Childhood IQ 'linked to adult drug use'

Behind the Headlines

Tuesday November 15 2011

Cocaine use was much higher in adults with high childhood IQs

Clever children are “likelier to take drugs” reported The Independent today. Several newspapers have reported that children who performed better at IQ tests when aged five and 10 are more likely to take illegal drugs such as cannabis and cocaine by the time they are 30. The link was particularly strong for women, who were over twice as likely to have recently used cannabis or cocaine than their counterparts with lower IQs.

These results are based on a large British study that had followed 8,000 people born in 1970. As part of ongoing research their IQs had been assessed when they were aged five and 10, with later surveys asking about various aspects of illegal drug use at the ages of 16 and 30. People who had higher childhood IQ were more likely to have used a number of illegal drugs, including cannabis and cocaine, although this did not seem to be the result of social status or distress as a teenager.

Although the research has found a puzzling gap in drug use between people with different IQs, the research did not directly address why this gap exists. While some news sources have speculated that it could be due to increased expendable income, the availability of drugs at university or due to coping with the pressures of intelligence, the truth is we simply cannot tell from this study. It will take further research to untangle this conundrum, and also to see if the results apply to the rapidly changing drug scene of today.


Where did the story come from?

The study was carried out by researchers from the UK Clinical Research Collaboration DECIPHer project in Cardiff and University College London. Some of the original data collection was also performed by researchers from Bristol University. The study was funded by a large number of UK medical research councils and published in the Journal of Epidemiology & Community Health, a peer-reviewed medical journal.

The research was generally covered appropriately by the Daily Mail, The Daily Telegraph and The Independent, which tended to emphasise the fact that we cannot tell why the patterns of usage in the study occurred. Most included commentary suggesting the theories that people with a high IQ might be more open to new experiences and keen on novelty and stimulation, or that drug use could be a response to feeling ostracised during childhood. However, when offering these theories the newspapers rightly made it clear that they were speculative and not directly supported by the research.


What kind of research was this?

This study aimed to examine how a group of children’s IQ scores related to their use of illegal drugs later in life. Participants were drawn from a long-running, ongoing prospective cohort study that has examined the lives of people who were born between the 5th and 11th April 1970.

The researchers were interested in this as they said that previous studies have linked high childhood IQ scores with excess alcohol intake and alcohol dependence in adult life. They also said that high IQ in childhood is associated with positive health effects such as lower rates of death in mid-to-late adulthood, a lower likelihood of smoking, greater physical activity and a higher intake of fruit and vegetables. They said that higher childhood IQ is also associated with ‘socioeconomic advantage’ in later life, i.e. that they are likely to have gone on to further education and to have a higher income.

This study only looked for whether there was an association between childhood IQ and drug taking, it did not assess the causes for the association.


What did the research involve?

The 1970 British Cohort Study is an ongoing longitudinal study of children born in Great Britain between the 5th and 11th of April 1970. At birth a total of 16,571 babies were enrolled, with follow-up assessments given at the ages of five, 10, 16, 26 and 29-30 years. Drug use was only assessed at 16 and 30 years.

At the age of five the children’s cognitive function was tested using vocabulary tests and drawing tests, which would test their visual-motor context and also test how well they could capture a concept in their image. At the age of 10 a different set of tests were used to test IQ.

At 16 years the study members reported their level of psychological distress and whether they had ever tried drugs such as cannabis, amphetamines, barbiturates, LSD, cocaine or heroin. The researchers also included a fictitious drug, which they called semeron. Any participants who said they had taken semeron had their data removed from the analysis due to the possibility it could be inaccurate.

At the age of 30, similar enquiries were made about ever having tried a drug, but the range of drugs asked about was widened to include ecstasy, magic mushrooms, temazepam, ketamine, crack cocaine, amyl nitrate and methadone. At this time the participants were asked if they had taken any of these drugs, and the response options were yes; yes in the past 12 months; and no. If the participants had taken three or more drugs then they were defined as a ‘polydrug user’, i.e. user of multiple drugs. At the age of 30 the participants were asked about their highest educational achievement, their monthly gross salary and their occupation. Social class was assigned using a recognised classification system.

The researchers used a standard statistical technique called multivariate logistic regression to calculate the likelihood of drug use in 30 year olds who had IQs in the top third when they were five years old and compared it to the IQs of those in the bottom third. They adjusted their analysis for a number of factors that may have influenced the results, including their parents’ social class when they were aged five, psychological distress of the individuals when they were 16, present social class, monthly income, and level of education at 30 years.


What were the basic results?

Nearly 8,000 people from the original cohort study were included in the analysis.

When the cohort was asked about drug use when they were 16, some 7.0% of boys and 6.3% of girls had used cannabis. Only 0.7% of boys and 0.6% of girls had used cocaine. Both boys and girls who reported using cannabis had higher average IQ scores at 10 years than those who reported never using cannabis. There was no difference in the child IQ scores of adolescents who had used cocaine compared to those who had never used it.

