Girls who are not breastfed as babies or who grow up without a father are likely to have their own children earlier, according to a long-term study of more than 4,500 women across the UK.

The researchers looked at four kinds of dirsuption that might affect the family lives of young girls and found that each one brought forward the age at which they first became pregnant by six months on average.

The study underlines how setbacks in childhood can continue to have an impact into adulthood.

"The early environment matters a lot because it calibrates what kind of world you're living in and what kind of resources you have available to you," said Daniel Nettle, a behavioural scientist at the Institute of Neuroscience at Newcastle University, who led the work. However, he stressed that children could rise above problems during their upbringing.

The study found that even after controlling for major differences such as the parents' income and social class, four factors during upbringing had an impact on the age at which a woman had her first child: whether she was breastfed; how involved her parents were in her upbringing; whether her father was present; and whether her parents moved house regularly.

The key message, said Nettle, is that what happens to children early on in their lives is hugely influential. "As a society, we should never forget that. If you get girls when they're 14 or 15 and say, 'Hey girls don't have babies young', it may not do a lot at that point if, in fact, there have been more influential events much earlier in their lives that set them on a path towards a desire to have short time horizons and have babies young. A lot of those things are to do with stress and wellbeing of kids in deprived areas."

Rather than just telling girls to use condoms, he added, authorities should think much more about the context of people's early years. The research is published today in the Proceedings of the Royal Society B.

Nettle's team used data from the National Child Development Study, an ongoing longitudinal study of all people born in the UK between 3 and 9 March 1958. The cohort has been repeatedly surveyed to gather information about family life and health. Their parents and teachers have also been interviewed to add independent sources of information about their lifestyles.

In his study, Nettle focused on 4,553 women in the cohort who had provided details of their first pregnancy and looked at four potentially disruptive factors in the first seven years of their childhood. Each factor was given a numerical value in the researchers' analysis. In the case of family moves, Nettle calculated whether the child moved home more times than average for a child in that particular area. Fewer moves than average gave a child a "no disruption" score of 0, more moves than average would give a "disrupted" score of 1.

Nettle's results showed that low paternal involvement in the first seven years of a girl's childhood meant an average reduction in the age of first pregnancy of 0.74 years; prolonged separation from a mother in childhood entailed a 0.64-year reduction in first pregnancy. And the effects could be added together, so that if a girl had two of the disruptions to her early life, her age at the time of her first pregnancy was around a year earlier than average. Having all four disruptions brought forward the age of first pregnancy by around two years.

Nettle acknowledged that the nature of the study meant it was hard to be absolutely sure that the early life factors were responsible for the age shift, but he was confident that his research had controlled statistically for other factors such as income and education.

Diana Kuh, based at University College London and director of the Medical Research Council's National Survey of Health and Development, said that Nettle's findings support and extend earlier work on the 1946 British birth cohort, published by Kath Kiernan and Ian Diamond in Population Studies. They looked at the factors that affected age at which people first became parents up to age 32 years in men as well as women in this older national cohort.

"She found many factors in childhood and adolescence, similar to the factors found by Nettle et al, were related to age at first childbearing. For example, Dr Kiernan found that low parental interest in the child's education was associated with earlier childbearing, which supports both the social and evolutionary theories that Nettle discusses."

She added that another related area of research showed that many characteristics of early conditions are related to when a woman experiences the menopause. "They include, for example, early social stressors (such as the experience of parental divorce or poor socioeconomic conditions while growing up) that were associated with an earlier menopause and others (like being breastfed and early cognitive ability) that were associated with a later age at menopause. Overall, I think there is growing evidence of early life effects on lifetime reproductive function and that both social and biological pathways are implicated and need to be further explored."

Nettle said the results could also be interpreted from an evolutionary point of view because it makes sense to have babies earlier if childhood is disrupted. "If your early world is full of stress, that may calibrate you onto a pathway of develop fast, don't think too far ahead about the future, get on with stuff at an early age. That makes perfect sense – if your environment is full of stress and danger, you probably should do that."

Nettle's work supports similar results found in other mammals, such as rats and rhesus monkeys, which show that the reproductive schedules of females can be influenced by environmental conditions during their early development. "Female rats whose mothers are kept under caloric restriction during pregnancy reach puberty earlier than controls," write the researchers in the Royal Society journal. "Female rhesus monkeys that experience poor maternal care develop an increased interest in infants during the juvenile period, suggestive of accelerated reproductive schedules."