strongly disagree) to 7 (very strongly agree) [45, 49]. Significant other subscale items 1, 2, 5, & 10, family subscale items 3, 4, 8, & 11, friends subscale items 6, 7, 9, & 12. The values of each item are added up, and mean scores are computed. The mean scores can range from 1 to 7, with one being the lowest and seven being the highest. According to Zimet GD et al., 1988; mean scores ranging from 1 to 3 indicate low perceived social support, scores ranging from greater than 3 to 5 indicate moderate perceived social support, and scores ranging from greater than 5 to 7 indicate high perceived social support [49]. Given the significance of perceived low social support for patients with mental illnesses, we dichotomised social support into low and medium/high for some of the analysis.

Self-Esteem

Self-esteem was measured using the single-item self-esteem scale, which measures overall self-esteem. Participants rate the single item on a five-point Likert scale ranging from 1 (not very true of me) to 5 (very true of me). Although shorter than the Rosenberg Self-Esteem Scale, the scale exhibits good convergent and comparable predictive validity [50].

Medication adherence

The medication adherence scale (MARS-5) assesses patients’ standard treatment adherence through five questions and five-level response formats (1 = always, 2 = often, 3 = occasionally, 4 = rarely, and 5 = never). Responses are added for a total score ranging from 5 to 25, with higher scores indicating greater adherence. They use the MARS-5 at a cut-off point greater than or equal to 20 [51, 52].

Substance use

Clients who used alcohol, tobacco, khat and cannabis (for nonmedical purposes) in the last year before data collection were considered current substance users [53].

Severity of illness

The Clinical Global Impression (CGI) scale, a subjective and objective measurement, was used to assess the severity of the illness. Responses 1–3 indicate mild, responses four indicate moderate, and responses 5–7 indicate severe on the CGI scale [54].

Suicidal ideation/attempt

Suicidal ideation and attempts were evaluated through “Yes” or “No” questions, with responses coded as 1 for “Yes” and 0 for “No.” These questions were adapted from the World Mental Health Survey Initiative’s module of the WHO’s Composite International Diagnostic Interview (CIDI), which has been validated for use in both clinical and community settings in Ethiopia [55].

Family relationships

Family relationships were measured as your relationship with your family—excellent, very good, good, fair, or poor.

Family participation in patient care

refers to instances where family members are actively involved, such as picking up medications during follow-ups, accompanying the patient to appointments, or communicating with healthcare providers about the patient’s care. This was measured using a single question, with a response of Yes/No.

Episode

An episode refers to a period during which a person experiences symptoms of a mental illness. It is often used to describe an acute phase of a disorder, such as a depressive episode, manic episode, or psychotic episode in psychiatric conditions. Eepisodic illness occurs if it happens more than once/year.

Relapse

occurs when a person who has been in recovery or remission experiences a return of symptoms. We assessed with single question have you had a history of illness relapse? Yes/No.

Statistical analyses

Low perceived social support was the dependent variable of this study. The independent variables were sociodemographic characteristics (sex, age, ethnicity, marital status, educational status, religion, job of participant, residence), clinical factors (type of mental illness, age of onset of illness, duration of illness, number of episodes of disease per year, drug side effect, objective severity of illness, number of episodes, hospital admission, waiting time in clinics), family relationships (family not participating in patient care, relationship with family) and substance use factors (alcohol use, tobacco use, khat use, cannabis use, family history of substance use) and other factors (legal issues, self-esteem).

This study followed a systematic approach to data management and analysis. After ensuring the accuracy and consistency of the data, this was meticulously checked, cleaned, coded, and then entered to the statistical software SPSS-28 for data analysis. The primary focus of this research was to assess the prevalence of low perceived social support, which was designated as the dependent variable and numbers of associated factors.

