INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The prevalence of childhood asthma is increasing (1, 2), although the underlying reasons for this remain obscure. An important familial component has long been recognized in asthma and atopy (3), although the relationship of familial predisposition to changes in asthma prevalence has not been studied. A difficulty in such studies is that the definition of asthma and, in particular, the relationship of asthma to wheezing illness in infancy and early childhood occurring only with upper respiratory tract infection remains controversial (4). Wheezing illness in older children and persistence of wheezing starting in the preschool years is associated with atopy (5), but prospective studies in infants and younger children have suggested that reduced small airways function, particularly in male children, is more important than atopy in predisposing infants and children younger than 2 yr of age to wheezing associated with viral infection (8, 9).

Historically, much childhood wheezing illness associated with clinical evidence of upper respiratory tract infection was diagnosed as "wheezy bronchitis," although this approach fell into disuse with the demonstration that it was often associated with underdiagnosis and undertreatment (10). Nevertheless follow-up studies have demonstrated better outcome in childhood (11) and adulthood (12, 13) for those with wheezy bronchitis compared with those with asthma, suggesting that this classification, although subjective and imprecise, has prognostic value. The increasing evidence that much lower respiratory illness in infancy and early childhood, previously labeled as wheezy bronchitis, differs from asthma in its etiology (4, 8, 9), and prognosis (12) suggests that wheezy bronchitis and asthma may be distinct clinical syndromes.

We have previously demonstrated independent effects of childhood asthma or wheezy bronchitis and atopy on outcome in middle age (13), leading us to speculate that clinical phenotype and atopy may have independent effects on outcome in families. Previous work has demonstrated the complex nature of the heritability of asthma and atopy and has suggested a significantly increased risk of asthma in children of parents with asthma and elevated total IgE levels (14). The influence of different parental wheezing syndromes (asthma or wheezy bronchitis) on outcome in the next generation has not been studied, nor has the effect of recent changes in childhood asthma prevalence (2, 15) on outcomes in children of families expected to be at high and low genetic risk. We hypothesized that the effects of parental asthma, atopy, and wheezy bronchitis on outcome in children differed. In order to study this we investigated outcomes for wheezing illness and atopy in the children of nuclear families based on highly selected middle-aged probands drawn from the follow-up cohort we have previously reported (12, 13). Probands with histories of persistent atopic asthma were selected in order to identify families with strong genetic predisposition to asthma and atopy in which children would be at high risk of developing asthma and atopy (14), whereas a comparison group at low risk was based on probands with neither childhood or adult symptoms nor atopy. In addition families based on probands with transient childhood wheezy bronchitis were also studied to investigate whether this syndrome has similar heritable effects to those in atopic asthma.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Probands were selected from subjects originally studied as children in 1964 (16). At that time, 287 children with a parentally reported history of wheeze were identified in a larger random community sample of 2,511 schoolchildren 10 to 15 yr of age. All of the children with a history of wheeze were reviewed by a single clinician and divided into 121 with a clinical diagnosis of asthma (defined as "recurrent dyspnea of an obstructive type without other demonstrable cause") and 167 with a clinical diagnosis of wheezy bronchitis (defined as "wheeze occurring only in the presence of infection"). Twenty-five years later all accessible persons within these groups together with a comparison group of 167 subjects selected from the nonwheezing group within the original study population were recontacted. On the basis of the information obtained at the 25-yr follow-up study (12), we selected 79 probands, known to have children, divided into four groups based on childhood history:

  Group 1. Asthma diagnosed in childhood, continuing symptoms in middle age, and atopic in middle age (n = 20).

  Group 2. Wheezy bronchitis diagnosed in childhood, asymptomatic in middle age and atopic in middle age (n = 12).

  Group 3. Wheezy bronchitis diagnosed in childhood, asymptomatic in middle age, and nonatopic in middle age (n = 18).

  Group 4. Asymptomatic both in childhood and middle age and nonatopic (n = 29).

