(CNN) British teens with the highest test scores are less likely to smoke cigarettes yet more likely to drink alcohol and smoke pot compared with teens with lower scores, according to a study published Wednesday in the British Medical Journal Open.

Although some people believe smart students simply have a tendency to experiment, James Williams and Gareth Hagger-Johnson, co-authors of the new study, say these patterns of substance use may continue into adulthood.

"Our research provides evidence against the theory that these teens give up as they grow up," said the authors, both affiliated with University College London.

Why young people use alcohol and pot

In the decade ending in 2014, England saw a decrease in the percentage of 14-year-olds who admitted smoking and drinking. In 2004, 12% of 14-year-olds surveyed said they smoked cigarettes regularly, 23% said they drank alcohol at least weekly, and 17% had tried cannabis. Ten years later, the figures were 4%, 6% and 9%, respectively.

In the United Kingdom, the drinking age is 18, though teens who are 16 or 17 and accompanied by an adult can drink (but not buy) beer, wine or cider with a meal.

The news was good overall, but why certain teens use drugs remains unclear. Wanting to understand more, Williams and Hagger-Johnson surveyed more than 6,000 students from public and private schools across England.

Using questionnaires, they regularly tracked each student's use of tobacco, alcohol and cannabis from age 13 or 14 until age 19 or 20. Williams and Hagger-Johnson used national test scores taken at age 11 to rank students academically.

Some of their results provided no surprises.

During their early teens, high-scoring pupils were less likely to smoke cigarettes and more likely to drink alcohol than their peers with lower test scores. At this time, they were slightly more likely to say they used cannabis.

During their late teens, pupils with the highest scores were more than twice as likely to drink alcohol regularly compared with others, yet they also showed themselves to have less of a tendency to binge-drink. During this same period in their lives, the academically gifted students proved nearly twice as likely to use cannabis persistently and 50% more likely to use it occasionally compared with their peers with lower test scores.

One "potential explanation," Williams and Hagger-Johnson said, is that "higher-ability adolescents are more open to try cannabis but are initially cautious of illegal substances in early adolescence as they are more aware of the immediate and long-term repercussions that breaking the law might incur."

"Cognitive ability is also associated with openness to new experiences and higher levels of boredom due to a lack of mental stimulation in school," the co-authors added.

Another possibility: Clever teens may ingratiate themselves and be accepted by older peers who "facilitate access to alcohol and cannabis," they said.

Finally, drinking patterns may be related to "parental influence, since parents with high cognitive ability and socioeconomic status are known to drink alcohol more regularly," Williams and Hagger-Johnson said.

Meanwhile, average students were 25% more likely to use cannabis on occasion and 53% more likely to use it persistently in their early teens than pupils they outperformed on tests. Average-ability students also used weed more than lower scoring peers.

"Smoking cigarettes, drinking too much alcohol and smoking cannabis are widely accepted to be detrimental to your health," Williams and Hagger-Johnson said. Possible health effects include lung cancer, cirrhosis of the liver and psychiatric illnesses. Dependence or addiction is another possible consequence, but it's not just your health that may be affected. Another possible fallout for those who venture too deeply into "light" drugs is worse education and employment opportunities.

Results 'not entirely surprising'

"The cigarette smoking findings are consistent with the literature (students with lower academic standing being more likely to smoke)," Pat Aloise-Young, a psychologist and associate professor at Colorado State University, wrote in an email.

The alcohol results, though, are a "mixed bag," she said. Although in England, the findings indicate that drinking is higher for students with higher academic standing, in the United States, researchers typically find the opposite result in high school but the same result for college-age teens: That is, American college students consume more alcohol than their non-college peers, she explained.

"So these results are somewhat inconsistent with previous results; however, they are not entirely surprising," Aloise-Young said, since teens "have different views of smokers and drinkers."

"There is an important caveat, however," she said. The new study indicates that students with lower academic standing have higher levels of hazardous drinking, and this "is very important," despite the fact that the study authors do not focus on it.

"From a public health perspective, hazardous drinking is more likely to result in negative consequences such as driving while intoxicated, alcohol-related crimes and injuries and unwanted sexual encounters," Aloise-Young said.

According to Dolores Cimini, a psychologist and assistant director for prevention and program evaluation at the University at Albany's substance abuse counseling center, "a strength of the study was that it was conducted on a very large sample of young people." However, the fact that the study relied on self-reports and interviews means there might be issues "related to accuracy of recollections."

Cimini noted that recent studies conducted in the United States also note declining trends in substance use among teens. However, there's still cause for concern since several studies indicate a relationship between substance use during the late teens and education. The effects of excessive drinking and/or drug use tend to "accumulate over time," she said.

"Initially, disengagement related to substance use is manifested by spending less time studying and skipping classes," Cimini said. "As students become more disengaged ... the likelihood of earning lower grades, 'stopping out' or dropping out of college increases."

Add to that possible health effects: problems with short-term memory, other brain functions and sleep patterns caused by excessive drinking; or marijuana-related effects such as attention and concentration difficulties and decreased information processing abilities. All of these may "undermine the efficiency and effectiveness of study time" leading to lower grades, Cimini said.

Teen substance use can cheat students of a good education and negatively affect their lives, she said.

According to Dr. Amir Levine, a psychiatrist and assistant professor at Columbia University Medical Center, it's necessary "to differentiate between substance experimentation and problem drug use." Still, he finds the results "interesting."

"We usually think about youth who are not doing well in school as the ones that are prone to alcohol and drug use," Levine said; the new study tells us "it may not be so simple."