Pictured: Laboratory director Remy Kachadourian logs in samples of marijuana edibles for THC potency testing at state-licensed cannabis testing facility Steep Hill Halent of Colorado in Denver on March 31, 2015. (Joe Amon, Denver Post file)

Why are there so many conflicting marijuana studies?

The relationship of marijuana use to public health is one of the most confounding topics in science. As soon as one study is published making a claim about the safety of the drug, another is often published warning of the exact opposite.

Take, for instance, the link between marijuana use and cardiovascular disease. Last week, a study was released detailing a review of outpatient hospital data, including more than 20 million health records and 316,000 people who reported using marijuana. The study found that using the drug was associated with 26 percent increased risk of stroke and a 10 percent increased risk of developing heart failure – independent of other health factors such as tobacco use, hypertension, diabetes and obesity. While the study hasn’t yet been peer reviewed (it will be presented this weekend at the American College of Cardiology’s annual meeting), it’s attractive given the massive amount of data.

The problem is that its conclusion directly contradicts a number of other large, recent studies looking at the same question. One longitudinal study published last month looking at more than 5,000 subjects over a period of more than two decades found no association between marijuana use and cardiovascular disease. Another longitudinal study from Sweden published last month reviewing more than 50,000 men found similar results.

Related: Here’s what is conclusively known about marijuana’s health benefits and risks, from a review of 10,000 cannabis studies

These are by no means the only studies on this topic. You can spend a whole day reading through the literature on this question and still not find a clear answer.

The root of the problem comes from considering all of the limitations on studying marijuana. Because it is considered a Schedule 1 drug by the federal government – meaning it is considered to have no medical benefits – conducting a randomized controlled trial (the gold standard in scientific experiments) on marijuana use is essentially impossible.

Instead, scientists must rely heavily on self-reporting from patients, making it extremely difficult to know how much, how often or through what means subjects use marijuana – let alone what type of chemicals they’re ingesting alongside the drug. Meanwhile, without collecting massive amounts of data from marijuana users, conclusions about those with cardiovascular issues are going to be based on small numbers of people.

“There’s not great infrastructure in place for us to study cannabis well,” said Aditi Kalla, lead author of the study released last week and a researcher from the Einstein Medical Center in Philadelphia.

She notes that with more and more states legalizing marijuana, it’s becoming more acceptable to use the drug and for patients to discuss personal use with researchers. But marijuana remains a highly controversial and politicized topic. Regardless of the outcome of any marijuana study, scientists who tackle the issue do so knowing they will probably receive criticism.

“This is not meant to be an anti-cannabis study,” Kalla said of her work. “We know we have limitations. We want to guide future research.”

There are plenty of health implications from marijuana that research has substantiated fairly well, as evidenced by an extensive review of more than 10,000 marijuana studies by the National Academies of Sciences, Engineering and Medicine published earlier this year. We know from this research that the drug works to treat chronic pain, multiple sclerosis spasticity and nausea and pain in cancer patients. But we’re also pretty sure that marijuana use is linked to the development of schizophrenia and an increase in traffic accidents.

So it’s clear from the science so far that use of the drug has both pluses and minuses. But there are still many questions about marijuana that need to be answered – especially on cardiovascular disease.

It would be wrong to justify not legalizing marijuana simply because of public-health issues. There are reams of evidence linking alcohol to terrible health outcomes with no net benefits – and yet few people seem ready to go back to banning booze. Still, given that heart disease is the leading cause of death in the United States, marijuana’s debatable link to cardiovascular issues should be a top health policy concern for everyone – supporters and opponents of marijuana alike. This is about making consumers aware of the risks of their choices. The more we know, the better.