Serious mental illness (SMI) is a mental disorder which impairs functioning in social or occupational roles.¹ Women with SMI may fill a number of these roles, including motherhood.² Children of mothers with SMI may have poor school performance and impaired social skills, and are at an increased risk of developing mental illness.³ Children living with parents with mental illness are at greater risk of psychological and behavioural problems, and poor coping mechanisms, with a 2021 study showing children of women with schizophrenia were more likely to have disruptive disorders than their peers.^4,5 Children of mothers who have a mental illness are less likely to perform well academically.⁶ For mothers with SMI, there is a fear of loss of custody of their children.³ Research on mental illness in women with SMI tends to focus on the peripartum period;⁷ however, interventions (including parenting skills training) for mothers with SMI throughout their lifespan have been shown to improve parenting and are associated with better outcomes for these mothers and their children.⁵

The fertility rate of women with SMI is comparable to that of women without SMI, but women with SMI are more likely to have unplanned pregnancies and may mother an above-average number of children.² Studies examining prevalence rates of motherhood in SMI have predominantly been conducted in developed nations on convenience samples. A 2011 New York-based study in a psychiatric hospital found that 38.5% of female psychiatric inpatients with SMI were mothers.⁸ A representative, but older, study conducted in 2002 in the United States (US) found the prevalence of motherhood in women with SMI to be 67.2%.⁹ Studies from community outpatient departments in Australia and London showed similar prevalence rates to those in the American representative study, at 59% and 63% respectively.^10,11 In the US-based ‘MOMs’ study on a non-representative sample of 379 women from mental health community healthcare centres and hospitals, most women had three children. However, only 33% of women with SMI were caregivers to their children.² In Canada, 50% of mothers with SMI lived with their children, while in Australia, approximately 60% of children whose parents had SMI lived with their parents.^12,13 The 2020 fertility rate in SA was estimated at 2.4 children per woman, with the frequency of motherhood in women with SMI unknown, as is the percentage of these women who are primary caregivers to their children.¹⁴

Women with SMI face difficulties such as poverty, unemployment, poor social support, stigma, residual mood or psychotic symptoms of their SMI and side effects of psychotropic medication.² Mothers with SMI have additional stressors which have largely been identified through qualitative studies.¹⁵ In a UK-based qualitative study, women with SMI identified their role as mothers as central to their recovery and a strong motivator to prevent relapse, while another woman with a psychotic disorder in the UK described worsening alcohol use because of the pressures of the parenting role.^16,17 Women with SMI in a focus group in a Massachusetts community clinic felt stigmatised for being mothers, and worried about their children being removed from their care. One mother described severe emotional pain as a result of this happening.¹⁸ Mothers perceived that the removal of children from their custody was the primary parental objective of social services.¹⁷ Suggestions from the mothers themselves for interventions included parenting skills and support groups. One South African study (based in Gauteng) described the experiences of 10 mothers living with chronic mental illness who were receiving outpatient psychiatric services at the time of interview.¹⁹ Like the women in studies from high-income countries, they expressed a wish for family support and training in parenting skills.¹⁹

Children in Southern Africa are already vulnerable, facing challenges such as malnutrition and human immunodeficiency virus (HIV).²⁰ Children of mothers with SMI are even more vulnerable, with a recent study showing children of mothers with schizophrenia are more likely to have complex chronic conditions, including respiratory and cardiac disease.²¹ A South African qualitative study highlighting unemployment in these mothers, preventing them from meeting their children’s basic needs.¹⁹ A study from New Zealand revealed the invisibility of children of parents with mental illness. In this mixed-method study, less than half the records of parents with SMI included information on their children’s well-being.²² Children of mothers with SMI require attention as they are at risk of intergenerational transfer of mental illness, as well as academic and behavioural issues.^4,7

Providing support and interventions for mothers with mental illness improves outcomes for both mothers with SMI and their children.⁷ Almost all patients in one qualitative study felt that services could be improved to help them with parenting.¹⁷ A review article by Seeman et al.¹² identified a number of interventions to prevent custody loss in women with schizophrenia. These include: relapse prevention, planning in advance for admission, and attending parenting skills courses.¹² Where delusions may affect a woman’s ability to mother, treating these may improve her mothering ability.²³ Importantly, a review and meta-analysis proved that providing appropriate support for mothers with SMI could lessen poor outcomes for children of parents with mental illness.²⁴ Interventions cited included psycho-education of parents regarding parenting skills, and adolescent children regarding their parents’ mental illness. Mowbray et al. also identified parental support groups, as well as financial management of household expenditure, as effective interventions.²

This study is a retrospective quantitative review of files of female psychiatric inpatients at Tara H. Moross Hospital, which is a specialist psychiatric hospital in Johannesburg, South Africa. This hospital has a total of 140 beds for psychiatric inpatients and receives patients from a large catchment area. Approximately 120 adult female patients are admitted per year with SMI who require medium-term care. For the purpose of this study, SMI is defined as having a mental disorder with associated functional impairment.¹

Female inpatients admitted to the biological wards at Tara H. Moross Hospital with an SMI between 1 July 2021 and 30 June 2022 were included in this study. No patients less than 18 years old were included in this study, as patients included in this study were admitted to the adult biological wards at Tara H. Moross Hospital. For the purpose of this study, a mother was defined as a patient who self-reported being a mother in interviews recorded in clinical notes.¹⁵ In the clinical files of three patients, motherhood status was not recorded. These patients were excluded from data analysis.

