Women who use cannabis during pregnancy may be more likely to give birth prematurely, according to a comprehensive new study.

The team of Canadian researchers were quick to point out though that their findings are tentative, and that more work is needed to fully determine whether there is a causal link between marijuana consumption and premature births.

That said, the sample size used by the researchers to inform their analysis is huge — more than 600,000 pregnant women.

The team, with members from Ottawa Hospital, the Children’s Hospital of Eastern Ontario, the provincial birth registry BORN Ontario and the University of Ottawa, took the data from administrative records in Ontario. Essentially, the study involved all the women who gave birth in the city between 2012 and 2017, before cannabis was legalized.

All of these women were asked about their marijuana use at an early prenatal visit. Of this large sample, only 1.4 percent self-reported using cannabis while pregnant — a total of 9,427 women.

The rate of preterm births for these women was 12 percent, while for non-users it was only 6 percent. But the researchers noted that other factors, namely lifestyle habits or circumstances, could be at play which more strongly determine premature births among cannabis-users.

The data collected by the researchers showed that women who self-reported cannabis use were more likely to be teenagers, underweight and poor.

They also self-reported higher rates of tobacco, alcohol and opioid use during pregnancy compared to non-users of marijuana.

To better isolate the impact of cannabis use during pregnancy, the researchers employed a matching technique of users to nonusers who share some of these risk factors.

For example, a 17 year old marijuana user with little income who drinks alcohol would be paired with another 17 year old with little income who drinks alcohol, but who does not use cannabis.

The researchers created matches for all marijuana users. Those for who there was no match were left out of the analysis, while those with many were compared with an average measure of their matches. In doing so, the researchers hoped to create two groups which were as similar as possible along a number of variables, typically risk factors associated with birth complications, except for marijuana use.

The researchers found evidence of more likely poor birth outcomes for cannabis users, including premature births, infants born of a smaller size and admissions to intensive care.

For premature births, the risk for self-reported cannabis users stood at 10.2 percent versus 7.2 percent for non-users. For admissions to a neonatal intensive care unit, users experienced a 19.3 percent risk compared to 13.8 percent for non-users.

While significant, the analysis suggests the risks are not overwhelming and that further research is required to explore other differences between and within users and nonusers which could be crucial to birth outcomes.

For instance, noting that a pregnant women is underweight is one thing, but what other metrics of a person’s health should we consider to give more validity to the matching technique?

Also, the question of when and how a pregnant woman uses marijuana is significant. Does it make a difference whether it cannabis is used at the beginning or at the end, or throughout, a pregnancy? Does it make a difference whether a pregnant women smokes marijuana or consumes edibles? And perhaps most importantly, how intensively does a pregnant women use marijuana?

At least in part because of these uncertainties, the researchers do not conclusively say that marijuana use was directly linked with poor birth outcomes among these women. But the significant association they have found gives greater urgency for more quality research. The team’s measured and comprehensive findings are a welcome addition to a discussion which is all too often overblown with unfounded hysteria.

The study, published in the Journal of the American Medical Association, is available online.