Women who struggle with their mental health have an almost 50% higher risk of preterm births, according to the biggest study of its kind.

The research, published on Tuesday in the Lancet Psychiatry, examined data from more than 2m pregnancies in England and found about one in 10 women who had used mental health services had a preterm birth, compared with one in 15 who did not.

The study also found a clear link between the severity of previous mental health difficulties and adverse outcomes at birth. Women who had been admitted to psychiatric hospital were almost twice as likely to have a preterm birth compared with women who had no previous contact with mental health services.

And women with history of mental health difficulties faced a higher risk of giving birth to a baby that was small for its gestational age (75 per 1,000 births compared with 56 per 1,000 births).

The study recommends that when pregnant women are first assessed by doctors and midwives they should be sensitively questioned in detail about their mental health.

One of the reports authors, Louise Howard, professor emerita in women’s mental health at King’s College London, said such screening would help identify “clear red flags for a possible adverse outcome”.

She said once these mental health risk factors are identified they can be treated, potentially reducing instances of maternal deaths, stillbirths, premature births and underweight babies.

Howard said: “Mental illness is a treatable problem … maternity professionals will be focused on thinking about other modifiable risk factors such as smoking and obesity, but they may not have thought about some of the additional risks that women with mental illness may have.”

She said the study highlighted the need for well-resourced mental health services.

“These services are under huge pressure,” she said. “But this study highlights the need for these different interventions to be available, because they really do make a difference.”

Howard, who chaired a committee that drew up 2014 Nice guidelines on antenatal and postnatal mental health, said these guidelines should be updated in light of the study.

She said: “What we didn’t emphasise in 2014, because we didn’t have the data then, is how important it is to take a really detailed history of when mental health problems occurred and how severe they were.”

Howard added: “The significance of the study is the association of adverse birth outcomes with the severity of illness. It also emphasises the need for really good training of midwives and obstetricians around mental health issues.”

The study found that 0.65% of pregnant mothers who had been admitted to psychiatric hospital had a stillbirth, compared with 0.45% of women who had no pre-pregnancy mental health care. Howard said: “These are quite small numbers, but they’re still devastating in terms of stillbirths.”

Previous smaller studies have also established a link between adverse birth outcomes and mental health. But Howard said the difference with this study was its size and comprehensiveness. “So it’s all women rather than just women who attended a particular hospital in a particular area.”

The study did not explore the reasons behind the link between poor mental health and adverse birth outcomes.

Howard said: “Mental illness itself is likely to be associated with some of these adverse outcomes, because it impacts the stress system and is therefore likely to cause problems. But it is also associated with other factors. If you’ve got mental illness, you’re more likely to smoke because you’re feeling stressed. Similarly with substance misuse. And there are other stressors such domestic violence, or a history of childhood abuse, or poverty.”

She said mental health medications may also be linked to problems in pregnancy. “Antipsychotics are associated with obesity, and we know that obesity is a major risk factor for these outcomes. Some antidepressants may be associated with some of these outcomes as well. That’s why we have to think really carefully about who would benefit from medication or whether psychological therapy could be used instead.”

The study involved researchers from the University of Exeter, King’s College London, the London School of Hygiene & Tropical Medicine and the University of Liverpool.