After making statistical adjustments to their analysis, the researchers found that:

  • 30-year-old women with the higher IQs at age five were twice as likely to have tried cannabis relative to those with lower childhood IQs (odds ratio (OR) 2.25, 95% CI 1.71 to 2.97).
  • 30-year-old women with the higher IQs at age five were also twice as likely to have tried cocaine (OR 2.35, 95% CI 1.71 to 2.97)
  • At age 30 men with higher IQ scores at the age of five had a greater likelihood of having taken amphetamines, ecstasy or more than three drugs compared to men who had the lowest IQ scores at the age of five.
  • Having a higher IQ at five increased the odds of having taken amphetamines by 46%, ecstasy by 65% and multiple drugs by 57% relative to men who had a lower IQ at age five.
  • The IQ score at age five did not affect the likelihood that women had taken amphetamines, cocaine or multiple drugs by the age of 30.
  • However, similar to the women with higher childhood IQ the use of cocaine and cannabis was also more likely in men who had a higher IQ at the age of five relative to those with a lower IQ, but the gap in the likelihood of drug use between the highest and lowest childhood IQ groups was narrower in men.
  • For example, having a higher IQ at age five increased the odds of using cannabis by 83% and cocaine use by 73% compared to the lowest childhood IQ group in men. This is less than the more than twofold increase in likelihood seen in the women who had higher childhood IQs.

The same type of analyses were then performed, but instead comparing the people with the top and middle third IQ scores at the age of 10 to those who had scores in the bottom third. The research found that the highest IQ scores at 10 years were associated with cannabis use but not cocaine use at 16 years. The odds of cannabis use were threefold higher for boys and 4.6-fold higher for girls in the top third compared to the bottom third.

Men and women aged 30 who had IQs in the top third when they were 10 years old were more likely to have taken cannabis, cocaine, ecstasy, amphetamines and multiple drugs than those 30 year olds whose IQ scores had been in the bottom third when they were aged 10.

Again, for cannabis and cocaine use the relative odds in women tended to be larger than men. So, for example, women who had the highest IQs at 10 years were over three times as likely to have used cannabis or cocaine than women who had been in the lowest third. Men were just over two times more likely.

The researchers had focused more on the comparisons between the highest and lowest IQ thirds rather than comparing the middle against the bottom group.


How did the researchers interpret the results?

The researchers said that children with a higher IQ were more likely to use illegal drugs in adolescence and as an adult, and that their findings were independent of the effects of parent social class, psychological distress during adolescence and adult socioeconomic advantage.

They said that “potential pathways linking high childhood IQ with later illegal drug misuse are likely to be varied and require further explorations”.



This longitudinal prospective cohort study that followed individuals born in 1970 to the age of 30 years found that a higher IQ during childhood was associated with an increased risk of trying drugs by age 16 and 30. In particular, they found that the risk of taking cannabis or cocaine by the age of 30 was particularly great in the third of women with the highest childhood IQs compared to women who had IQs in the lowest third at that time.

This study had several strengths. It was large, in that it contained data from almost 8,000 people – although the initial study group contained over 16,000 participants but many were lost during the long-term (30 year) follow-up. It is not clear what caused this high proportion to not participate in the follow-up assessments.

Another strength was that the study adjusted for parental and adult socioeconomic position and asked about a variety of drugs. It is worth noting that the drugs that people were asked about when they were 30 included a bigger range than those at the 16-year survey, and that because they were asked only twice about drug use we do not know how use may have changed between the ages of 16 and 30 in those who were users. Another important point to note is how the availability of drugs and drug use patterns have changed over this time and whether these results would apply to children who are growing up and going through adolescence now.

Ultimately though, it should be remembered that this study did not look at why childhood IQ may be associated with later illegal drug use. There could be any number of plausible theories put forward, such as the children with higher IQs being more likely to go on to university were they may have more access to drugs, these children growing up to be more open-minded to drug taking or more willing to take risks, but the research cannot tell us whether these are actually true. Having found patterns in the way the participants used drugs, it would seem the next step would be to closely look at why these patterns exist.

Analysis by Bazian

Edited by NHS Choices

Links to the headlines

Clever children more likely to end up on drugs. The Daily Telegraph, November 15 2011

Women with high IQs 'more likely to take illegal drugs in their thirties'. Daily Mail, November 15 2011

Clever children 'likelier to take drugs'. The Independent, November 15 2011

Links to the science

White J, Batty GD et al. Intelligence across childhood in relation to illegal drug use in adulthood: 1970 British Cohort Study. Journal of Epidemiology & Community Health, 2011


How helpful is this page?

Average rating

Based on 2 ratings

All ratings

2  ratings
0  ratings
0  ratings
0  ratings
0  ratings

Add your rating