Skin test reactivity and number of siblingsBMJ 1994; 308 doi: https://doi.org/10.1136/bmj.308.6930.692 (Published 12 March 1994) Cite this as: BMJ 1994;308:692
- Universitatskinderklinik, Lindwurmstr 4, D-8000 Munchen 2, Germany.
- a University Children's Hospital, Munich, Germany Respiratory Sciences Center, Department of Pediatrics, University of Arizona, Tucson, Arizona, USA University Children's Hospital, Leipzig, Germany Research Centre for Environment and Human Health, Medis Institute, Munich, Germany University Children's Hospital, Halle, Germany
- Correspondence and reprint requests to: Dr
- Accepted 16 November 1993
Objective: To investigate the relation between skin test reactivity in children and number of siblings.
Design: Cross sectional survey among school-children aged 9-11 years. Skin prick tests in the children and self completion of written questionnaire by their parents.
Subjects: 5030 children in Munich and 2623 children in Leipzig and Halle, Germany.
Main outcome measures: Atopic status assessed by skin prick tests.
Results: After possible confounders were controlled for, the prevalence of atopic sensitisation decreased linearly with increasing number of siblings (odds ratio=0.96 for one sibling, 0.67 for five or more siblings; P=0.005). In atopic children the severity of the skin test reaction as assessed by the weal size was not associated with the number of siblings.
Conclusions: Factors directly or indirectly related to the number of siblings may decrease the susceptibility of children to become atopic. Thus, declining family size may in part contribute to the increased prevalence of atopic diseases reported in Western countries over the past few decades.
Public health implications
Public health implications
The prevalence of atopic diseases is on the increase
The development of hay fever has been linked to decreasing number of siblings
This study shows that an objective measure of atopy (skin test reactivity) decreases with increasing numbers of siblings
Increased exposure to infections early in life through siblings or socioeconomic factors related to the family size may explain these findings
The increase in the prevalence of atopic diseases over the past few decades may in part be associated with decreasing family size and higher socioeconomic status
A strong inverse association between the number of siblings and the prevalence of hay fever in British children was reported by Strachan.1 As the presence of older siblings had a stronger effect than the presence of younger siblings, the author suggested that factors such as viral infections early in life may prevent the development of allergic sensitisation. His observations, however, were based on parents' answers to a questionnaire. It could be argued that the effects observed may have been affected by recall bias, as parents with many children may not remember a relatively mild disease such a hay fever as accurately as parents with only one or two children.
The aim of this report was to investigate the relation between the number of siblings and an objective measure of atopic sensitisation in children living in eastern and western Germany. We performed skin prick tests in 9-11 year old schoolchildren in a survey in Leipzig and Halle, East Germany, and Munich, West Germany, and related these findings to answers to a parental questionnaire on household size, socioeconomic status, and occurrence of disease.
Study areas and populations
In former West Germany, all fourth grade pupils (n=7445) at all primary schools in Munich, a city of about 1.3 million inhabitants located in the south-western part of the country, were included in the study. All schoolchildren (n=3105) attending the fourth grade of a random sample of 39 schools in Leipzig and of 23 schools in Halle were studied in former East Germany. Leipzig, a city of about 535 000 inhabitants, is located in the southeastern part of the country in close proximity (35 km) to Halle, a city of about 300 000 inhabitants. These two cities are heavily polluted due to private coal burning and industrial emissions, whereas Munich has moderate industry but heavy car traffic.2 In all three study areas, climatic conditions are very similar.
Children in Munich were studied between September 1989 and July 1990 and children in Leipzig and Halle were studied between October 1991 and July 1992. The main purpose of both surveys was to determine the prevalence of asthma and allergies in western and eastern Germany.6 Therefore, field workers performing the skin prick tests were blind to the hypotheses being tested.
A self administered questionnaire was distributed through the schools to the parents. The questionnaire had been developed according to international recommendations.3 Questions concerning diagnoses and typical symptoms of respiratory and allergic disorders in the child and the parents were included. In addition, the sociodemographic characteristics of the family, such as the number of siblings, the parents' education, and the presence of pets at home, were assessed by questionnaire.
