Child, adolescent, and caregiver mental health difficulties and associated risk factors early in the COVID-19 pandemic in South Africa

Participant sample

Two hundred and fifty-four parents and caregivers completed the online survey. The vast majority, 246 (96.9%), had access to outside space. The majority of the children were in the care of their biological parents, 234 (92.1%), while 6 (2.4%) were fostered and 2 (0.8%) were adopted. For further information on the demographic of the sample please see Table 3.

Table 3 Sample socio-demographic characteristics

This study also recorded mental health information for both the child and the parent/carer. These included whether the child had an SEN/ND and the type of SEN/ND as well as whether the child or the parent/carer had been diagnosed with any other mental health condition such as depression or anxiety, or diagnosed with a neurodevelopmental disorder such as autism spectrum disorder (ASD) or attention deficit hyperactivity disorder (ADHD). See Table 4 for more detail.

Table 4 Reported mental health status of children and their parents/carersThe Strengths and Difficulties Questionnaire (SDQ)Child gender and age

There were significant differences in mental health scores between children and adolescents, such that children had higher mental health scores on the SDQ scale than adolescents, F(1, 240) = 4.88, p = 0.028, η2p = 0.020 (see Fig. 1). There was no main effect of gender, F(1,240) = 1.89, p = 0.171, η2p = 0.008, and no interaction between child age and gender on mental health scores, F(1,240) = 0.511, p = 0.476, η2p = 0.002.

Fig. 1figure 1

The effects of age and gender on mental health

Emotional problems scale Children had higher scores for emotional problems than adolescents; however, this difference was not significant, F(1, 240) = 3.44, p = 0.065, η2p = 0.014. There was also no significant gender difference in emotional problems, F(1, 240) = 0.605, p = 0.437, η2p = 0.003, and no interaction between child age and child gender, F(1, 240) = 2.17, p = 0.142, η2p = 0.009.

Conduct problems scale There was a significant difference in scores on the conduct problems scale between children and adolescents, with higher conduct problems endorsed in children than adolescents, F(1, 240) = 4.748, p = 0.030, η2p = 0.019. There was no difference in conduct problem scores by gender, F(1, 240) = 2.071, p = 0.151, η2p = 0.008, and there was no interaction between child age and child gender, F(1, 250) = 0.002, p = 0.968, η2p = 0.000.

Hyperactivity scale There was a significant difference between children and adolescents. Children were reported to be more hyperactive than adolescents, F(1, 240) = 10.91, p = 0.001, η2p = 0.043. There was also a significant gender difference, with males being more hyperactive overall compared to females, F(1, 240) = 4.37, p = 0.038, η2p = 0.018. There was no interaction between age and gender, F(1, 240) = 0.74, p = 0.391, η2p = 0.003.

Peer problems scale There was no significant difference in peer problems reported between children and adolescents, F(1, 240) = 2.68, p = 0.103, η2p = 0.011, or between males and females, F(1, 240) = 1.58, p = 0.209, η2p = 0.007. There was also no interaction between child age and child gender, F(1, 240) = 2.81, p = 0.095, η2p = 0.012.

Prosocial scale There was no main effect for child age. Both children and adolescents reportedly displayed high prosocial behaviour, F(1, 240) = 1.45, p = 0.230, η2p = 0.006. There was, however, a significant main effect for gender, with females displaying more prosocial behaviours than males, F(1, 240) = 56.15, p = 0.014, η2p = 0.025. Finally, there was no interaction between age and gender, F(1, 240) = 0.88, p = 0.350, η2p = 0.00. For further details on SDQ scores across child gender and age, See Appendix 1.

Household income

A 2 (age: child vs adolescent) × 3 (income: low vs middle vs high) between subjects factorial ANOVA revealed no main effects for any of the variables. This suggests that neither low, middle or high income made any difference to children’s or adolescent’s mental health, all p’s > 0.05. There was also no significant interaction observed between the variables. The same was observed for the SDQ subscales. No main effects were observed for the emotional, conduct, hyperactivity, peer problems or prosocial scales, and there were no significant interactions, all p’s > 0.05. This suggests that the children’s mental health as reported by their parents/carers was not affected by their age, nor total household income. Please see Appendix 2 for the ANOVA results.

Family composition

There was no difference in mental health scores between children raised in single adult households and those with more than one adult in the household (p = 0.183). As reported by parents/carers, children raised in single adult households exhibited fewer emotional problems than those raised by more than one adult (p = 0.021), and these children also displayed more prosocial behaviours than those raised by more than one adult (p = 0.010). There was no difference in conduct problems, hyperactivity levels, or peer problems between those children raised in single parent households compared to those raised by more than one adult (all p’s > 0.05, see Table 5 for more detail).

Table 5 SDQ scores by family compositionPresence of SENs/NDs

Due to the small number of children and adolescents with SENs or NDs, we were not able to stratify by age. Participants with SENs/NDs had significantly higher mental health scores than those without (p < 0.001).

The effect of SENs/NDs (present or not) on mental health scores were examined more closely across the five SDQ subscales. As reported by parents and carers, youth with SENs/NDs had higher emotional mental health scores (p = 0.003), were reported to be more hyperactive (p < 0.001), and experienced more peer problems (p = 0.038) than those without SENs/NDs. Finally, there was no difference in misconduct scores (p = 0.052), or prosocial behaviour between participants with and without SENs/NDs (p = 0.068, see Table 6 for more detail).

Table 6 SDQ scores of youth with and without SEN/NDThe Depression Anxiety Stress Scales (DASS)Child age

There were no differences in depression scores (p = 0.145), or anxiety levels (p = 0.133) between parents/carers who reported on children compared to those reporting on adolescents. However, parents/carers who reported on a child experienced significantly higher levels of stress compared to those who reported on an adolescent (p = 0.009). See Table 7 for more details.

Table 7 Parent/carer DASS scores by child age, family composition and SEN/ND present or not present in the childHousehold income

There were no differences in levels of depression (F(2,211) = 0.624, p = 0.537, η2p = 0.005), anxiety (F(2,211) = 1.154, p = 0.318, η2p = 0.01), or stress levels (F(2,211) = 1.34, p = 0.264, η2p = 0.013 across low, middle or high income groups. See Table 8 for descriptive statistics.

Table 8 One-way ANOVA comparing parent/carer DASS scores by household incomeFamily composition

There were no statistically significant differences in levels of depression (p = 0.290), anxiety (p = 0.163) or stress (p = 0.804) between single parents and those who had assistance raising their families. See Table 7 for more details.

Presence of SENs/NDs

The following section examines how having a child with an SEN/ND affects parent/carer mental health. No differences were found in depression (p = 0.259), anxiety (p = 0.327) or stress (p = 0.091) levels between parents/carers who had children with SENs/NDs and those who did not. See Table 7 for more details.

Correlations between SDQ and DASS scores

Moderate, positive correlations were found between total child/adolescent SDQ scores and parent/carer DASS scores for depression (r(244) = 0.37, p < 0.001), anxiety (r(244) = 0.35, p < 0.001), and stress (r(244) = 0.45, p < 0.001). Parents/carers who experienced higher levels of depression, anxiety and stress also reported higher mental health problems overall in their children/adolescents.

留言 (0)

沒有登入
gif