Study examines peanut allergies in England

Monday February 7 2011

“Middle-class children 'face twice the risk of nut allergy' than those from poor families,” reported the Daily Mail. Covering the same story, BBC News said that “boys are more likely to be diagnosed with a peanut allergy than girls.”

These findings are from a study which looked at large amounts of data, collected between 2001 and 2005, for almost 3 million people registered at just over 400 GP surgeries in England. It found that in 2005, about 5 in 10,000 people in England had a recorded peanut allergy. As reported, the study also found that boys were more likely to be diagnosed with a peanut allergy than girls up to the age of 18, and that there were differences between socioeconomic groups.

This large study provides a good estimate of the prevalence of peanut allergies. However, the findings cannot tell us why these trends have been observed, and whether there is a difference between socioeconomic groups because affluent people have better access to healthcare, or because poorer children are protected in some way, as the Daily Mail suggested. These trends can only be explained by further research.

Where did the story come from?

The study was carried out by researchers from the University of Edinburgh and the Maastricht University Medical Center in the Netherlands. Funding was provided by the NHS Health and Social Care Information Centre. The research was published in the peer-reviewed Journal of Allergy and Clinical Immunology .

BBC News and the Daily Mail provided generally good coverage of this study.

What kind of research was this?

This time trend study examined the incidence and prevalence of peanut allergies between 2001 and 2005 in England, based on diagnoses made by GPs. The study data came from a large national database of information collected from general practices. The researchers say that other studies have assessed how prevalent peanut allergies are, but in general they have only looked at a relatively small sample of the population, which may not be representative of the population as a whole.

This type of study allows researchers to look at the number of new diagnoses occurring over time in a large population, and the number of people with the diagnosis at any point in time. Such studies are useful in telling us how common a condition is, and in identifying trends over time, such as an increase or decrease in new diagnoses. However, they cannot tell us why these changes occur.

What did the research involve?

The researchers used a large national database of information collected by UK general practices to identify people diagnosed with peanut allergies in England between 2001 and 2005. The database contained anonymous information about almost 3 million patients registered with 422 English GP surgeries.

For each year, the researchers recorded the number of people registered at each GP surgery, and then identified all patient records that had been given a code meaning the patient had been diagnosed with a peanut allergy. They identified which cases were new (first) diagnoses of peanut allergy each year, and how many people had an existing diagnosis.

This data allowed the researchers to calculate what proportion of the English population has new or existing diagnoses of peanut allergy each year. They then looked at the rates of the condition in different groups and over the study period to see if they could identify any trends. For example, they looked at whether having a different socioeconomic status was related to the risk of having a peanut allergy, with people’s socioeconomic status assigned based on their postcode.

What were the basic results?

In 2005, there were eight new cases of peanut allergies diagnosed for every 100,000 people followed up for the full year. In the same year, the prevalence of peanut allergies was 51 in every 100,000 people (in other words, 51 people in every 100,000 had a peanut allergy). This meant that for the entire population of England, an estimated 4,000 new cases of peanut allergy were diagnosed in 2005, and 25,700 people had an existing peanut allergy.

The researchers found that between 2001 and 2005, the number of new cases each year remained relatively stable, but the number of existing cases doubled from 24 in 100,000 people in 2001 to 51 in 100,000 people in 2005. For people under 18 years of age, the likelihood of having a peanut allergy was higher in boys than girls. The condition was more common in higher socioeconomic status groups than lower socioeconomic status groups.

How did the researchers interpret the results?

The researchers say their study has shown that peanut allergies are a lot less common than previous studies have indicated. However, they say that the difference in the figures between this study and previous studies may partly be because some cases of peanut allergy are missed in general practice records.

To clarify the matter, they suggest that “further research is needed to assess the true frequency of peanut allergy in the population and whether there has been a true increase in recent years.”


This study’s main strength is its use of data on a large group of people registered with their GP surgery. This population should provide a reasonably good representation of the population as a whole. There are some other points to note:

  • As these data were collected as part of routine data collection by GPs, and not specifically for the purposes of this study, there may be some differences in how diagnoses were made and recorded. It is possible that not all patients will have received the “gold standard” (best) method of diagnosing peanut allergy, which is a double-blind placebo-controlled food challenge.
  • There may also be some inaccuracies arising from categorising people’s socioeconomic status based on their postcode alone.
  • To be identified as having a peanut allergy in this study, people would have to see their GP about this condition. The study would have missed those who were unaware of their allergy or whose GPs were unaware of it.
  • The researchers note that the change in prevalence over time could be due to increased awareness of the condition in patients and GPs, leading to improved diagnosis rates.
  • Previous studies have reportedly estimated the prevalence of peanut allergy in children as between 4 in 1,000 and 19 in 1,000. The rates in this study were substantially lower, with the prevalence in children about 1 in 1,000 in the 0-4-year-old age group, about 2 in 1,000 in the 5-9 and 10-14 groups, and 0.7 in 1,000 in the 15-19 group. The authors suggest this may be partly due to the prevalence being underestimated in their study because of missed diagnoses, and partly due to overestimation in previous studies due to the methods used to select participants. They suggest that the true rates are likely to lie somewhere between their estimate and those of previous studies.

Overall, this study provides helpful estimates of how common peanut allergies are in England. Although various trends have been identified, such as the link with socioeconomic status, further studies would be needed to investigate why these links might exist.

Analysis by Bazian
Edited by NHS Choices