The following information was captured for all female admissions: age; highest level of education; relationship status; source of income; primary DSM-5 diagnosis on discharge;²⁶ number of previous psychiatric admissions; medical comorbidities; substance use (clinical history of use before admission or a documented substance disorder); history of non-adherence (documented directly in clinical notes, or listed as a v-code); family planning; length of admission and discharge location. For mothers with SMI, the following were captured: maternal age at birth of first child; number of children; primary residence of minor children (<18 years old) before the mother’s admission and reason for referral to social services.

Motherhood status was recorded in 118 of 121 files. Approximately 63% (N = 74) of the female patients in the 1-year study period self-identified as mothers. Demographic data of mothers and non-mothers are summarised in Table 1. There was no significant difference in the relationship status, education level or employment status between mothers and non-mothers. Nearly half of mothers (N = 35) were single, with 28% in relationships at the time of admission. Employment levels in the sample were low, with only 9% of mothers (N = 7) and 14% (N = 6) of non-mothers being employed (p = 0.508).

Table 2 summarises the admission characteristics of mothers and non-mothers. The admission length was not associated with motherhood status, with mean admission lengths for mothers at 107 days, and 89 days for non-mothers (p = 0.151). Mothers who lived with their minor children had a mean admission length of 83 days, while those mothers who did not live with their minor children had a mean admission length of 119 days (p = 0.056). Of mothers who had been previously admitted (N = 58), 84% reported non-adherence to treatment before the studied admission. The most common diagnosis for the sample was psychotic disorders (47%), followed by bipolar and related disorders (36%). There were no significant differences in the diagnoses received by mothers and non-mothers (p = 0.0585). The diagnosis of major neurocognitive disorder (MNCD) was the primary diagnosis in 11% of mothers, while no non-mothers received this diagnosis. The mean age of women with a diagnosis of MNCD was 57 ± 13 years. The sample size precluded the use of logistic regression to determine potential confounding variables in MNCD being more common in mothers compared with non-mothers.

Mothers were more likely than non-mothers to have a medical comorbidity (p = 0.002). The burden of medical comorbidities in mothers was 80% (N = 59), compared with only 52% (N = 23) in non-mothers. Human immunodeficiency virus was present in 27% of the total sample. More mothers had HIV (39%, N = 29), compared with non-mothers (7%, N = 3) (p = 0.0001). The average age of women with HIV was 39 years ± 11. The second most common comorbidity was hypertension, which was diagnosed in 34% (N = 25) of mothers compared with 18% (N = 8) of non-mothers (p = 0.068). The mean age of women with the diagnosis of hypertension was 52 ±10. In the overall sample, 33% (N = 41) of patients had documented substance or alcohol use, with no difference in use between mothers (34%, N = 25) and non-mothers (32%, N = 14) (p = 0.235). In mothers, 7% (N = 5) used cannabis, 9% (N = 7) used alcohol and 16% (N = 12) used more than one substance.

Post-discharge, 88% (N = 65) of mothers returned home, with mothers and non-mothers equally likely to be referred to placement facilities. (p = 0.072).

The frequency of motherhood in this sample is 63%, higher than a study conducted in New York with a reported frequency of 38.5%. Both were convenience samples and not generalisable, but the fertility rate in SA is higher than that in the US, and this may contribute to the differences noted in similar inpatient populations. The TFR of women with SMI in our study was markedly lower at 1.15 children per woman, compared with the general South African population, which was 2.34–2.35 children per woman between 2020 and 2021, the same time period as the study was conducted.³⁰ The lower fertility rate may be partly explained by women with SMI having half as many live births as those without SMI.³¹

The high unemployment rate in this study may highlight the occupational impairment in those with SMI, as the unemployment rate of females in the South African general population over the same period was between 35.5%-37.3%.³² Mothers in our study had a 9% employment rate – comparable with mothers with SMI in Michigan, US, with an 11.3% full-time employment rate.² Only 34% of the total sample completed school, lower than the 66.9% of the general female population in the 2022 South African census data.³³ A quarter of mothers in this sample had their first child at age 19 or less, and only 43% of mothers were currently involved in the care of their children, which does not allow for childcare alone to be a reason for not completing school; however, pregnancy during high school may account for school dropout.³⁴ The age of onset of psychiatric illness was not captured, nor was whether mothers cared for their children during infancy. These all limit the interpretation of the lower level of education of mothers in our sample, compared with the general female population in South Africa.

