Thursday October 1 2009
Sweet enough already?
“Lots of sweets makes kids thuggish adults,” said The Mirror today. The newspaper reports that research has found that more than two in three people (69%) with a violent record by the age of 34 had “scoffed confectionery every" day when they were 10 years old. The newspaper quoted experts who think that this aggression comes from not learning patience in childhood.
The research, involving 17,500 people, is the first to look at adult violence in relation to childhood diet. However, there are other possible explanations for this link including the fact that difficult children might be given more sweets. It should be noted there was a high proportion of people who ate sweets every day in both the violent and non-violent groups. Also, it appears that less than 0.5% of children (about 81) in this study became violent offenders.
Overall, this study on its own does not provide strong enough evidence to support media explanations for the supposed link, which would need more study through dedicated research. Regardless, common sense tells us that eating too many sweets is not good for children’s health.
Where did the story come from?
Dr Simon Moore and colleagues from Cardiff University carried out this research. The study was funded by a grant from the Economic and Social Research Council, and published in the peer-reviewed British Journal of Psychiatry.
What kind of scientific study was this?
This was a retrospective analysis of data from a prospective cohort study, the British Cohort Study. This research collected data on newborns at regular intervals from 1970 onwards. It followed 17,415 babies born in the UK in one particular week in April that year, and also collected data on their families. The researchers estimate that 95-98% of all births in that week were included.
The researchers explain that diet has been associated with behavioural problems, including aggression, but that the long-term effects of childhood diet on adult violence have not been studied. Using the previously collected data they attempted to see if eating sweets and chocolates at 10 years of age was a predictor of convictions for violence in adulthood, up to 34 years of age.
Since 1970, there have been seven periods of data collection that used questionnaires to ask about health, education, social and economic circumstances. These took place when the study participants were aged 5, 10, 16, 26, 30, 34 and 42. The researchers only used the data from 5, 10 and 34 years of age.
At 10 years old, participants were asked how often they ate sweets, and at 34 years, they self-reported violent offending data and additional information on socioeconomic status. A computerised system was used to ask the questions about violent offending. Some additional questions from the data collection at five years old were used to classify children’s early development and their parents’ style of parenting.
Responses to the questions about eating confectionary at the age of 10 were converted into two possible answers: every day or less often/never. The results were analysed using a rare event logistical model, which takes into account that only 0.47% (possibly about 81 children) became violent offenders.
What were the results of the study?
Overall, 69% of respondents who were violent by the age of 34 years reported that they ate sweets nearly every day during childhood. Sweets were eaten this regularly by 42% of those who were non-violent.
The researchers say that children who ate confectionery daily at age 10 were significantly more likely to have been convicted for violence at age 34, a relationship that was robust when controlling for ecological and individual factors.
What interpretations did the researchers draw from these results?
The researchers say that children who ate confectionery daily at age 10 were significantly more likely to have been convicted for violence at age 34 years and that this link remained significant even after controlling for a number of other environmental and the individual’s life factors.
What does the NHS Knowledge Service make of this study?
This analysis of the British Cohort Study has the advantage of a large sample size. As it is designed prospectively, it also avoids the chance of reverse causation, i.e. the possibility that in some way violent offending might determine dietary habit. However, there are limitations to this study, some of which are mentioned by the authors:
- As a general population cohort study it was not designed to specifically examine the nature of diet and how it might be related to behaviour in the long-term. This increases the chance that the original study did not include questions about aspects that later became important. For example, the study does not appear to have asked about family income.
- The researchers collapsed the responses about how much confectionery was consumed into two categories, known as a binary variable (every day or less often/never). Analyses using this method means that important links between the amount or type of confectionary eaten may have been lost. The approach increased the chance of finding a statistical link for the rare event, (e.g. offending), but at the expense of useful information.
- The absolute number of children who became violent offenders is not reported in this publication and this also makes it difficult to be sure that the difference in eating habits between a small number of violent offenders and a large number of normal adults is statistically significant.
- The detail of questions asked by the self-reported computer-assisted interview are not reported and the context of how such sensitive information is collected should be considered when assessing how reliable the responses might be. The accuracy of information provided could have been checked against other records or by face-to-face interview. The number of people choosing not to respond to these questions was not published.
Overall, this study on its own does not provide strong enough evidence to guide childhood dietary advice, although common sense says that eating too many sweets is probably not good for children. Before the newspapers’ explanation for a link can be believed there must be studies specifically designed to investigate the issue from the outset.