The remaining subjects within the follow-up cohort either could not be classified using these criteria or did not have children, excluding them from consideration for the present study. This approach was used in order to study families based on clearly phenotyped parent probands assigned to discrete groups based on natural history. Given the uncertainties involved in classification of the parent probands, this approach concentrated on clearly phenotyped probands in order to reduce the potential for misclassification.

Atopy was regarded as the presence of one or more positive skin prick tests with a weal diameter of 2 mm or more greater than the negative control to cat, grass, or house dust mite, or a positive RAST test (> 0.35 IU/ml) to these allergens, or a total serum IgE level greater than 100 IU/ml after the definition of Cookson and coworkers (17).

All 79 probands were contacted by letter and telephone. Sixty-nine probands and their families agreed to participate (87.3%). Of the remaining 10 probands seven refused to participate, one had died, and two had moved out of the study area. One of the 69 probands declined to participate, although other members of the nuclear family were seen. One proband had divorced and remarried; both his current and previous spouses and children of both marriages were seen. In the 69 families studied one spouse had died (of a nonrespiratory cause), one spouse refused to participate, and six spouses could not be contacted because of divorce or separation. Of the 143 children from these families 133 agreed to participate.

After initial contact subjects were seen at home by a research nurse, and a semistructured questionnaire based on the ATS DLD78A questionnaire (18) was administered. Symptoms were reported as being experienced ever or as current (within the previous 12 mo) and the presence or absence of asthma by self-report. Subjects were thereafter invited to attend the testing center for skin prick testing for common aeroallergens and venepuncture for total and specific IgE determination. Children were interviewed with their parents to obtain information about events in infancy and early childhood.

Skin Prick Testing

Skin prick testing was carried out using the method of Hendrick and coworkers (19) using house dust mite (D. pteronyssimus), cat, and mixed grasses (Dome Hollister Steir, Spokane, WA). These allergens have previously been shown to provide a satisfactory screen for skin test reactivity in this population (20). Skin tests were performed on 65 probands, 58 spouses, and 119 children and were regarded as positive if the maximum weal diameter was at least 2 mm greater than the negative control for any of the allergens tested.

Total and Specific Serum IgE Measurement

This was carried out using standard radioimmunoabsorbent (RAST) techniques (Pharmacia Diagnostics, Milton Keynes, UK) for house dust mite (D1), grass (G6), mixed epithelia (EX1), and mixed molds (MX1). Blood for total and specific IgE determination was obtained from 64 probands, 55 spouses, and 95 children. Four probands, seven spouses, and 38 children refused venesection. Total IgE values were log-transformed prior to analysis to normalize the distribution. Specific IgE determinations were regarded as positive if the result exceeded 0.35 IU/ml.

The study was conducted over a 16-mo period (June 1993 to October 1994). It was not possible to perform skin prick testing, total IgE measurement and RAST testing in a single season or at a single time of day, although greater than 80% of the samples were taken between 2:00 and 9:00 P.M. Analysis of the season in which the subjects' atopic status was determined revealed no significant or consistent effect attributable to season.

Statistical Analysis

Data were analyzed on a personal computer using the package STATA (Stata Corporation, Austin, TX). Chi-square tests were used to seek associations between categorical variables as appropriate. Analysis of variance (ANOVA) was used to compare groups where data were normally distributed. The Mann-Whitney two-sample statistic and the Kruskal-Wallis rank sum tests for equality of populations were used to compare groups where data were not normally distributed. Multivariate regression techniques were used to investigate the effect of parental sex and atopic status on outcomes for atopy in the children. Ethical approval was obtained from the Joint Ethical Committee of Aberdeen University and Aberdeen Royal Hospitals NHS Trust.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

