INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Although it is often claimed that asthma is not a single disease entity, clear separation of subphenotypic groups has not yet been accomplished. This status of unclear segregation is likely attributable to (1) variability of signs and symptoms with age and environmental exposure patterns, (2) similarities in signs and symptoms to some other illnesses, especially lower respiratory illnesses (LRIs) early in life, and (3) the lack of knowledge in regard to the major mechanisms providing the basis for the pathogenesis of this (or these) disease(s). Available genetic evidence suggests that asthma is a complex disease with inheritance patterns suggestive of an oligogenic or polygenic basis (1, 2) and with critical environmental factors important in the degree to which the disease is manifested. The identities of these genes and factors have remained elusive.

One approach to unraveling the heritable genetic components contributing to a complex disease is to examine the heritability patterns of the individual phenotypic characteristics that are invariantly or most closely associated with the disease. This approach assumes that these characteristics may have less complex and thus more directly definable genetic bases. Although these characteristics are often referred to as intermediate phenotypes, we will use the phrase "phenotypic markers" here so as to avoid confusion with "subphenotypes" of disease, in which at least partially distinct clusters of the phenotypic markers serve to identify disease subgroups.

Regarding phenotypic markers, several allergic characteristics have been shown closely associated with asthma, although none has been invariantly associated. Several attempts have been made to distinguish allergic (or atopic) asthma from nonallergic asthma (3), but this division remains controversial. That is, upon dividing asthmatics into those with and without skin test reactivity to common aeroallergens, both groups still showed a close association to total serum IgE (4). Without evidence for other distinguishing characteristics, the possibility remains that all asthma is atopic and that not all of the appropriate allergens were included in the battery used for testing.

For those asthmatics with positive skin test reactivity, several studies led to the realization that asthma was preferentially associated with sensitization to a group of antigens distinct from those associated with allergic rhinitis or other forms of allergic disease (5). Studies in many if not most parts of the world have shown house dust mite to be the common dominant antigen associated with asthma, although pet-related allergens, cockroach, and mold are also part of this group of asthma-related allergens (8). In some environments, however, and especially desert environments, house dust mite is of lesser importance and the mold Alternaria appears to be a dominant asthma-linked allergen (12). Indeed, in our own U.S. Southwest desert environment, Alternaria is the only sensitizing allergen consistently associated with asthma in children raised in this environment (14).

Respiratory viruses have long been suspected of participating in the development of asthma, but separating a role for inducing the development of asthma from that of initiating asthma attacks (in those with chronic asthma) has proved difficult. Additional complicating factors include the suggestion that at least some respiratory viral illnesses in early life may provide protection from asthma and/or allergy (16). Others either increase the risk for subsequent asthma or are themselves the first symptomatic expression of asthma (17), and still others appear to only mimic the symptoms of asthma early in life but are not associated with the symptoms after the first 3 yr of life (18).

Markers of disease (such as IgE and skin test reactivity for asthma) may or may not be linked genetically to involvement in disease susceptibility. That is, the markers might represent allelic forms of inherited genes or they might be induced by environmental stimuli or develop secondarily as a result of some aspect of the disease process. One way to test for a possible genetic contribution of a marker to disease induction would be to determine, first, if inheritance patterns suggest Mendelian transmission of the marker from parent to the child and, second, if the presence or level of the marker in the parent increases the risk of disease in the child. An example of the first step is the demonstration that total serum IgE levels appear to be regulated via pathways that have Mendelian inheritance patterns (19). Nevertheless IgE concentrations in the parent have not yet been shown to be associated with asthma in the child, and, in fact, one recent study casts some doubt on this possibility (22).