Parents were asked how many siblings their child had; how many years of education the mother and the father of the child had; whether there were pets at home; who smoked in the household; and whether a member of the family had ever had asthma, hay fever, or eczema (atopic dermatitis). Children with one or more first degree relatives with any of these diseases were defined as having a family history of asthma or atopy.
Skin prick tests
The sensitivity to six common aeroallergens (Dermatophagoides pteronyssinus, grass, birch and hazel pollen, cat and dog dander) was assessed by skin prick tests; the same multitest device (Stallerkit) was used in all three study areas. Results obtained with this device had been validated in a pilot phase in Munich.4 In this pilot study, weal reactions obtained with the multitest were compared with weal reactions elicited by the puncture method using the same allergen extracts in 31 atopic and 33 non-atopic children. Validation of results of skin tests by radioallergosorbent testing has been reported and discussed elsewhere.4
A child was considered sensitised to a specific allergen if a weal reaction of 3 mm or more to this allergen was present, the reaction to the negative control being previously subtracted from that to the respective allergen. Children with a weal reaction >=3 mm to one or more of the six allergens tested were considered to be atopic.
The prevalence of atopic sensitisation was calculated for numbers of siblings in both West and East Germany. The Mantel-Haentzel statistic was computed to estimate statistical significance over all strata. A regression analysis was used to assess the effects of the number of siblings on the weal size of skin prick test reactions in atopic and non- atopic subjects. A categorical logistic regression model5 was used to assess the independent effect of the number of siblings on atopic sensitisation after possible confounders such as a family history of asthma or atopy, sex, the socioeconomic status of the family, the presence of pets at home, and the study area were simultaneously controlled for. To assess a linear trend for the predictor variable, the logarithms of the odds ratio obtained from the logistic regression were plotted against the different levels of the predictor variable. Furthermore, linearity was tested by comparing the goodness of fit X2 of the categorical and linear model. If the fit of the data did not deviate significantly from the linear model, the slope of the regression line was obtained from a logistic regression model that included the same predictor variables plus the number of siblings as a scored ordinal term. The 95% confidence intervals around the regression line for each odds ratio were calculated on the logarithmic scale using the technique proposed by Rothman.6 Calculations of confidence intervals from logistic regression when the risk decreases with increasing “exposure” may be problematic.6 For this reason, odds (and their confidence intervals) of being “non-atopic” were first calculated from the logistic model. To make results more intuitively understandable, reciprocals of these odds and their confidence intervals are reported. The SPSS software package version 4.0 was used for all computations.7
All study methods had been approved by the ethics committee of the Bavarian Medical Association. Informed consent for skin prick tests was obtained from the parents.
The questionnaire was distributed to 1854 children in Leipzig, 1251 children in Halle, and 7445 children in Munich. Parents of 1500 (80.9%) children in Leipzig, 1134 (90.6%) children in Halle, and of 6537 (87.8%) children in Munich returned the questionnaire. The proportion of children without German nationality was 0.5% (eight) in Leipzig, 0.3% (three) in Halle, and 23.0% (1507) in Munich. To keep the study populations as similar as possible with regard to ethnic background, the analysis was restricted to children with German nationality. As children in Leipzig and Halle were very similar with respect to sociodemographic characteristics, prevalence of respiratory and allergic disorders, and results of skin prick tests, data for these two populations were pooled. Of the eligible German children for whom the questionnaire was returned, skin prick tests were obtained in 2335 (89.0%) children in East Germany and 4451 (88.5%) children in Munich. Subjects not tested for skin reactivity to allergens did not differ in their family size nor in their family history of atopy from subjects undergoing skin prick tests.
Atopic sensitisation is about three times as prevalent in West Germany as in East Germany.8 Skin test reactivity decreased significantly with age and was more prevalent when the tests were performed in the months January to May than in June to December (data not shown). The table shows the prevalence of atopic sensitisation in children with different numbers of siblings in both study areas. After the study area was adjusted for, the prevalence of atopic sensitisation decreased significantly with increasing numbers of siblings. This relation between number of siblings and atopic sensitisation was similar in boys and girls (data not shown). When cut off points of 4 mm and 5 mm were used a similar, significant relation was found, but when a weal size of >=2 mm was defined as atopy the association disappeared. This finding may be attributable to a better discrimination of atopic from non-atopic children when a larger weal size is used as cut off point. In atopic and non-atopic children, no association was found between the weal size of the skin prick test reaction and the number of siblings.