There were no significant differences between the children on whom complete and incomplete data were available in terms of age, sex, grouping, or history of wheezing symptoms or cough (Table 1). Outcomes for symptoms and atopy in the 95 children on whom complete data were available are shown in Table 2. Skin prick testing and total and specific IgE determination data were available for both parents of 83 of these children. Current wheezing symptoms, defined as wheeze within the previous 12 mo, were significantly less frequent in children of both atopic and nonatopic wheezy bronchitic probands (Groups 2 and 3, respectively) than in either of the other two groups (² = 11.6, 3 df, p = 0.009). Wheezing symptoms at any time and cough within the previous 12 mo were also less frequent in the children of wheezy bronchitic probands, although these differences did not reach significance. Similarly, current asthma (defined as self-reported asthma with symptoms within the previous 12 mo), was less common in the children of wheezy bronchitic probands, although the differences were not significant. There were no differences between the children of atopic wheezy bronchitic probands (Group 2) and those of nonatopic wheezy bronchitic probands (Group 3) in symptomatic outcome. The children of asymptomatic nonatopic probands (Group 4) had a high prevalence of current wheezing symptoms (33%) and of ever having had wheezing symptoms (36%).

There were no significant differences in the prevalence of at least one positive skin prick test or at least one positive RAST test between the groups of children studied. Raising the threshold for skin test positivity to a weal size or at least 3 or 4 mm greater than the negative control did not affect this result. Similarly, there were no differences between the groups in the cumulative weal size to cat, house dust mite, and grass or in the total RAST test scores obtained. Overall IgE responses to specific allergens were similar in the children of the four groups studied. Log-total IgE levels differed significantly between the groups of children studied. The highest log-total serum IgE levels were found in the children of atopic probands, both asthmatic and wheezy bronchitic, and were significantly higher than the values in children of nonatopic probands (p = 0.0002, one-way ANOVA).

Results for the parent probands are presented in Table 3. The majority of probands were male. As expected, current wheeze, skin prick test and RAST positivity, and total IgE levels were higher in the probands in the asthmatic and atopic groups. Three probands in Group 2 reported current wheeze, and several probands in Groups 3 and 4 had positive skin prick and RAST tests representing changes in their wheezing and atopic status since the follow-up study in 1989 on which the selection criteria were based. As expected log-total serum IgE levels and the prevalence of skin prick and RAST test positivity were higher in the parent probands selected on the basis of atopy (Groups 1 and 2). There were no significant differences between groups among the spouses (Table 4).

Multivariate logistic regression analysis using positive skin prick or RAST tests in the children as the dependent variable showed weak associations between these outcomes in children and positive skin prick or RAST tests in their parents. There were no significant effects of proband sex observed. The high levels of skin prick and RAST test positivity observed in the children of nonatopic asymptomatic probands were not explained by the influence of atopic spouses. A similar proportion of children of atopic and nonatopic spouses were skin test positive (six of 14 children of atopic spouses compared with 10 of 22 children of nonatopic spouses) or RAST test positive (six of 14 compared with 11 of 22 children of nonatopic spouses).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Several studies have shown a strong relationship between atopy, assessed either as total serum IgE levels or positive skin prick tests and wheezing illness both in mid-childhood (21) and in adulthood (22), although the relationship does not appear to be present for total serum IgE levels and wheezing in the first 2 yr of life (9). Atopy, particularly when more severe, has also been associated with poorer outcome in some, but not in all, long-term follow-up studies of asthmatic subjects (23, 24), an effect that may be independent of the nature of childhood wheezing illness (13). We have shown a significantly lower prevalence of current wheezing symptoms, although not of lifetime wheezing symptoms, and, in males, poorer lung function in the children of both atopic and nonatopic wheezy bronchitic probands compared with children of asthmatic probands (25). This suggests that there are significant familial associations in this pattern of transient wheezing illness independent of atopy.

There is evidence from family studies of significant heritability of total serum IgE levels (3) that may be independent of specific IgE responses (26) and that may involve different genetic mechanisms (27). Linkage studies have suggested that total serum IgE levels are under the control of a locus on chromosome 5q (28, 29), whereas a wider atopic phenotype has been linked to chromosome 11q (30), although this work has proven more controversial (31). As can be seen from Tables 3 and 4, there appears to be a closer familial aggregation of total IgE levels than of specific IgE responsiveness. Although total serum IgE levels were significantly higher in children of atopic probands, specific IgE responsiveness to cat, house dust mite, and grass was similar in children of both atopic and nonatopic parent probands.