The analysis consisted of several steps. Initially, bivariate logistic regression was used to evaluate the relationships between independent and outcome variables. Variables with p-values less than or equal to 0.2 in the bivariate analysis were selected for further investigation to control for potential confounding effects but to allow for all potentially significant variables to be included the analysis [56]. These selected variables were then included in a multivariate logistic regression analysis.

In addition to logistic regression, descriptive statistics and other relevant analyses were conducted to understand the data comprehensively. To measure the strength of associations between the dependent and independent variables, bivariate analysis utilized adjusted odds ratios with a 95% confidence interval. Finally, we used a p-value of ≤ 0.05 and 95% CI were employed to see statistical significance.

Results

Background findings

A total of 638 patients with mental illness were included as participants two individuals consented but did not complete the questionnaires. The participants had a mean age of 35.5 ± 11.7 years, and slightly over half were female (51%). The mental illness diagnoses among participants included schizophrenia for 43%, depression for 30.2%, bipolar disorder for 7.9%, anxiety disorder for 6.1%, other psychotic disorders for 10.7%, stress/trauma-related disorder for 0.6%, and somatization disorder for 1.1%. Particularly, antipsychotic medications were taken by 47.3% of the participants, while 21.2% took both antipsychotics and antidepressants. Furthermore, 31.3% of all participants had been under treatment for more than five years, and 49.7% reported low self-esteem, while 13.7% exhibited poor drug adherence (Table 1).

Table 1.

Sociodemographic clinical and substance use characteristics of people with mental illness in an outpatient clinic, 2023 (n = 636)

Variables	Categories	Frequency (n = 636)	Percept (%)
Age	Mean ± Sd	35.5 ± 11.7
Sex	Male Female	317 319	49.8 50.2
Marital status	Married Not married.	253 383	39.8 60.2
Living condition	Living with family Living alone	541 95	85.1 14.9
Level of education	Literate Illiterate	451 185	70.9 29.1
Job of participant	Employed Not employed	89 547	14 86
Residence	Rural Urban	205 431	32.2 67.8
Mental illness	Other disorders Schizophrenia	362 274	56.9 43.1
Age of onset illness	≤ 25yrs. > 25yrs.	287 349	45.1 54.9
Duration of illness	> 5yrs. 6months-5yrs.	235 401	36.9 63.1
No, of episode/yr.	No episode Episode	298 338	46.9 53.1
Hospital admission /year	Yes No	209 427	32.9 67.1
Number of Admission	No admission 1 Admission ≥ 2 Admission	427 121 88	67.1 19 13.9
Comorbid illness	Yes No	77 559	12.1 87.9
Duration of Rx.	≤ 5 year. > 5 year.	437 199	68.7 31.3
Relapse	Yes No	412 224	64.8 35.2
Suicidal ideation	Yes No	250 386	39.3 60.7
Suicidal attempt	Yes No	126 510	19.8 80.2
Family Hx. MI	Yes No	144 492	22.6 77.4
Family Hx. Subs.	Yes No	115 521	18.1 81.9
Waiting time	30 min–1 h. 2–3 h.	534 102	84 16
Family participates in care	Yes No	536 100	84.3 15.7
Objective severity	Mild Moderate & Severe	493 143	77.5 22.5
Subjective Severity	Mild Moderate & Severe	424 212	66.7 33.3
Tobacco Use	Yes No	43 593	6.8 93.2
Alcohol Use	Yes No	141 495	22.2 77.8
Khat use	Yes No	78 558	12.3 87.7
Cannabis Use	Yes No	04 632	0.6 99.4
Self-esteem	Low self-esteem High self-esteem	316 320	49.7 50.3
Drug adherence	Poor adherence Good adherence	87 549	13.7 86.3

Prevalence of low perceived social support among people with mental illness in Gondar, Ethiopia

Among the 636 included participants, 77 individuals (12.1%) reported low perceived social support, while 271 (42.6%) reported moderate perceived social support and 288 (45.3%) indicated high levels of social support. Additionally, the breakdown of social support sources shows that a significant portion of people with mental illness experienced low social support from significant others (109 participants, 17.1%), family (54 participants, 8.5%), and friends (121 participants, 19%) (Fig. 1).