We herein test whether parental serum IgE level and parental skin test reactivity show significant contributions to asthma in the child. This is tested in the context of two phenotypically distinct subgroups of asthma. As previously reported, asthma in our desert environment shows a strong association only with the allergen Alternaria (14) and yet this association occurs in only about half of the asthmatic children at age 6. Thus, an opportunity was provided to divide asthma into those with and without sensitivity to Alternaria and examine both groups for other phenotypic differences and for evidence of parental influences in each type of asthma. We hypothesized that Alternaria-negative asthma is associated with increased LRIs early in life and is not linked to parental asthma whereas Alternaria-positive asthma is linked to parental asthma and parental IgE level. We tested these hypotheses in the children and parents of the Tuscon Children's Respiratory Study (23).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

The families included in this report are participants in a longitudinal prospective assessment of childhood respiratory health, the Tucson Children's Respiratory Study (23). A total of 1,246 infants were enrolled at birth between May 1980 and October 1984. Physician-diagnosed active asthma in the parents was determined from the enrollment questionnaire. Children were followed for the development of LRIs through age 3. At the first sign of an LRI, the parents were instructed to take the child to see the physician and signs and symptoms of the illness were recorded. The wheezing LRI variable used in the analyses herein was based on the age of the first wheezing LRI in the first 3 yr.

When the children were an average of 6 yr old (mean ± SD, 6.2 ± 0.9), the child and both parents were skin prick tested to local aeroallergens, a blood sample was collected for total serum IgE analysis and a detailed health questionnaire completed. Active asthma in the child was ascertained at age 6 by parental report of a physician diagnosis of asthma and by at least one exacerbation of asthma during the previous year. The age of asthma diagnosis was also taken from the questionnaire obtained when the child was 6 yr old. Data from the age 11 questionnaire were used only to assess the remittance patterns for the age 6 current asthma.

To assess influence of ethnicity, the children were divided into three groups based on parental ethnicity, those with Anglo parents (64.4%), those with Hispanic parents (15.4%), and those with parents of other (or mixed) ethnicities (20.2%).

Informed consent was obtained from the parents of participating children and the study was approved by the Human Subjects Committee of the University of Arizona.

Serum IgE Levels

Serum IgE levels were assayed by Phadebas Radioimmunosorbent Test (PRIST) using commercially available kits (Pharmacia Diagnostics, Piscataway, NJ). Total serum IgE determinations were completed on 534 children, 743 mothers, and 554 fathers. All samples were assayed in duplicate. The lower limit of detection in this assay is 0.1 IU/ml.

Allergy Skin Prick Tests

House dust mix, Alternaria alternata, Bermuda grass (Cynodon dectylon), careless weed (Amaranthus palmeri), olive tree (Olea europaea), mesquite tree (Prosopis glandulosa), and mulberry tree (Morus alba) were tested on 760 children, 776 mothers, and 613 fathers (all allergens obtained from Hollister-Stier Laboratories, Everett, WA). Tests were read at 20 min and the sizes of the wheals elicited by each allergen were recorded as the sum of the two diameters (in mm) measured at right angles to each other. Wheal sizes of 3 mm or greater (or in one case as noted 6 mm or greater) were considered positive.

Statistical Analysis

Statistical analyses were performed using the Statistical Package for the Social Sciences for UNIX version 4 (SPSS Inc., Chicago, IL). Statistical analyses for total serum IgE values were performed using log-normally distributed values but are reported as the geometric mean (GM) and geometric 95% confidence interval (95% CI) expressed in IU/ml. The 95% CI of the odds ratios (OR) were calculated using standard techniques (24). Logistic regression analysis was used to examine the independent effects of parental asthma and the presence of at least one wheezing LRI in the first year of life on the prevalence of asthma in the child. (Children who had nonwheezing LRIs were included in the No LRI group.) Risk factors were examined separately for Alternaria skin test positive and negative subjects. All p values of < 0.05 (two-tailed) were considered statistically significant for all tests.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Formation of Asthma Groups Based on Alternaria Responses

Among 741 children who had both questionnaires and allergy skin tests at age 6, the overall prevalence of asthma at age 6 was 10.8%. Positive Alternaria skin test responses were present in 17.1% of children. Asthma prevalence among those positive for Alternaria skin test was 31.5% (n = 40). Despite a much lower prevalence of asthma among Alternaria skin test negative children (6.5%), the Alternaria skin test negative group was substantially larger (n = 614 versus 127) and thus 50% of the asthmatic children were from this group (n = 40). For further analyses, the children were divided into four groups based on the presence or absence of a positive Alternaria skin test response at age 6 and the presence or absence of a physician diagnosis of asthma also at age 6. In Table 1, individuals who had and who did not have complete data sets are compared and found not to differ.

