Sometimes, in science, a topic can be so widely researched, and various results confirmed or refuted so many times, that it actually has the opposite effect from intended – with so much information out there, the actual results and conclusions can kind of get lost along the way.

That’s why we have meta-analyses: studies of studies, which combine the results from multiple pieces of research, weight them according to quality, and synthesize them into one overall conclusion. A new paper, published this week in BMJ Global Health, has done precisely that, for a topic whose effects are felt literally throughout life: the link between racism and poor pregnancy outcomes.

“A growing body of epidemiological evidence documents the health impacts of racism,” the authors write. “In particular, disparities in fetal, neonatal and maternal health outcomes have been reported, with racialized women experiencing worse outcomes.”

It’s long been known that people who face discrimination because of their skin color, ethnicity, or nationality have higher rates of maternal death and adverse pregnancy outcomes than those who don’t – but by considering the pooled results of 24 previous studies, the team behind this new meta-analysis were able to provide some context to those figures. “It is… not race, but racism, that is largely the root cause of racialized health disparities,” they write.

And the effect is not small: for pregnant people who experienced racism, the overall odds of a premature birth were estimated to be increased by 40 percent – even when low-quality studies were excluded, the odds were still increased by close to one-third over their non-racialized peers. Meanwhile, the chance that they have a baby which measures small for its gestational age – something which is linked to multiple complications after birth – was estimated to be increased by nearly one-fourth.

“Racism has far-reaching implications on the experiences of racialized individuals,” explained the authors. “As an upstream factor, it shapes other social determinants of health such as employment, poverty, education, and housing. Relating more directly to health, racism can impact what services and resources are available, such as referral to specialist care, access to health insurance, and access to public health services.”

But the problem goes deeper than that, they note. Even accounting for that upstream effect, “stark disparities” still exist between women who face racial discrimination and those who don’t: for example, “black women with higher educational attainment have better outcomes than black women with lower educational attainment,” the authors explain, and yet “they continue to have worse outcomes than white women with lower educational attainment.”

While the vast majority of the studies in the analysis were carried out in the US – a detail which the authors acknowledge as a limitation – the research was able to draw on the experiences of pregnant people from a wide range of racial and ethnic backgrounds, including Black or African American, Hispanic, Mãori, Pacific, Asian, Aboriginal, Romani, Turkish, and more. Across all the studies, and the thousands of women included, one pattern emerged: “significant positive associations (between experiencing racial discrimination and an adverse pregnancy event) and non-significant associations (trending towards positive) were reported, with no studies reporting significant negative associations.”

Of course, naming the problem is only half the battle – the question is, what can we do about it? The authors have some suggestions on how to curb this disturbing disparity.

“It is critical to work towards decolonizing and improving medical training by universally removing well-documented examples of racial bias which continue to perpetuate health inequities,” they write. “This includes the lack of teaching on dermatology and differential disease presentations in non-white individuals, inaccuracies in pulse oximetry technology, unsubstantiated race-based adjustments to measuring renal function, and inadequate teaching around individual biases and the social drivers of health inequities.”

Such biases can be deeply embedded in medical training, and obstetrics, the study of pregnancy, childbirth, and the postpartum period, is no exception. It was only in 2018 – notably well after humans first started giving birth – that researchers realized pelvis shapes differ by ethnicity, and that doctors had likely been performing unnecessary and potentially traumatic interventions on women with non-European backgrounds for centuries.

“Dismantling the structures and policies that enable institutional and interpersonal racism, underlying racial and/or ethnic disparities in health and intersecting social inequalities is essential to improve overall health in societies,” the authors conclude.

“Globally, it is critical for public and global health scholars, educators and practitioners to research and fight these phenomena to contribute to better and sustainable health outcomes at the population level.”