Dr. Ruchi Gupta did everything right to protect her daughter from peanut allergies — and it backfired.
Dr. Gupta was a young doctor researching food allergies. She knew the advice. At the time, in 2006, scientists recommended delaying the introduction of possible allergens until age 3 [1].
But one day, her daughter broke out in hives when her older brother touched her cheek after eating a peanut butter and jelly sandwich. She had developed a peanut allergy anyway.
It was just one case in a rising epidemic of peanut allergies.
And nobody seemed to know what was going on or how to stop it.
This case highlights something doctors got horribly wrong.
Table of Contents
- A Mystery: The Rising Tide of Allergies
- Early Hypotheses and Response
- A Clue
- A Theory Emerges
- The Dilemma
- A Breakthrough
- Aftermath
- References
A Mystery: The Rising Tide of Allergies
Parents didn’t always have to fear peanut butter sandwiches. Back in 1997, just 0.4% of American children had peanut allergies [2].
But by the time Dr. Gupta and a team of researchers published data in 2011, the figure had hit 2% [3].
That’s a 400% increase in a very short time. And it kept climbing. By 2016, it reached 2.2% [4].
Children’s bands like The Wiggles — which my 3 kids are obsessed with — even started performing songs encouraging parents not to bring peanuts to preschools.
And this rapid increase wasn’t just happening in the U.S. A study of two birth cohorts from the same region of the Isle of Wight in the U.K. compared children born in 1989 to those born in 2001. At age 3, only 0.5% of those born in 1989 were allergic to peanuts. For those born in 2001, that number was 1.2% [3].
Suddenly, kids with peanut allergies seemed to be everywhere. And it’s a serious issue. Peanut allergies usually don’t resolve with age, and the consequences can be deadly [5].
This wasn’t a slowly developing trend. It was an explosion.
But what was driving it?
Early Hypotheses and Response
At first, no one was sure. But an early theory suggested that exposing very young children to peanuts might trigger allergic reactions. Their immature immune systems might mistake peanut proteins for a threat, generating an allergic response. A study even indicated that maternal peanut consumption during pregnancy and early infant exposure could increase risk [6].
In response to the rising rates — and without definitive evidence — the American Academy of Pediatrics issued guidelines in 2000 recommending that children avoid peanuts until age 3 [7].
This left scientists scrambling for answers.
Some theories looked at broader health trends. It wasn’t just peanut allergies increasing. Allergic and autoimmune issues in general were becoming more prevalent in modern, industrialized countries [8].
At the same time, infectious diseases were becoming less common [8].
According to the “hygiene hypothesis,” these trends are connected. The theory suggests that reduced exposure to infectious agents — particularly those that co-evolved with humans — may lead to increased susceptibility to allergic and autoimmune diseases [8].
Others explored changes in the gut microbiome. For example, babies delivered by C-section display significantly different gut microbiota, due to the absence of maternal vaginal microbes. These differences may play a causal role in allergies and immune-related disorders [9].
Similarly, when pregnant women are obese, smoke, or experience dietary restrictions, it appears to predispose the developing baby to allergic disease later in life. This has been linked to increased Th2 cytokine signatures in cord blood — a marker of allergic potential [10].
It’s possible all of these contribute. But none fully explained why peanut allergies were getting worse — and so fast.
A Clue
An important clue emerged in 2008.
By that time, the advice to delay peanut introduction had become mainstream in countries like the U.S. and the U.K. Yet allergy rates were still climbing.
So researchers studied Jewish children living in the U.K. and in Israel — two genetically similar groups with very different cultural practices. In Israel, parents hadn’t adopted the peanut-avoidance guidelines. For babies aged 8 to 14 months, the median monthly consumption of peanut protein was 7.1 grams. In the U.K., it was 0 grams [11].
According to the dominant theory, children in Israel should have had higher rates of peanut allergy.
But the opposite was true.
Jewish children in the U.K. were 10 times more likely to have a peanut allergy than those in Israel [11].
This finding raised a crucial question: Could early exposure to peanuts actually prevent peanut allergies [11]?
Could the official guidelines have made things worse?
A Theory Emerges
Meanwhile, a new understanding of allergy development was forming — the Dual Exposure Hypothesis.
There are two main ways the body can encounter potential allergens: through the skin or through ingestion. The hypothesis suggests that skin exposure can sensitize the immune system — especially if it occurs before the body has had a chance to develop oral tolerance via the gut [12].
This theory helps explain why children with eczema — a condition that compromises the skin barrier — are at significantly higher risk for peanut and egg allergies. Their skin may allow environmental allergens to pass through and trigger sensitization [12].
So, it’s not that food allergies cause eczema — as had long been assumed — but that eczema may cause food allergies.