Phenotypic Characterization of Alternaria-positive and Alternaria-negative Asthma

Male gender was prevalent to a similar degree in the Alternaria-positive and -negative asthma groups (65% and 57%, respectively). Prevalence of maternal smoking was also similar and did not differentiate asthma from control groups (data not shown).

Alternaria skin test positive children with asthma (Alternaria-positive asthma; n = 27) were found to have 3-fold higher total serum IgE levels than Alternaria skin test negative children with asthma (Alternaria-negative asthma; n = 28): GM (95% CI) = 219 (130 to 367) versus 72 (37 to 140) IU/ml, respectively; p = 0.01. Nevertheless, both asthma groups had significantly higher IgE levels compared with their respective nonasthmatic control groups: GM (95% CI) = 107 (76 to 149) IU/ml for Alternaria-positive control groups (p = 0.02) and GM (95% CI) = 26 (22 to 30) IU/ml for Alternaria-negative control groups (p < 0.001); n = 63 and 396, respectively.

The results in Table 2 demonstrate that the two groups of asthmatics also showed marked differences from one another in skin test responses to each of the aeroallergens, with the Alternaria-negative asthma group much less reactive to all of the aeroallergens tested (overall reactivity 30% versus 85%, respectively; p < 0.0001). Nevertheless, neither asthma group was distinguishable from its respective nonasthmatic group with respect to the prevalence of the individual allergen responses.

Both asthma groups showed increased prevalences of wheezing LRIs in the first 3 yr of life compared with the nonasthma groups (Table 3). The two asthma groups did not differ from each other as to the overall prevalence of wheezing LRIs in the first 3 yr of life. However, the distributions for age at which the first wheezing LRI occurred were markedly different, with the Alternaria-negative asthma group having twice as many LRIs in the first year of life compared with the Alternaria-positive asthma group (p = 0.01). In contrast, the Alternaria-positive asthma group tended to have more LRIs in the third year.

The two asthma groups also differed with respect to the age at which the current asthma was diagnosed. As shown in Figure 1, asthma was most often diagnosed among Alternaria-negative children between ages 0 and 3 yr, whereas the peak period for diagnoses of asthma in the Alternaria-positive children was between ages 3 and 4 yr.

View larger version (39K): [in this window] [in a new window]   Figure 1.   The age of asthma diagnosis for children in Alternaria-positive (n = 40) and Alternaria-negative asthma (n = 40) groups at age 6.

Another difference in these two types of asthma related to differences in subsequent remission rate after age 6. Remission of the age 6 current asthma by age 11 was more frequent among the Alternaria-negative group (39%) compared with the Alternaria-positive group (9%; p = 0.003).

Parent-Child Relationships for Alternaria-positive and Alternaria-negative Asthma

The prevalence of asthma and asthma subphenotypes among children in relation to parental history of asthma is shown in Table 4. Alternaria-positive asthma was influenced by asthma in both the mother and the father. In contrast, for Alternaria-negative asthma, a significant influence of maternal asthma, though not paternal asthma, was evident.

From a clinical point of view, it is important to emphasize that among Alternaria skin test positive children, a maternal asthma history was associated with a 53% prevalence of asthma (10 of 19) and a paternal asthma history was associated with a 52% prevalence of asthma in the child (12 of 23) in contrast to a 23% prevalence for a negative parental history (19 of 83). For Alternaria skin test negative children, a maternal history of asthma was associated with a 21% asthma prevalence (11 of 53) in contrast to 5% without a maternal history (26 of 516, irrespective of paternal history).

