To the Editor,
‘Dietary diversity’ (DD), is defined as the number of different foods or
food groups consumed over a given reference period1. A
report from the European Academy of Asthma, Allergy and Immunology found
that increased DD may reduce the risk for development of food and
aero-allergen sensitization via its effect on the microbiome and
increased intake of immune-modulatory nutrients2. The
report summarized 14 papers reporting the role of DD on allergy
outcomes. Only two studies reported on the association between increased
DD and food allergen sensitisation3,4, one of which
found a reduced risk4. One study reported an
association between increased DD and decreased aero-allergen
sensitization4.
The aim of this study is to assess the effect of DD in the first year of
life on food and aero allergen sensitization over the first ten years of
life in a population birth cohort.The Food Allergy and Intolerance
Research (FAIR) birth cohort included children born on the Isle of Wight
(UK) (n = 969) between 2001-2002 who were followed up
prospectively3. Demographic and reported familial
allergy data were collected at 12 weeks gestation using the validated
ISAAC questions4 . Infant and child allergy data was
collected at 3, 6, 9 months and 1, 2, 3 and 10 years.
Food introduction data were collected via a standardized questionnaire
at ages 3, 6, and 9 months5. Specific information was
collected regarding breastfeeding duration, introduction of infant
formula feeding and age of introduction of 21 foods categorized at the
same time periods7. The 21 foods were fruits (citrus
fruit, non-citrus fruit, strawberries, tomato), vegetables, potato,
pulses, non-wheat cereals (rice, oats), meat (chicken/turkey, lamb,
beef, pork) and food allergens (cows’ milk products, egg, wheat, peanut,
tree nuts, fish, soy and sesame). DD was calculated as the number of
foods introduced by 3, 6, and 9 months. Cumulative DD was calculated as
the number of different foods eaten by 6 months plus number of foods
eaten by 9 months6,7. Portion size and frequency of
intake data was not available.
Skin prick tests (SPT) were performed to milk, wheat, egg, cod, peanut
and sesame at 1, 2, 3 and 10 years. A new variable was calculated for
children diagnosed with any food sensitization in the first ten years of
life (“over the first ten years”). The following aeroallergens were
tested via SPT at the same time points: house dustmite, cat and grass.
Ethical approval was obtained from the NRES South Central - Southampton
B Research Ethics Committee (REF 10/H0504/11). All parents consented and
children provided assent.
Data were double entered on SPSS versions 20 and 21 (SPSS Inc, Chicago,
USA). Descriptive statistics with means (standard deviations) or counts
(frequencies) were calculated. Univariate analysis was carried out to
assess the associations between DD, food and aero allergen
sensitization. Logistic regression models were fitted to describe the
relationship between the binary food sensitization variables, DD and
other related covariates. If independent variables (family history of
allergic disease, maternal history of food sensitization and food
allergy, parity, maternal smoking prior to pregnancy, breastfeeding
duration, timing of introduction of solid foods, and any eczema in the
first year of life) were found to be statistically significant at the
p=0.05 level in the univariate analysis, the variables were entered into
a multivariate model to understand the variables at each time point that
are independently associated with food sensitization in the first 10
years of life. Spearman correlations were performed to examine
relationships between count data. All significance tests were two sided
and analyses were performed with SAS version 9.4 (SAS Institute Inc.,
Cary, NC, USA). We made no adjustments for multiple comparisons because
the hypotheses were made a priori .
Data was available on 851 mother-infant dyads, for whom we had
information on DD, food and aero allergen sensitization on at least one
occasion up to 10 years of age. There were 763 (89.7%) children with
data at 1 year, 658 (77.3%) at 2 years, 670 (78.7%) at 3 years and 589
(69.2%) at 10 years. Demographic, environmental and allergic
characteristics are shown in Table 1.