This was a breakthrough.
The idea is often summarized like this:
“If it’s through the skin, allergies begin. Through the diet, they stay quiet” [13].
In this view, the delay strategy backfired. Avoiding peanuts allowed children to first encounter peanut proteins through their environment — possibly via skin contact — without the protective effect of early ingestion.
The Dilemma
Even as this understanding developed, doctors were stuck in a tough spot.
There were roughly 9,500 food-allergy-related hospitalizations per year among U.S. children between 2004 and 2006 [14].
Without high-quality data, physicians couldn’t ethically recommend feeding allergenic foods to infants — no matter how promising the theory sounded.
Still, the American Academy of Pediatrics took a cautious step in 2008: they withdrew their 2000 recommendation to delay peanuts until age 3, citing insufficient evidence that delaying introduction helped [15].
They no longer said "delay" — but they weren’t yet saying "introduce early" either. That would require stronger evidence.
A Breakthrough
That evidence arrived in 2015.
The landmark LEAP (Learning Early About Peanut Allergy) study was published that year. It was a randomized clinical trial of 640 high-risk infants, aged 4 to 11 months. One group was instructed to consume peanut products regularly. The other was told to avoid peanuts until age 5 [16].
The results were dramatic.
Among infants who had no prior sensitivity (based on a negative skin-prick test), 13.7% of the avoidance group developed a peanut allergy. In the consumption group, the number was just 1.9% — a relative reduction of 86% [16].
But what about infants who were already sensitized?
The study also included a separate cohort of 98 infants who had positive skin-prick tests at the start. Here, too, early introduction helped. Peanut allergy developed in 35.3% of the avoidance group, compared to just 10.6% in the consumption group [16].
This was the robust data doctors had been waiting for.
That same year, a coalition of medical organizations issued guidelines recommending early peanut introduction for high-risk infants between 4 and 11 months [17].
In 2017, the guidance was expanded to include moderate-risk infants [18].
And finally, in 2021, guidelines recommended early introduction for all infants, regardless of risk — between 4 and 6 months of age [19].
“To prevent peanut and/or egg allergy, both peanut and egg should be introduced around 6 months of life, but not before 4 months. Screening before introduction is not required” [19].
This is exactly what I’ve done with my three children.
Aftermath
This reversal in peanut allergy guidelines is a fascinating example of how science is supposed to work. It responds to data. It self-corrects.
But updated guidelines don’t always change behavior.
So what actually happened?
A recent study looked at this. Researchers used a large database of electronic health records from the American Academy of Pediatrics network. They created three cohorts of children aged 0–3:
- Preguidelines cohort: Sept 2012 – Aug 2014
- Postguidelines cohort: Sept 2015 – Aug 2017 (after LEAP results)
- Postaddendum cohort: Feb 2017 – Jan 2019 (after 2017 expansion) [20]
They found a clear pattern.
- From the preguidelines to postguidelines cohort, the relative risk of developing a peanut allergy fell by 35%
- From preguidelines to postaddendum cohort, it fell by 45% [20]
By the most recent period, only 0.45% of children had developed a peanut allergy, based on their medical records [20].
That’s almost identical to the 1997 rate of 0.4% — before the allergy epidemic began [2].
So we’ve come full circle.
Bad guidance likely fueled the rise in peanut allergies. Better guidance — based on better science — is helping bring those numbers back down.
It’s a remarkable story of course correction.
References
1. https://foodallergyprevention.org/for-parents/
2. https://pmc.ncbi.nlm.nih.gov/articles/PMC12511798/
4. https://pmc.ncbi.nlm.nih.gov/articles/PMC6317772/
5. https://www.ncbi.nlm.nih.gov/books/NBK538526/
6. https://pubmed.ncbi.nlm.nih.gov/10410914/
7. https://jamanetwork.com/journals/jama/article-abstract/2841304
8. https://pmc.ncbi.nlm.nih.gov/articles/PMC2841828/
9. https://pmc.ncbi.nlm.nih.gov/articles/PMC10816971/
10. https://pmc.ncbi.nlm.nih.gov/articles/PMC10960768/
11. https://pubmed.ncbi.nlm.nih.gov/19000582/
12. https://www.mdpi.com/2072-6643/14/13/2565
13. https://jamanetwork.com/journals/jama/article-abstract/2841304
14. https://www.cdc.gov/nchs/products/databriefs/db10.htm
15. https://pubmed.ncbi.nlm.nih.gov/18166574/
16. https://www.nejm.org/doi/full/10.1056/NEJMoa1414850
17. https://www.jacionline.org/article/S0091-6749(15)00785-X/fulltext
18. https://www.jacionline.org/article/S0091-6749(16)31222-2/fulltext
