Printer friendly version
Does breastfeeding protect from depression in adulthood?
05 September 2012
Journal of Psychotherapy and Psychosomatics
A study published in the last issue of Psychotherapy and Psychosomatics examines the relationship between breastfeeding and onset of depression in adulthood.There is increasing evidence on long-term effects of mother-pup interaction in rodents and nonhuman primates, but observations on the effects of early life experiences on human adults’ health do not generally consider the period before 6 years of age. Adverse events beyond that age can be reported in retrospective studies and are related to lasting modulation of the HPA system and incidence of psychiatric disorders. In humans breast-feeding is a protective factor against several somatic disorders.
Regarding psychiatric disorders, early weaning is related to an increased risk of alcohol dependence, while there are conflicting results on the association with schizophrenia. So far, there is no information on breast-feeding and prospective affective disorder. Using a case control approach, a group of German researchers investigated the association of being breast-fed versus bottle-fed with depression later in life in humans. Female and male adult inpatients (n = 52) with a diagnosis of major depression (DSM-IV) were included in the study. Using questionnaires blood donors were screened for psychiatric disorders and eventual enrolment as never-depressed healthy controls (n =106). Twenty participants fulfilled the criteria for lifetime major depressive disorder, which led to their exclusion. Only participants who had been breast-fed for at least 2 weeks were categorized as breast-fed. In both groups 2 mothers reported having fed expressed milk, which
also led to exclusion from the analysis. To avoid bias we did not exclude individuals whose mothers were deceased, suffered from dementia or could not be contacted due to refusal by the enrolled individual. In these cases either the participant himself or another relative provided the respective information.
The investigators found that 61/84 (72.6%) never-depressed controls, but only 22/48 (45.8%) depressed patients had been breast-fed. The X 2 test (p<0.002) showed ‘no breast-feeding’ to be associated with depression in adulthood. The risk for major depression was considerably increased in bottle-fed compared to breast-fed subjects. Controlling for risk factors of depression and maternal education known to be related to breast-feeding, bottle-feeding (p = 0.005), but not age, gender and
maternal education ( NS) was found to be related to depression. Age at onset of the affective disorder did not differ between breast-fed and bottle-fed depressed patients, while age (p<0.03) and number of episodes were higher in breast-fed compared to non-breast-fed individuals (p<0.04). Controlling for gender, age and maternal education, the investigators found that duration of breast-feeding was not significantlyrelated to depression in adulthood (NS). This is the first report demonstrating that a history of not being breast-fed is associated with subsequent major depression. Due to the correlative nature of
these data, however, the Authors of this study could not conclude this to be a causal relationship.
First, breast-feeding could be an indicator for the quality of the mother-infant interaction with other aspects of the interaction being the protective mechanisms. Second, breast-feeding may boost oxytocin-induced affiliative behavior in the mother, which is regarded as stress-protective. Third, the composition of breast milk might be advantageous for brain development and thereby prevent depression. Fourth, breast-feeding may lower the risk for diseases, like hypertension, which have been shown to be associated with an increased risk for depression.
In conclusion, this is the first report showing an association between breast-feeding and the occurrence of depression later on in adulthood. The limited sample size and the inevitably retrospective nature of this analysis are limitations. Also, the herein reported association may not necessarily reflect a causal relationship. Thus, replications in larger samples as well as mechanistic studies are warranted.