Depressive Symptoms in Parents of Children With Chronic Health Conditions

A Meta-analysis

Martin Pinquart, PHD


J Pediatr Psychol. 2019;44(2):139-149. 

In This Article


The present study reports the results of the first meta-analysis on depressive symptoms among parents of children with a broad range of chronic health conditions. Depressive symptoms were elevated by 0.46 SD units when compared with parents of healthy/nondisabled children and test norms. Twelve studies with structured clinical interviews provided a weighted mean depression rate of 20.9%—a rate that was 2.98 times higher than the rate in the general population (Steel et al., 2014). The largest elevations of depressive symptoms were observed in parents of children with neuromuscular disorders and cancer. In addition, we found greater elevations of parental depression if the child's condition has lasted for a shorter period at the time of the assessment, in mothers compared with fathers, and in parents from less economically developed countries rather than developed countries. The present results assist with identifying parents with the greatest need for depressive symptoms screenings.

When comparing the results of the present meta-analysis with the results of a previous meta-analysis on depressive symptoms in children and adolescents with chronic conditions (Pinquart, M. & Shen, Y. 2011), we find greater elevations of symptoms in the parents than in the children. This difference may indicate that parents are more depressed because they are more aware of the negative consequences of the health condition of the child than the child is (Aldridge, Shimmon, Miller, Fraser, & Wright, 2017), that parents experience more stressors than their children (because of high involvement in illness management; Naar-King et al., 2009), or that children tend to underreport their depressive symptoms or show repressive coping and avoid thoughts about threatening or distressing aspects of their health condition (Phipps & Srivastava, 1997).

Condition-specific elevations of depressive symptoms were similar to those reported by Easter et al. (2015) and Vermaes et al. (2005), although Pai et al. (2007) reported lower elevations of psychological distress in parents of children with cancer compared with the present meta-analysis. This difference may have been based on the use of different outcome measures (psychological distress in general vs. depressive symptoms in particular) or on the fact that Pai et al. (2007) included only studies from western, developed countries where elevations of depressive symptoms were lower than in the other countries.

We only found partial support for the suggestion that parental depressive symptoms would be highest if their child has a life-threatening or life-limiting condition, such as cystic fibrosis, HIV infection/AIDS, or progressive neuromuscular disorders. While parents of children with neuromuscular disorders and cancer showed moderate elevations of depressive symptoms, between-group differences were small in the case of cystic fibrosis, and even statistically nonsignificant in the case of HIV infection/AIDS. As most HIV-infected children had not yet developed AIDS, some sources of depressed mood (such as expecting to lose the child in the near future) were probably not yet present. Although cystic fibrosis is a life-limiting condition, survival rates have increased. While in 1985, patients were expected to live, on average, to the age of 25 years, the expected mean survival was 47.7 years for children born in 2016 (Cystic Fibrosis Foundation, 2017). Increased length of survival reduced one source of depressive symptoms in parents of children and adolescents with cystic fibrosis (Lowes & Lyne, 2000).

The present results indicate that the fact whether a disease is life-threatening or life-limiting may be less relevant for predicting parental depressive symptoms than other characteristics, such as expected length of survival, seeing the child suffering, or high caregiving demands (Lawoko & Soares, 2006). In addition, parental perceptions of life-threat often differ from the actual life-threat (with many parents being overly optimistic), and parental subjective perceptions of life-threat may be more relevant for parental mental health than the related objective life-threat (Mack, Cook, Wolfe, Grier, Cleary, & Weeks, 2007).

The highest between-group differences in depressive symptoms were found in parents of children with neuromuscular disorders, such as Duchenne muscular atrophy—a lethal disease that is characterized by progressive muscle degeneration and weakness, with death often occurring before the age of 20 years (Bothwell, Dooley, Gordon, MacAuley, Camfield, & MacSween, 2002). Depressive feelings in these parents may result from lack of control over the course of disease, chronic stress because of caregiving demands, perceptions of social isolation, and the near death of their child (Bothwell et al., 2002). The data on families with children with cleft lip and palate showed that some chronic conditions of children do not, on average, lead to elevated parental depressive symptoms. Treatment of cleft lip and palate is usually completed in early childhood, and the condition does no longer affect the daily life of most parents in the long run (Feragen et al., 2017).

The moderation effect of the duration of chronic conditions suggests that, in the case of nonprogressive diseases, caregiving demands decline over time and parents increasingly adapt to the chronic condition of their child which leads to lower levels of depressive symptoms. Lower between-group differences in depressive symptoms in the case of longer durations of the chronic condition may also be because of the fact that some chronic conditions tend to become less severe over time (e.g., asthma; Ko, Song, & Clark, 2014), or that young patients increasingly take responsibility for their own disease management (Naar-King et al., 2009). If the latter plays an important role for parental depression, we should also find lower between-group differences among parents with older children. However, this was not the case in the present meta-analysis.

The greater elevations of depressive symptoms observed in samples with higher percentages of women could be based on the higher involvement of mothers in the care of their ill children, mothers' stronger emotional responses to the disease of their children (Hyde et al., 2008), or gender differences in coping with stressors, such as using a ruminative coping style (Hyde et al., 2008). We did not find a moderator effect of parental marital status; this may have been based on the fact that fewer studies were available for testing a moderator effect as compared with the other moderator analyses, or on the restricted variance of this moderator variable, as no separate data were available for unmarried parents. The observed higher elevations of parental depression in less economically developed countries can be explained by the lower access to high-quality health care for the ill children (Telfer, 2009), as well as lower availability of high-quality support services for parents of children with chronic diseases (Witbooi, 2013). Finally, the lack of moderator effects of the aspects of the study quality and publication status indicates that our results were robust with regard to these criteria.