Fig. 1.

Source of Low perceived social support among people with mental illness in psychiatry outpatient in Gondar, Ethiopia, 2023 (n = 636)

Factors associated with low perceived social support

During the final bivariate logistic regression analysis for perceived social support, factors which had a p-value of less than or equal to 0.2 were identified. These were: (i) sociodemographic characteristics (living condition of participants, residence), (ii) clinical factors (duration of illness, type of drug, duration of treatment, relapse, suicidal ideation, suicidal attempt, family history of mental illness, objective severity of illness, subjective severity of illness, adherence), (iii) family relationship factors (relationship with family, family participating in patient care), and, (iv) others factors (legal issues, self-esteem) were fitted to multivariate logistic regression.

In the final model, living alone, having a relapse of illness, family not participating in patient care, moderate and severe objective severity of illness, low self-esteem and poor drug adherence were significantly associated with low perceived social support at p-value ≤ 0.05 and 95% CI (Table 2).

Table 2.

Multivariate Logistic regression on sociodemographic, clinical and substance use related factors on Low perceived social support among people with mental illness in Gondar, Ethiopia, 2023(n = 636)

Variables	Categories	Social support		COR (95% CI)	AOR (95% CI)	P-value
		High	Low
Age	Mean ± Sd	35.5 ± 11.7
Living condition	Living with family Living alone	487 67	54 28	1 3.76 (2.23,6.35) *	1 2.40 (1.24,4.63) **	0.009
Residence	Rural Urban	173 381	32 50	1 1.40 (0.87,2.27)	1 1.32 (0.75,2.32)
Duration of illness	> 5yrs. 6month-5yrs.	198 356	37 45	1 0.67 (0.42,1.08)	1 0.83 (0.47,1.46)
Type of drug	Other drugs Antipsychotic	286 268	49 33	1 0.71 (0.44,1.15)	1 0.81 (0.46,1.45)
Duration of Rx.	≤ 5yrs. > 5 year.	388 166	49 33	1 1.57 (0.97,2.53)	1 0.79 (0.44,1.42)
Relapse	No Yes	208 346	16 66	1 2.48 (1.39,4.39) *	1 2.13(1.13,4.02) **	0.019
Suicidal ideation	No Yes	348 206	38 44	1 1.95 (1.22,3.12) *	1 1.51 (0.88,2.60)
Suicidal attempt	No Yes	454 100	56 26	1 2.10 (1.26,3.52) *	1 1.24 (0.60,2.56)
Family Hx. MI	No Yes	424 130	68 14	1 0.67 (0.36,1.23)	1 0.66 (0.33,1.32)
R/ship with family	Excellent Very Good Good Fair Poor	46 127 265 92 24	03 20 34 08 17	1 2.41 (0.68,8.50) 1.96 (0.58,6.67) 1.33 (0.33,5.26) 10.86 (2.89,40.77) *	1 2.58 (0.68,9.82) 2.17 (0.59,7.97) 0.83 (0.19,3.60) 3.55 (0.83,15.22)
Family participates in the treatment	Yes No	491 63	45 37	1 6.40 (3.85,10.64) *	1 4.67(2.49,8.76) **	0.001
Legal issues	Yes No	26 528	01 81	3.98 (0.53,29.79) * 1	0.13 (0.01,1.07) 1
Objective severity	Mild Moderate & sever	444 110	49 33	1 2.71 (1.66,4.42) *	1 2.51(1.41,4.45) **	0.002
Subjective Severity	Mild Moderate & sever	378 176	46 36	1 1.68 (1.04,2.69) *	1 1.20 (0.66,2.20)
Self-esteem	High self-esteem Low self-esteem	295 259	25 57	1 2.59 (1.57,4.27) *	1 2.36(1.34,4.15) **	< 0.003
Drug adherence	Good adherence Poor adherence	486 68	63 19	1 2.15 (1.21,3.82) *	1 1.99(1.02,3.89) **	< 0.044

** Variables that have a significant association with perceived social support

The results indicated that patients with mental illness who lived alone were 2.40 times (AOR = 2.4, 95% CI; 1.24,4.63) more likely to report low social support than patients living with their family.