Parental Total Serum IgE and Skin Test Reactivity in Relation to Asthma in the Child

As described previously, the prevalence of asthma at age 6 showed a strong relation to the total IgE level in the child at that same age. Furthermore, both the maternal and paternal total serum IgE levels were associated with the child's total serum IgE level (Table 5). These data are in accord with previous segregation analyses suggesting a pattern of codominant inheritance for a major gene regulating serum IgE in this population (1). Despite this close association between parental and child IgE level, there was, nevertheless, no increased risk for asthma in the children associated with the IgE level of either father or mother (Table 5).

It is evident from the data in Table 6 that positive skin tests in the parents showed a significant influence on skin test reactivity in the child. Like IgE, however, no significant association was evident between the prevalence of either type of asthma in the child and the presence or absence of skin tests in the parents. The data from this analysis yielded the same relationships (or lack thereof) if the criterion for a parental positive skin test reaction was increased to 6 mm. Further, for both parental skin test and the serum IgE relationships shown in Tables 4 and 5, the analyses were repeated limiting the population to only those with complete data for all variables (n = 340). Within the limited data set, the results were the same: significance for the parent-child relationships for skin test and total IgE but no evidence of a link for these parental variables to asthma in the child.

Familial Relationships Assessed by Logistic Regression

Applying logistic regression to these parent-child relationships provided confirmation of the bivariate findings described previously. Both maternal and paternal asthma contributed to risk for asthma in Alternaria skin test positive children (OR 3.4, p = 0.05 and 3.9, p = 0.02, respectively) with early life wheezing LRIs contributing additional risk (OR 4.5, p = 0.003). In contrast, among Alternaria skin test negative children, maternal asthma but not paternal asthma provided significant asthma risk (OR 4.3, p = 0.002 and OR 1.9, p = 0.8, respectively). Addition of wheezing LRIs in early life added an independent contribution of risk for asthma (OR 4.1, p = 0.004). These results were independent of group size with respect to including or not the children on whom we did not have sera.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In a previous study, we demonstrated the strong association of asthma with skin test reactivity to Alternaria (14), a ubiquitous mold that is both an indoor and outdoor allergen in the Southwest desert environment of Tucson, Arizona (25). We show herein that asthma also occurs in this population among children with negative responses to Alternaria skin tests (although at a much lower prevalence). Our results demonstrate that dividing asthma by the presence of Alternaria sensitization provides a segregation of several phenotypic manifestations and thus the division appears to serve as a means of classifying two distinct subphenotypes of asthma in children at age 6.

The Alternaria-negative asthma group had a lower mean level of total serum IgE, although both groups of asthma had increased levels compared with their respective nonasthma control groups. The data agree with previous reports of close association between the prevalence of asthma and the total IgE level (3). The basis of that association remains elusive; the data are consistent with either IgE contributing to the induction of asthma or asthma contributing to the induction of IgE.

The fact that just one allergen, Alternaria, was significantly related to asthma in our population provided a unique opportunity to carry out this study. It is not our intention to suggest that Alternaria has a more important link to asthma than have other allergens in populations from other environments. On the other hand, we emphasize that an Alternaria/asthma relation is not completely unique to our semi-arid region. Several recent studies have emphasized that this allergen is also important (though not necessarily as the only allergen) in many other environments (6, 12, 26, 27).

The prevalence of asthma was unrelated to sensitization to each of the aeroallergens tested when assessed within the respective Alternaria skin test groups. We cannot at the present time rule out the possibility that asthma prevalence may be linked to one or more allergens not being skin tested. However, the identified relationship of asthma to total IgE does not appear to be related to the known group of asthma-associated allergens. Though we tested only for house dust (mix) at age 6, both house dust mite and also cat sensitization were infrequent in this population at age 11 and neither showed a significant relation to asthma (14). Cockroach allergen has also been tested in another population from this environment and found to be an infrequent sensitizer (28). Thus, it will be of interest to determine if similarly comparing other populations for asthmatics with sensitization to the asthma allergen group versus asthmatics without in other locales provides similar phenotypic and parental history distinctions.