The DD by 3, 6 and 9 months was 0 (IQR 0), 11 (IQR 9-13) and 16 (IQR
14-17). Cumulative DD at 6+9 months was 27 (IQR 24-30). Table 2 shows
the univariate results for the associations between DD and food
sensitization at each time point. A higher DD at 6 months decreased the
odds of food sensitization at 1 year (OR 0.86, 95%CI 0.74-1.0, p=0.04),
2 years (OR 0.84, 95%CI 0.73-0.96, p=0.01) and over the first 10 years
combined (OR 0.92, 95% CI 0.88-0.98, p=0.01). Likewise, a higher DD at
9 months decreased the odds of food sensitization at 2 (OR 0.81, 95%CI
0.67-0.98), p=0.03) and 3 years (OR 0.83, 95% CI 0.7-0.98, p=0.03)
(Figure 1). A higher cumulative DD (6+9 months) decreased the odds of
food sensitization at 1, 2 and 3 years. There was no association between
DD at any age and food sensitization at 10 years.
Looking at covariates, maternal history of food sensitization, parity,
maternal prenatal smoking, later introduction of solid foods and eczema
in the first year of life were associated with food sensitization at one
or more time periods (supplementary file 1). In the multivariate model
(Table 2) a higher DD at 6 months decreased the odds of food
sensitization at 1 year (OR 0.79, 95%CI 0.65-0.96, p=0.02) and 2 years
(OR 0.84, 95% CI 0.73-0.96, p=0.01). A higher DD at 9 months decreased
the odds of food sensitization at 2 years (OR 0.79, 95%CI 0.65-0.96,
p=0.02), but not at any other age. A higher cumulative DD decreased the
odds of food sensitization at 1 and 2 years. There was no association
between DD at any age and aeroallergen sensitization at any age (data
not shown).
Overall our data demonstrates a reduced risk of food sensitization up to
2 years of age in those consuming a higher diversity of foods in the
first 6-9 months of life. On the whole, the association was maintained
when relevant confounding variables were accounted for, but was not
apparent at age 10 in either the univariate or multivariate analyses.
This aligns to some extent with research by Roduit et
al.6 and Markevyech et al.8, who
used similar methods and study designs, thus adding support to the
evidence base promoting dietary exposure in early life to prevent food
allergy, rather than dietary avoidance.
Roduit et al.6, included participants from 5 European
countries, with a similar sample size (n=856) to the FAIR cohort. They
assessed intake of a smaller number of foods (between 6-15) with
children followed up at 4.5 and 6 years. In addition to a reduced risk
of food sensitization, they also found that increasing infant DD was
protective of asthma and was positively associated with increased
expression of a marker for regulatory T cells, providing a plausible
mechanism. Markeyvech et al8, using data from the
German LISA cohort (n=2518), reported higher infant DD was associated
with decreased risk of allergic sensitization to aeroallergens up to age
15 years among children with early skin symptoms. It is not clear why in
the FAIR cohort, the reduced risk of food sensitization exists until age
2, but not until age 10, as in the case of clinical food
allergy9, or why we did not find an association with
aeroallergen sensitization. This may be due to a more pronounced effect
on the microbiome in the first 1-3 years of life than in later
childhood10. Early life microbiome has particularly
been associated with early life food allergen
sensitization11. Differences in aeroallergen
sensitization may be due to different range of allergens tested.
Strengths of this study are the prospective design, large numbers of
foods considered and low attrition rate. Although we considered a range
of confounding variables, there may be unknown or unmeasured factors
that we have not explored. In summary, higher food diversity during the
first 6-9 months of life is associated with a decreased risk of food
allergic sensitization up to 2 years of age in adjusted models and up to
3 years of age in unadjusted models in our population based cohort.
Kate Maslin PhD1,2,
Kaci Pickett MSc3,
Suzanne Ngo3,4 MD,
William Anderson3,4 MD,
Taraneh Dean PhD2,5,
Carina Venter PhD2,3,4
1.School of Nursing and Midwifery, University of Plymouth, Plymouth, UK.
2.The David Hide Asthma and Allergy Centre, St. Mary’s Hospital, Isle of
Wight, UK
3.University of Colorado School of Medicine, Colorado, USA
4.Allergy and Immunology Section, Children’s Hospital Colorado,
Colorado, USA
5.University of Brighton, Brighton, UK
Acknowledgements: We thank all the FAIR study participants and
families for their participation over the years.
Impact statement: Higher food diversity during the first 6-9
months of life is associated with a decreased risk of food allergic
sensitization up to 2 years of age in adjusted models and up to 3 years
of age in unadjusted models, providing further evidence that a varied
infant diet should be encouraged.