An unidentified man smokes medical marijuana during karaoke night at the Cannabis Cafe in Portland, Ore. (Rick Bowmer/Associated Press)

If you're confused about what marijuana use really does to people who use it, you're not alone. For years, the scientific research on health effects of the drug have been all over the map.

Earlier this year, one study suggested that even casual marijuana use could cause changes to the brain. Another found that marijuana use was also associated with poor sperm quality, which could lead to infertility in men.

But marijuana advocates point to other research indicating that the drug is far less addictive than other drugs, and some studies have found no relationship between IQ and marijuana use in teens.

Researchers at the Center for Brain Health at the University of Texas in Dallas sought to clear up some of the confusion with a study that looked at a relatively large group of marijuana users and evaluated their brains for a slew of different indicators.

What they found was complex, but the pattern was clear: The brains of marijuana users were different than those of non-marijuana users. The area of the brain responsible for establishing the reward system that helps us survive and also keeps us motivated was smaller in users than in non-marijuana users. But there was also evidence that the brain compensated for this loss of volume by increasing connectivity and the structural integrity of the brain tissue.

Those effects were more pronounced for marijuana users who started young.

"The orbitofrontal cortex is one of the primary regions in a network of brain areas called the reward system," explained Francesca Filbey, lead author of the study and an associate professor of the neurogenetics of addictive behavior at the University of Texas in Dallas. "It helps us determine what is good for us and what keeps us sustained. "In this case, the orbitofrontal cortex plays a role in drug use because drug use and things associated with it --paraphernalia for example -- are associated with the rewarding effects of drugs."

The study, published Monday in the Proceedings of the National Academy of Sciences, used MRI scans to look at the brains of 62 non-marijuana users and 48 regular marijuana users, 27 of whom used marijuana but not other drugs. The marijuana users reported ingesting the drug about three times a day --  very heavy use -- and had used it for an average of about 10 years.

Looking first at atypically heavy marijuana use is necessary to identify structural effects, if there are any, Filbey said. Subsequent research would look at varying levels of use to find out if differences compared to non-users are still observed. For example, in Colorado, only about 22 percent of marijuana users reported using the drug about once a day at most.

Relative to other studies, this one had a fairly large sample size; it also excluded participants who had symptoms of psychosis, brain injury or neurological disorders in order to reduce the likelihood that the tests would pick up on other confounding factors. And it looked at three brain characteristics: the volume of the orbitofrontal cortex, how connected that part of the brain was to other areas, and the structural integrity of the white matter.

"We found that while the orbitofrontal cortex was smaller, there was greater functional and structural connectivity," said Filbey. "The white matter seemed to have greater integrity than the [non-marijuana using group]. And the connection between the orbitofrontal cortex and other areas were stronger."

That's potentially positive news suggesting that whatever impact marijuana use might have on the size or volume of that part of the brain, it may be offset by better connectivity and structural soundness. "It suggests that there is definitely a more complicated pattern that the brain seems to be able to compensate for any kind of loss in order to keep that network maintained," Filbey said.

However, Filbey's research also indicated that some of these findings could be affected by how young the person was a the time that they started using marijuana and how long marijuana use continued. People who started younger appeared to show more protective effects -- stronger connectivity and integrity in their white matter. But the longer the person used the drug, those features began to erode. After about six or seven years of use, connectivity and integrity begin to erode.

"So whatever compensatory effects that there may be originally seemed to diminish with prolonged use," she added.

There is no proven connection between the structural characteristics of this area in the brain and certain behaviors in humans. And it is possible that the small orbitofrontal cortex observed in marijuana users predated their marijuana use. A 2012 study found that smaller orbital frontal cortex volume at 12 years of age appeared to predict the initiation of marijuana use later in life.

Further long-term research is needed to determine what if any effects cessation has on brain structures of marijuana users.

[This post has been udpated.]

Voters in Oregon, Alaska and the District of Columbia voted to legalize marijuana in the 2014 midterm elections. They joined Colorado and Washington, which did so in 2012. So could other states follow suit in 2016? The Washington Post's Niraj Chokshi breaks down the possibilities. (Jorge Ribas/The Washington Post)