The relation of household size to atopic sensitisation was further assessed in a multivariate logistic regression analysis. The prevalence of atopic sensitisation was significantly correlated with the number of siblings (P=0.005; figure) when various factors were controlled for: family history of asthma or atopy (odds ratio=1.6; 95% confidence interval 1.4 to 1.8; P<0.0001); male sex (1.2; 1.1 to 1.4; P=0.0005); higher maternal education (1.1; 0.9 to 1.2; P=0.4); maternal smoking (0.8; 0.7 to 0.9; P=0.02); the presence of pets at home (0.9; 0.8 to 1.0; P=0.1), and the West German study area (2.6; 2.1 to 3.2; P<0.0001). Controlling for age and month of testing did not change the results.
Our results suggest that factors directly or indirectly related to the number of siblings decrease the susceptibility of children to becoming atopic. These factors do not affect the severity of atopic sensitisation as assessed by the weal size of the skin prick test reaction in atopic subjects.
Our findings are consistent with an earlier report by Strachan, who found a strong inverse correlation between the number of siblings and the prevalence of hay fever in British children.1 He speculated that declining family size, improvements in household amenities, prevention of viral infectious diseases, and higher standards of personal cleanliness - which are characteristics of what has been called Western lifestyle - may be associated with the increase of atopic diseases seen over the past few decades. The association was stronger for older siblings than younger siblings. Our data do not allow us to disentangle the effect of having older or younger siblings on the development of atopic sensitisation, since parents were asked only how many siblings their children had. Therefore, if children with older siblings are less susceptible to becoming atopic than are children with younger siblings, our results would probably have shown stronger effects had we been able to study the association between sensitisation and number of older siblings separately from that between sensitisation and number of younger siblings.
The factors that cause a decrease in the prevalence of atopic sensitisation with increasing numbers of siblings are not known. However, firstborn children are usually exposed to common infections after their enrolment in kindergarten or school, whereas children born subsequently are often exposed much earlier, through their siblings. Thus, having more siblings may be associated with an increased exposure to viral infections early in life.9 Recent immunological studies support the hypothesis of a downregulation of IgE production by viral and bacterial infections.*RF 10-12* It has been shown that two mutually exclusive patterns of sytokine release are produced by different clones of T helper cells, and it has been suggested that a predominant activation of TH1-like T helper cells in the course of recurrent viral or bacterial infections may prevent the proliferation of TH2 clones and the development of allergic disease.10 Our results suggest that having more siblings may prevent the development of atopic sensitisation. Once atopy is present, the number of siblings does not affect the size of the skin reaction. It is thus possible that factors associated with a large number of siblings may interfere with the process of allergic sensitisation in early life. The size of the weal, on the other hand, may depend on other genetic and environmental factors such as degree of exposure to allergens.
A higher number of siblings may not only reflect an increased exposure to viral infections early in life but may be related to other factors such as the socioeconomic status of larger families, nutrition, household amenities, or personal cleanliness.1 In addition to these environmental factors, inborn differences between children of increasing birth order may be important. Increasing parity is associated with changes in levels of sialic acid on neonatal lymphocytes.13 Immunogenicity of these cells can be altered by removal of sialic acid residues.14 It is not known whether these immunological changes are associated with any changes in disease prevalence.
Several reports have shown an increase in the prevalence of hay fever*RF 15-18* and eczema15,17,19 over the past few decades. Our results suggest that factors such as declining family size and subsequent decreased exposure to common infectious diseases may indeed be associated with the increase in the prevalence of atopic diseases in Western countries.
This work was supported by the Bundesministerium fur Forschung und Technologie (German Ministry for the Sciences and Technology). Dr Martinez was supported by a SCOR grant (HL 14136) from the National Institutes of Health.