People with mental illness who had relapsed illness within the last year were 2.13 times more likely (AOR = 2.13, 95% CI;1.13,4.02) to report low social support than patients who did not have relapsed illness.

People with family who did not participate in their care were 4.67 times odds more (AOR = 4.67, 95% CI; 2.49,8.76) likely to report low social support than patient who did have participant family in their care.

People with moderate and severe objective severity of illness were 2.51 times (AOR = 2.51,95% CI;1.41,4.45) more likely to report low social support than people who have mild objective severity of mental illness.

People with low self-esteem were 2.36 (AOR = 2.36, 95% CI;1.34,4.15) more likely to report low social support when compared to patient with high self-esteem.

People with poor drug adherence were almost two times more (AOR = 1.99, 95% CI;1.02,3.89) likely to report poor social support than patients who had good drug adherence.

Discussion

Perceived social support from family, friends, and significant others varies depending on several factors, such as the nature of the relationships, the level of emotional closeness, and the type of support provided. Family provides unconditional and long-term support, while friends offer companionship and emotional backing. Significant others, including partners and spouses, contribute emotional intimacy, companionship and at times financial assistance. This high prevalence of low perceived social support more common from friends and significant other. People with mental illness often report receiving less social support from friends and significant others than from family members. This difference can be attributed to several factors, including misconceptions about mental illness that may cause friends and significant others to distance themselves. Furthermore, those facing mental health challenges often lean more on their families, who typically feel a sense of responsibility that friends and significant others may not share. As a result, this dynamic creates a perceived gap in support from non family relationships.

The current study found that 12.1% of participants reported low perceived social support, 42.6% reported moderate perceived social support, 45.3% indicated high levels of perceived social support. These findings are consistent with results from an Ethiopian study and other studies among patients with mental illness [10, 33, 57] that indicate a considerable number of people with low levels of social support.

The current study found that the greatest number of people reported low perceived social support from friends (19%) and significant others (17.1%), with 8.5% of people reporting low perceived social support from family. Goldberg et al. [58], in the USA, looked at social support among patients with mental illnesses and reported similar findings that the most frequent source of support was their families. A study from Sweden among patients who were receiving inpatient treatment and receiving care in the community showed that social support obtained from family members was higher in percentage than support received from non-family members [59]. The reasons why social support from family might be higher than that from friends might be multifactorial. As a result of their common value cultural norms and enduring bonds family members typically have stronger emotional ties. They also frequently have a strong sense of duty to look out for one another particularly in times of illness or need. Family members often also reside close by or are frequently physically present which makes it simpler to provide social support. In many cultures family members are the main persons to turn for assistance more generally [60]. In Ethiopia family bonds are strong and mainly serve as social support instead of a social burden which can be beneficial [61, 62].

However, contrary to what we found, Egyptian psychiatry patients reported the highest levels of social support was from their significant others (who could be any unique person in the patient’s life such as their boyfriend or girlfriend or a caregiver) [19]. Support from a significant other, such as a spouse, is often high for several reasons [63]. This may be because relationships with significant others tend to involve deeper emotional bonds, and they are often in committed relationships, fostering a greater sense of responsibility and care [64]. These factors make significant others a main source of support, often topping the help provided by friends, colleagues, or even family members in some cases. While 17% of participants in the current study reported low social support from significant others, the majority of participants were happy with the support they received from these key people in their lives. We could not identify any comparable studies from Ethiopia that separately examined types of support.

