This study has several interesting findings. First, we found evidence of sensitization to foods or aeroallergens in nearly half of this unselected rural Chinese cohort of twins. The prevalence of sensitization may have been even higher if regional pollens, such as trees and grass, had been included. Moreover, food allergy is more common among children than adults and is most prevalent among children in the first few years of life. The children in our study were older (>10 years). Therefore, the rate of food sensitization may have been even higher in this population if younger children had been studied. In contrast to findings among the general population17-20
, birth order within each twin pair did not have a significant effect on the rates of sensitization to the vast majority of allergens tested. This may be partly explained by the common prenatal environment that each twin pair shared, regardless of whether the twins were MZ or DZ.
The prevalence of atopy in our cohort is unexpected based on our hypothesis, and much higher than that found in other rural populations in Europe.21-23
Since 1999, almost all of the 15 studies performed in rural areas in Europe21
have reported a lower prevalence of atopic sensitization among farm children compared with non-farm children.22
Growing up on a farm has been found to have a protective effect on atopic sensitization as well as allergic disease. In the PARSIFAL23
study, atopic sensitization (using in vitro
methods) was identified in 22.7% of farm children, which was lower than our findings. It has been postulated that the protective farm effect is due, at least in part, to exposure to livestock and farm animals. However, in our multivariate logistic regression, the presence of farm animals was not protective for atopic sensitization. However, this may be accounted for by the pervasive exposure (≥75%) to farm animals. We cannot adequately assess rural vs. urban differences among this cohort. Other proxies for microbial exposures, including exposure to pets, mice, and cockroaches, were also not found to be protective factors for atopic sensitization. We did not, however, explore whether our study subjects had these same exposures in early childhood. We hypothesize that they may have, given the limited mobility of this population.
The reasons for the high prevalence of sensitization may in part be due to the presence of a pan-allergen. Of the four most common sensitizing allergens, three share a homologous protein: tropomyosin. This could explain the high prevalence of sensitization to these particular allergens. Cockroaches are commonly found in homes worldwide but specific allergen levels within mainland China are not well documented in the literature. Studies have previously established detectable levels of house dust mite allergen in Hong Kong,24
and mainland China.26
However, only 6.5% of participants (n=138) were sensitized to all three allergens (shellfish, cockroach, and dust mite) in our study. There are likely to be other reasons that are yet to be determined.
Second, in this study cohort, the most common sensitizing food allergens were shellfish (16.7%) and peanut (12.3%), with prevalence higher in children than adults (p<0.001 for both allergens). This is consistent with previous literature. The common food allergens in China are fish, shrimp, and crab.27, 28
One retrospective review of anaphylaxis within a Hong Kong emergency department (in a predominantly ethnic Chinese community) reports that 71% of all food-related reactions were due to seafood.29
However, the prevalence of seafood allergy in mainland China is less well defined. In our cohort, shellfish mix (clam, crab, oyster, scallops, and shrimp) was the most commonly sensitizing food extract. In contrast, sensitization to fish mix (cod, flounder, halibut, mackerel, and tuna) was much less common (2.0%). It is possible that these types of fish are not the most commonly ingested fish among this population.
In general, MZ twins had higher concordance rates than DZ twins. However, we did not find a substantial difference between MZ and DZ twins with regard to atopy or atopic disease. Our findings suggest that zygosity itself is not a risk factor. This is supported by Strachan et al30
who previously demonstrated that genetically identical twins are often discordant in their expression of atopy.
Finally, and most interestingly, we found a dramatic dissociation between skin test positivity and prevalence of allergic disease among this rural Chinese cohort. A recently published ISAAC study31
found that the link between atopy and asthma symptoms is strongest among populations with increased economic development. The average annual net income of Anqing farmers in 2006 was 2970 Chinese Yuan per person (~US$407) while in the urban Anqing population, it was 9483 Chinese Yuan per person (~US$1299). Overall, Anqing (both rural and urban) has experienced economic growth at a rate of over 10% annually. Compare this to the gross national income per capita for Hong Kong, which is approximately US$25,000. 31
Hong Kong is a highly westernized city with a subtropical climate. In contrast, Anqing consists of rural farming communities and is located 943 km (586 miles) north of Hong Kong. Despite the high rate of atopic sensitization in our cohort, the prevalence of self-reported physician-diagnosed allergic disease was quite low, as expected for the region.
In contrast to the high rates of sensitization to foods (nearly 30% of children and over 20% of adults), only 3.5% reported a history of FA. Prior studies in Western populations have established that self-reports of FA generally overestimate (by up to 4 times) the true prevalence,32, 33
indicating that the true prevalence of FA in this Chinese cohort is likely to be even lower. Also it has been reported that the diagnostic capacity of SPTs to foods is highly variable.34, 35
Food challenges were not performed to confirm the diagnosis in our study. However, when specific food allergens were examined, for example, peanut and shellfish, only a small minority of subjects with self-reported FA to these allergens had symptoms in addition to positive skin test results. Therefore, despite relatively high sensitization rates to foods, the rate of disease (based on self-report combined with positive skin test) was comparatively quite low (<1%), thereby further indicating that skin testing to foods has a relatively high false positive rate. We do not believe that parasitic disease accounts for the dissociations, based on lack of peripheral eosinophilia, a crude marker for parasitosis.
Although this cohort was tested to only 5 aeroallergens, the rate of sensitization was quite high (38.5%) in contrast to a low prevalence of reported asthma (current or past) of <1% in this unselected Chinese rural population. The clinical significance of wheal sizes of positive SPTs to aeroallergens is unclear. The rates of allergic rhinitis, hay fever, and pet allergy were comparable to that of asthma.
This study is limited because only sensitization was evaluated and the prevalence of allergic disease in this cohort was not confirmed, particularly for FA. However, food challenges would have been impractical and costly in such a large cohort. The Multi-Test II device has been shown to produce larger wheal sizes than the ALK lancet for example36
and may have the potential to increase the rate of falsely positive results.37
Overall, the wheal sizes were small for the foods and aeroallergens tested in our study. The Multi-Test II is easy to administer and the technique has greater potential to be consistent among multiple operators across various locations and allows for the application of multiple extracts with one application.
In conclusion, this is the one of the largest unselected cohorts of children and adults in China to be assessed for sensitization to both foods and aeroallergens. Sensitization was common in this agricultural Chinese population, particularly to shellfish, peanut, dust mites, and cockroach. Atopy was observed in nearly half the total number of subjects. The prevalence of atopic sensitization varied by age and by sex. While sensitization was common, the prevalence of allergic disease was comparatively quite low. It remains to be determined if the rate of allergic disease will increase in this rural community with expanding urbanization and lifestyle changes that become more akin to those in westernized countries.