The findings are based on data from 1,229 participants in the Great Smoky Mountains Study, a long-term study of residents in 11 counties near the Appalachian Mountains in western North Carolina, where Hispanics and Latinos are underrepresented and Native Americans are overrepresented compared to the rest of the US. A cohort of children in the study were enrolled as young as age 9 and have now reached their 30s. From 1993 to 2015, researchers tracked data in numerous areas of interest, including mental health, education, work attainment, and use of drugs and alcohol. The researchers defined problematic cannabis or marijuana use as daily consumption or a habit that meets diagnostic guidelines for addiction. They tracked participants' patterns of use from the college years (ages 19-21) into adulthood (ages 26-30).

They found more than three-fourths (76.3 percent) of participants didn't use or develop a problem with marijuana during this period. The remaining quarter developed problems that researchers grouped into three profiles: those with limited problems, persistent problems, and delayed problems.

* Limited users (13 percent)
Limited problematic users had trouble with marijuana either while in school before age 16 or in their late teens and early 20s, but their habits dropped off as they aged. Researchers were somewhat surprised that this group reported the highest levels of family conflict and instability during childhood as compared to others in the study; these factors are often associated with more drug use. "When this group of children left home, they seemed to do better," Dr. Hill said. "They didn't have as many children at a young age, and they went further in their education when they were 19 to 21 compared to those with persistent and delayed profiles."

* Persistent users (7 percent)
This group had trouble with marijuana beginning as young as 9 years old, and their chronic use continued into their late 20s and early 30s, the data showed. Large portions of this group had anxiety disorders in both childhood (27 percent) and at ages 19-21 (23 percent). They had the highest rates of psychiatric disorders and involvement in the criminal justice system, and most said the majority of their friends were drug users, too. "This suggests that a focus on mental health and well-being could go a long way to prevent the most problematic use," Dr. Hill said.

* Delayed users (4 percent)
This is a small but unique group that made it through adolescence and early adulthood without problematic marijuana use, only to become habitual users between ages 26 and 30. Blacks were five times as likely as whites to be delayed problematic users in their late 20s and early 30s after not having trouble between the ages of 19-21, a peak time for most marijuana users. More than half of delayed users were both bullied by peers and mistreated by caregivers as children, yet also had lower rates of anxiety, alcohol use, and other hard drug use compared to persistent users. "What we don't yet understand is how childhood maltreatment didn't prompt earlier problematic use of cannabis between ages 19 and 21-how individuals could be resilient to that kind of adverse experience for so long," Dr. Hill said. "One theory is that they were somewhat protected by having fewer peers in late adolescence who were substance users, but this is one of the questions we will continue to seek answers for."

In addition to Dr. Hill, study authors include E. Jane Costello, PhD, and William Copeland, PhD, of Duke, and Lilly Shanahan, PhD, of the Jacobs Center for Productive Youth Development, University of Zurich.