The current study found that several factors were associated with general social support. These included that those living alone were over twice as likely to have low social support when compared with patients residing with family. This is not surprising given that those who live alone are more likely to have less regular contact regarding the structural aspects of social support. A similar study supports this finding that the type of family that one lives in influences perceived social support in Nepal [65] and a study in South Africa with schizophrenia suggests that social support is crucial for the recovery [34]. When living alone, people with mental illnesses may not be able to get support with everyday duties. Patients with mental health problems like depression or anxiety might isolate themselves from others. They may experience heightened feelings of vulnerability, making them reluctant to ask for support from friends or family, due to fear of being stigmatised or burdening them. This low social support for mental illness could result in worsening symptoms and a lower QoL [6, 66], even where relationships with family are strong or they live with them, family members may not be involved in managing their medical care. When families actively participate in their family members’ treatment, their social support system may be enhanced through improved communication, more understanding, and stronger emotional ties.

In this study, patients with frequent relapses were more than twice as likely to report low social support. This finding aligns with the association between relapse and poor social support found among people with mental illness in Amanuel Mental Specialised Hospital, Ethiopia, which found that those with poor support were more than three times likely to relapse the disease than those with excellent social support [43]. A study from Taiwan likewise found a substantial association between the family domain of social support and the remission status of a patient with schizophrenia, and patients on remission showed higher levels of social support [67]. This reduction in social support during mental health relapses can result from various factors. If relapses are chronic, people may feel overwhelmed or unable to provide ongoing support. Family members or close friends offering continuous care may experience burnout, leading to a gradual decrease in the support they provide as they feel incapable of maintaining the same level of support, and individuals themselves may also withdraw so as not to be a burden for others who give support [68–70].

It is believed that family participation in patient care can help change the healthcare system as well as contribute to improved quality of care for patients [71]. Patients with families who did not participate in patient care had 4.67 times greater odds of low perceived social support than patients with families who did participate in their care. This is not surprising, as families who do not participate in a person’s care are also less likely to provide broader support. In the context of mental illness, the perception of family support and participation in care may exhibit a bidirectional relationship. When patients perceive strong family support, they are more likely to involve their families in their care, leading to positive outcomes such as improved treatment adherence, reduced symptoms, and enhanced emotional well-being [72]. As families become more actively involved, patients feel even more supported, reinforcing this cycle of care. This bidirectional association illustrates how the patient’s perception of support both shapes and is shaped by family involvement, ultimately enhancing the management of mental health conditions and improving overall QoL [73].

The present study provides additional evidence concerning the severity of illness and perceived social support. Patients with objective severity, moderate and severe, were 2.5 times more likely to have low perceived social support than those with mild severity. This is supported by a study conducted that over 70% of older adults with severe mental illness in the USA did not receive the support they needed [74]. This may be because when the severity of mental illness increases, it can impact social support in several ways. For example, when severity increases, mood swings, paranoia, or delusions can happen, and it can be difficult for people to maintain healthy relationships, which may be an increased burden that lowers social support levels [75]. Additionally, moderate and severe illnesses may cause social withdrawal and a decrease in the ability to maintain relationships, which can lead to isolation and social exclusion. This might also be a result of the stigma associated with severe mental illness [76].

This study found that self-esteem was one of the predictors of low perceived social support, with those with low self-esteem 2.3 times more likely to have low perceived social support than those with high self-esteem. This is similar to other studies where high social support has been associated with high self-esteem, and patients with high self-esteem were found to have high social relationships with others [77]. This may be due to patients with higher self-esteem being more likely to approach others for help or support easily, and high self-esteem makes people feel more competent in social situations. Additionally, they may find it simpler to communicate their needs, feelings, and concerns to others. This open communication can strengthen relationships and raise the likelihood of receiving the proper support from others. Low self-esteem was more likely to report low social support compared to those with high self-esteem due to several interconnected psychological and social factors: Low self-esteem can make it difficult for people to ask for help or express their needs [78]. Low self-esteem can also make people more prone to being sensitive to rejection, and they may be less assertive [79]. This can make it difficult for them to seek out and maintain supportive relationships actively. All these factors combined can create a vicious cycle in which poor self-esteem weakens relationships and intensifies feelings of helplessness and loneliness. In general, there has been very limited research into the reciprocal relationships between self-esteem and social support, which have been studied rarely [80]There is no more available evidence on the association of self-esteem and social support, so this study may be used as a benchmark.

