Discriminant analysis of caregivers' psychiatric symptoms according to offspring psychopathology.

Associations between parent or caregiver depression and adverse child outcomes are well established.1 We previously examined correlations of symptoms of common mental disorders in caregivers with offspring psychopathology in a Brazilian sample.2 Sixty-eight primary caregivers of 110 children (age 6-15 years) were enrolled. Caregivers were assessed using the Self-Reporting Questionnaire3 (SRQ-20), which measures symptoms of common mental disorders. We used the Strengths and Difficulties Questionnaire (SDQ) to measure children’s symptoms. In our previous results, higher SRQ-20 scores in caregivers correlated significantly with offspring psychiatric symptoms (beta = 0.20; p = 0.04).

We further investigated this result by conducting a discriminant analysis. A multinomial logistic regression with Bayesian estimation (Mplus, 7.4) was carried out to identify the symptoms assessed with the SRQ-20, discriminating three groups of children: 1) asymptomatic (SDQ < 14, impact supplement score = 0); 2) symptomatic without impact (SDQ ≥ 14, impact supplement score = 0); and 3) symptomatic with impact (SDQ ≥ 14, impact supplement score ≥ 1).

The results (Table 1) showed that caregivers of symptomatic children without impact reported lower depressed mood, lower somatic (anxiety) and somatic (gastrointestinal) symptoms, more weight loss, and lower insight compared to caregivers of asymptomatic children. Caregivers of symptomatic children with impact reported lower depressed mood, lower retardation, more agitation, lower somatic (gastrointestinal) symptoms, more weight loss, and less insight than caregivers of asymptomatic children. Comparison between the two groups of caregivers of symptomatic children showed that the group with impact reported higher levels of early insomnia, lower retardation, lower agitation, higher anxiety (psychological and somatic), more somatic symptoms, and less insight.

Table 1

Results of multinomial logistic Bayesian regression analysis

	SDQ- vs. SDQ+ without impact		SDQ- vs. SDQ+ with impact		SDQ+ without impact vs. SDQ+ with impact
	Beta	p-value	Beta	p-value	Beta	p-value
Depressed mood	-0.25	0.04	-0.32	0.03	-0.18	0.11
Feelings of guilt	0.25	0.09	0.18	0.09	0.20	0.06
Suicide	-0.19	0.15	-0.13	0.17	0.21	0.07
Insomnia, early	-0.30	0.09	0.09	0.32	0.46	0.00
Insomnia, middle	0.38	0.06	0.01	0.48	-0.08	0.30
Insomnia, late	-0.13	0.32	0.01	0.50	-0.15	0.19
Work and activities	0.04	0.43	0.07	0.34	-0.08	0.31
Retardation, psychomotor	0.01	0.47	-0.34	0.02	-0.22	0.03
Agitation	0.28	0.08	0.44	0.01	-0.40	0.01
Anxiety (psychological)	0.20	0.15	0.24	0.06	0.40	0.01
Anxiety (somatic)	-0.32	0.02	-0.13	0.21	0.32	0.01
Somatic symptoms (gastrointestinal)	-0.54	0.01	-0.40	0.03	0.27	0.09
Somatic symptoms (general)	-0.06	0.34	-0.04	0.38	0.27	0.01
Genital symptoms	-0.00	0.50	-0.08	0.30	-0.19	0.07
Hypochondriasis	0.07	0.35	0.00	0.50	-0.18	0.08
Weight loss	0.54	0.00	0.53	0.01	-0.21	0.09
Insight	-0.34	0.00	-0.26	0.02	0.20	0.04

SDQ = Strengths and Difficulties Questionnaire applied to children; SRQ-20 = Self Reporting Questionnaire-20 applied to adult caregivers; SDQ- = asymptomatic children (SDQ < 14); SDQ+ = symptomatic children (SDQ ≥ 14); impact = symptomatic children with impact supplement score ≥ 1.

Bold font indicates statistical significance.

These results possibly demonstrate a trend toward symptomatology interaction between caregivers and their offspring. Weissman et al. studied the differential effects of a depressed mother’s treatment on her child, and found that children whose mothers were on escitalopram showed significantly greater improvement in symptoms and functioning as compared to children whose mothers were on bupropion or a combination of both.4 The authors also observed that maternal baseline negative affectivity (which captures high levels of stress, irritability, and anxiety) appeared to moderate the effect of maternal treatment on children. Possibly, these mothers are better treated with escitalopram, which enhances serotoninergic neurotransmission, as compared to bupropion, which enhances dopaminergic transmission.

In another study, Morgan et al. evaluated how maternal neural response to child affect is related to depression by using an fMRI task.5 They found that comorbid anxiety, chronicity of depression, and poor mother-child relationship emerged as predictors of altered maternal neural response to child affect. Few studies have sought to elucidate the mechanisms of parental-offspring psychopathology.

Despite the role of anxiety in these previous studies, in our analysis, we found that agitation, less retardation, less depressed mood, less somatic symptoms, and more weight loss seem to characterize the caregivers of symptomatic versus asymptomatic children. When comparing only caregivers of symptomatic children, those caring for children with impact presented higher levels of anxiety, which is in line with the existing literature. Greater knowledge of mechanisms underlying caregiver-offspring interactions is needed to improve treatment strategies.

Disclosure

The authors report no conflicts of interest.