The absence of clear familial aggregation in specific IgE responsiveness supports the concept that this may be determined by different genetic and environmental factors than those influencing total IgE levels. Smoking and age are known to influence total IgE levels and specific IgE responsiveness in some groups (32), but it is unlikely that these factors introduced major confounding effects into the present study as levels of current and lifetime smoking were similar among the parents and the ages of parents and their children within the four groups were similar. There were no significant differences in the spouses of atopic asthmatic and nonatopic asymptomatic probands with respect to wheezing symptoms or atopy, suggesting that the influence of the spouses did not account for the unexpectedly high levels of current wheeze or specific IgE responsiveness in their children. Multivariate analysis of skin prick and RAST tests in the children similarly failed to demonstrate any confounding effect of either proband sex or atopy in the spouses. Selection bias also appears unlikely given that greater than 85% of the families approached participated in the study and greater than 90% of the children in these families participated.

A major issue in the present study was the use of highly selected parent probands as the basis for the study of families. This approach limited the size of the population available for study, but it also reduced the potential for bias because of misclassification of the parent probands based on information obtained in the previous studies. The small numbers of children studied mean that conclusions must inevitably be tentative, but the use of clearly phenotyped parent probands allows a clear focus on the "core" clinical syndromes of interest.

Population studies in Aberdeen have demonstrated a significant rise in asthma prevalence between 1964 and 1984 (2, 15). The present study is of interest as it examines important and well-characterized subgroups of the original 1964 population together with outcomes in their children. The atopic asthmatic probands comprised all traceable persons who had childhood asthma within the original cohort of 2,511 children and who had continuing symptoms and atopy when studied 25 yr later (12, 13). Evidence from other follow-up studies suggests that this group is likely to comprise those individuals with the most severe and persistent asthma (24, 33). Children of parents within this group would be expected to be at significantly higher risk for the development of wheezing illness and atopy than children of asymptomatic nonatopic parent probands (14). As expected prevalence of symptoms and total and specific IgE responsiveness was high in this group; however, the most intriguing findings of the present study are the high prevalence of specific IgE responsiveness and wheezing symptoms in children of nonatopic asymptomatic probands in whom genetic "risk" would be expected to be much lower against the background of a significant rise in asthma prevalence in the wider population. Although the prevalence of wheezing symptoms within this group was high, it was nevertheless consistent with the findings of a large postal questionnaire survey of Aberdeen schoolchildren carried out during the same time period (15).

Our observations suggest that either the familial recurrence risk for asthma and atopy is significantly less than is generally assumed, which appears unlikely given the findings of many other investigators, or that there has been a significant rise in the prevalence of asthma and specific IgE responsiveness among children without a family history of asthma in the space of one generation. This latter explanation would be consistent with the increasing prevalence of childhood asthma observed in this, and in other, populations during this period (1, 2). Environmental changes during this period may, therefore, have led to the development of symptoms in large numbers of children who do not have a strong genetic predisposition to asthma or atopy. Our finding that total IgE levels in parents and their children correlate well in both atopic and nonatopic groups further reinforces the evidence from both segregation analyses and linkage studies (26, 28, 29) that high total IgE levels may well be a marker for families at higher genetic risk for asthma and atopy (9). However, our results suggest that total IgE levels may be a poor marker for risk of asthma and specific IgE responsiveness in children without a family history of asthma or atopy. It is possible that there may be different genetic determinants of asthma and atopy in this group, perhaps related to specific IgE responsiveness (34) or that asthma in this group is primarily environmentally determined. Future studies of the genetic influences on asthma and atopy need to take into account the potential phenotypic heterogeneity of these conditions and the effects of continuing changes in prevalence.