This study also showed that patients with poor drug adherence were two times more likely to have low social support than those with good drug adherence. This finding is parallel with a study done in China on different areas on patients with severe mental illness that shows statistically significant positive associations between poor drug adherence and low social support [81, 82]. Poor drug adherence can lead to social isolation and reduced perceptions of support [83]. While most studies focus on how low perceived social support influences poor drug adherence, some research also suggests that poor drug adherence itself can predict low perceived social support among people with mental illness. Thus, a reciprocal relationship has been reported where poor drug adherence leads to reduced social support, which may further worsen adherence and social functioning [84]. This may be because poor social support undermines adherence by removing the emotional, practical, and psychological benefits that help patients stay engaged with their treatment. Patients with low social support are more likely to struggle with maintaining regular drug adherence.

Strengths and limitations of the study

This study used a standardised tool, the Multidimensional Scale of Perceived Social Support. While this tool has not been validated in Ethiopia, it has been in other parts of Africa and demonstrated excellent internal consistency in this study. Under supportive supervision, data were collected by psychiatry nurses with experience and training in data collection. this study’s second strength was its adequate sample size. However, due to the cross-sectional nature of the study design, the study is not able to show a temporal relationship, and mental illness can influence how patients perceive the support they receive. There may be a risk of social desirability bias because the survey was a facility-based cross-sectional study, and most of the data in the study were obtained through interviewer-administered questionnaires. The findings may not be generalisable to people with mental illness outside outpatient clinics, nor other outpatient patients in Ethiopia.

Conclusion and recommendations

More than 10% of patients in the outpatient psychiatry clinic had low perceived social support. These findings underscore the importance of understanding and addressing variations in social support within the context of mental health, as it can have significant implications for the well-being and recovery of people with mental illness. To improve health outcomes, people with mental illnesses need to have significant efforts to increase social support for those who receive outpatient care. Particular attention should be given to those with patients living alone, relapse, suicidal ideation, no family participation in care, moderate and severe objective severity of illness, low self-esteem, and poor drug adherence. To address these challenges, educating patients and their support networks, fostering mutual understanding, and establishing structured social support systems capable of sustaining them through follow-up periods is essential.

Abbreviations

AOR

Adjusted odds ratio

CI

Confidence interval

COR

Crude odds ratio

MARS-5

Medication adherence rate scale

MSPSS

Multidimensional perceived social support scale

CGI

Clinical Global Impression

OR

Odds ratio

SD

Standard deviation

SPSS-28

Statistical package for social science

Author contributions

Conceptualization: Wondale Getinet Alemu, Lillian Mwanri, Clemence Due, Telake Azalea, Anna Ziersch. Data curation: Wondale Getinet Alemu. Formal analysis: Wondale Getinet Alemu. Investigation: Wondale Getinet Alemu. Methodology: Wondale Getinet Alemu, Lillian Mwanri, Clemence Due, Telake Azale, Anna Ziersch. Supervision: Wondale Getinet Alemu, Lillian Mwanri, Clemence Due, Telake Azale, Anna Ziersch Writing original draft: Wondale Getinet AlemuReview & Editing: Wondale Getinet Alemu, Lillian Mwanri, Clemence Due, Telake Azale, Anna Ziersch.

Funding

The authors received no specific funding for this work.

Data availability

This is part of PhD work, and all data generated or analyzed during this study are included in this article.

Declarations

Ethics approval and consent to participate

Flinders University Human Research Ethics Committee approved with reference number 5416.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.