Nearly half of mothers (47%, N = 35) were single, similar to the 58.3% of mothers with SMI in an American study who identified as single.⁸ The South African 2022 Census data showed that 31.7% of the female population were either married or living with a partner, which is similar to the 26% of women in our study identifying as married or in a relationship.³⁵

The most common SMI diagnosed among mothers and non-mothers was psychotic disorders (47%), followed by bipolar and related disorders (36%). In 11% of mothers with SMI, the primary diagnosis was MNCD, while no non-mothers had this diagnosis. Logistic regression was not possible because of the sample size, but the mean age of women diagnosed with MNCD was 57 years old. The age of the individuals with this diagnosis is a more likely explanation for this statistically significant difference, a likely confounding variable in this case. Mothers in this sample were also more likely than non-mothers to have HIV, which increases the risk of MNCD, and this could also contribute to MNCD being diagnosed in mothers.³⁶

The high HIV rate in mothers compared with non-mothers is likely related to the same risk factors being associated with HIV and pregnancy. The HIV prevalence in mothers is similar to that seen in women attending antenatal care in South Africa, which is estimated at approximately 44.3%.³⁷ High burden of medical comorbidities in mothers may further impact adherence, and HIV itself can have cognitive effects on the brain that could further worsen SMI in these mothers.³⁸

The frequency of substance use in this study was 33% – similar to the 41% prevalence of substance use in patients with bipolar and schizophrenia globally.³⁹ Substance use in mothers is associated with increased risk of substance use in the next generation.⁴⁰ Substance use may also precipitate episodes of mental illness in these women and further impair their functioning as mothers. Addressing substance use in this population may improve outcomes for this patient population, including those who are mothers and their children.

There was no association between motherhood and treatment adherence. A previous qualitative study indicated that for some mothers, motherhood improved adherence, while for other mothers the medication impaired their ability to mother.¹⁷

In 25% of cases in this study, mothers were reported to be currently caring for their minor children alone, meaning that, during the average 107-day admission, many minor children were without their primary caregiver. The absent father in the lives of the children of mothers with SMI is also highlighted in this setting, with only 9% of children living with both parents. A study on the general South African population showed that 45% of minor children are cared for by their mother, with 27% of minor children cared for by their grandmother.⁴¹ In this study, 43% of minor children of women with SMI were cared for by extended family, showing the significant role extended family has in the care of these children. This large difference can partly be explained by the impaired functioning women with SMI may have in their parental role, but also may highlight some of the structural differences in families within the South African setting compared to research findings in some high-income countries. For example, a study conducted in the US indicated that only 13.3% of children of mothers with SMI lived with extended family.² No documentation of the primary caregiver of minor children (in 7% of cases) may highlight the lack of attention directed by adult mental health practitioners to these vulnerable children.

Family planning was not documented in 19% of the mothers of reproductive age in this study, while 42% of mothers were on some form of contraceptives. Rates of unplanned pregnancies are higher in women with SMI than in the general population, and there are additional risks for women with SMI, such as teratogenicity associated with certain psychotropics.² Unplanned pregnancies are associated with antenatal and postpartum depression.⁴² Interventions targeting family planning could assist with reduced risk and improved mental health among mothers and their children.

Mothers in our study had their first child at a median age of 22 years, which is similar to the 21.79-year average age of motherhood in an American study on mothers with SMI.² In the latest census data, mothers in South Africa had their first children between the ages of 18 years and 20 years, similar to the population of mothers in our study.³⁰

Mothers were discharged home in 88% of cases. Providing appropriate parenting interventions has been shown to improve parental outcomes, yet no referrals were documented as being made for parenting skills or assistance in this study.²⁴ Mothers in previous qualitative studies believed the primary role of social services to be the removal of children from their care.¹⁷ In this study, no mothers were referred to social services for the primary reason of the removal of children. The average admission length of mothers with minor children was 83 days, with social services asked to review the safety of children in only 12% of cases. This is despite evidence suggesting that children of mothers with SMI are extremely vulnerable.²²

This study had several limitations, including the retrospective study design, which relied on details captured in files, which may have omitted clinical information. The study also used a convenience sample, and findings cannot be extrapolated to the general population. The sample size also limited the ability to determine confounding variables in statistical analysis. Further research could focus on assessing the effectiveness of interventions, such as culturally appropriate parental counselling, on the mental well-being of mothers with SMI and their children.⁴³

The mothers in this study represent a unique population – they have SMI, often with comorbid substance use, and physical illnesses, including HIV. Mothers were most often single and unemployed, with 25% of mothers caring for their minor children alone. In 43%, children were cared for by extended family. In providing support for these women, their children, and their extended families, providers should consider substance use interventions, HIV management, and recognise the role of the single, unemployed parent with SMI, within the extended family. Likewise, these factors should be considered in the further development of culturally sensitive tailored evidence-based interventions for mothers with SMI, to improve overall care for women with SMI and to mitigate the impact of parental SMI on the mental health and well-being of their children.