Babies who are looked after by relatives, friends, and neighbours while their mothers are at work, are less likely to be breastfed, suggests research published online in the Archives of Disease in Childhood.
This applies to both full time and part time care, and all strata of society, the research indicates.
The findings are based on 18,050 infants, who were part of the UK Millennium Cohort Study, which tracks the long term health and wellbeing of children born between 2000 and 2002.
In the UK in 2005 only one in four mothers breastfed any amount for six months - the minimum period for exclusive breastfeeding recommended by the World Health Organization (WHO) - with rates among mothers from poor backgrounds even lower.
The researchers wanted to find out whether childcare arrangements and socioeconomic factors had any impact on breastfeeding rates and assessed whether the infants had been breastfed for at least four months, which was the WHO recommended minimum until 2003.
Formal childcare provision was classified as a nursery, crèche, childcare centre or registered childminder. Informal childcare was classified as a friend, neighbour, relative or unregistered childminder.
Around a third of infants were breastfed for at least four months. In all, 7% (1430) were placed in informal childcare between birth and the age of 4 months and 2.3% (360) were placed in formal childcare.
The analyses showed that, after taking account of whether the mother had returned to work, infants in informal childcare were 50% less likely to be breastfed than those looked after only by a parent, and those in formal childcare around 15% less likely to be breastfed.
Informal childcare arrangements reduced the likelihood of an infant being breastfed across the board, irrespective of whether provision was part time or full time.
But when it came to formal childcare, only full time provision reduced the likelihood of breastfeeding - and then only for mothers with a degree, those from managerial and professional backgrounds, and those with partners.
By contrast, lone mothers were 65% more likely to breastfeed an infant provided with formal childcare.
"It is likely that for many mothers, it is not childcare use in isolation that influences the decision to breastfeed, but a chain of antenatal decisions about infant feeding, childcare and employment," comment the authors.
But the fact that informal childcare had the strongest impact on breastfeeding irrespective of social and economic factors prompts them to conclude that UK breastfeeding campaigns should target everyone, not just those from disadvantaged backgrounds.