A part of our initial hypothesis was that Alternaria-negative asthma occurs via distinct pathways, possibly as a virus-initiated event begun early in life. We thus predicted that Alternaria-negative asthma would be more closely associated with a prevalence of early wheezing LRIs than would Alternaria-positive asthma. Table 3 indicates that when considered for all wheezing LRIs over the first 3 yr of life, this prediction was not borne out. However, analyzing the age patterns of LRI during those first 3 yr revealed rather striking differences for the two forms of asthma, with Alternaria-negative asthma showing twice the prevalence of wheezing LRIs in the first year of life compared with Alternaria-positive asthma.

An explanation for the mechanism by which wheezing LRIs occur less frequently in the first year of life for those children who will subsequently become part of the Alternaria-positive asthma group versus the Alternaria-negative asthma group (Table 3) is not immediately obvious. Possibly some factors suppress or enhance susceptibility to wheezing LRIs, respectively, in these groups. We can speculate that the noted maternal contribution to Alternaria-negative asthma might be supplying a factor that enhances susceptibility to early wheezing LRIs. Still another possibility might be that early wheezing LRIs themselves provide some form of suppression for the development of Alternaria-positive asthma. Finally, the low prevalence of early LRIs in the Alternaria-positive group may occur because those who later develop asthma are selected from among children with a higher resistance against first year LRIs. These children thus may have a certain level of susceptibility to Alternaria-positive asthma that is different from or maybe even suppressed among children suffering from first year LRIs.

This predominance of wheezing LRIs in the first year of life for Alternaria-negative asthma corresponds with the difference noted in the temporal pattern of diagnosis of asthma for this group (as shown in Figure 1). Alternaria-negative asthma was diagnosed in each of the first 3 yr of life more frequently than Alternaria-positive asthma. In contrast, Alternaria-positive asthma diagnoses were most frequent between 3 and 4 yr of age.

Alternaria-negative asthma was not only diagnosed earlier, it also remitted by age 11 at a greater rate than Alternaria-positive asthma. An interesting corollary to this greater rate of remission among Alternaria-negative asthmatics is a recent finding that the risk of subsequent persistent episodic wheezing associated with early life respiratory syncytial virus (RSV) LRIs decreases after age 6 (29). Thus we suggest that Alternaria-negative asthma is a subtype of asthma that occurs early in childhood, tends to remit near adolescence and is temporally linked to a susceptibility to viral wheezing LRIs occurring primarily in the first year of life.

In addition to phenotypic markers separating Alternaria-positive and -negative asthma, differences in the relationships to parental history of asthma were also noted. A familial inheritance pattern linking paternal and maternal asthma to asthma in the child has been previously reported in this and other populations (1, 18, 30). We demonstrate here that significant contributions from both parents occur only in Alternaria-positive asthma.

Although asthma risk appears to be inherited, roles for individual parental phenotypic characteristics in the transfer of asthma risk have not yet been established. Given the close association of total serum IgE with asthma within individuals (3) and studies supporting the transmission of a major gene regulating IgE level (19), we hypothesized that such an IgE regulating gene might thus contribute significantly to asthma heritability. The data presented here, however, indicate that parental total serum IgE levels showed no detectable relationship to asthma in the child. These data thus cast doubt on the possibility that the purported major gene regulating total serum IgE level makes a significant contribution to the development of asthma in the child. These data, together with the strong IgE/asthma association evident within the children themselves, lead us to speculate that increased IgE concentrations in the asthmatic children may occur as a result rather than a cause of asthma. Thus it may be that additional regulatory mechanisms together with environmental exposures that permit or facilitate increased IgE levels are activated subsequent to the development of the chronic condition of asthma. In this regard, Burrows and coworkers recently reported data from a population distinct from that used herein, demonstrating that parental history of asthma was a much more important risk factor for asthma in the child than was parental serum IgE level (22). Certainly, a mechanism in which asthma "drives" IgE concentrations more than IgE drives asthma can also account for the reported complete lack of asthma among those individuals with the lowest IgE levels (3).

Like IgE levels, skin test reactivity to aeroallergens in the parents was shown to be linked to skin test reactivity in the child. But for this variable also, no association was detectable between skin test reactivity in the parents and asthma in the child. These findings argue that the inherited susceptibility to aeroallergen sensitization is not an important causative factor in the development of asthma.

In contrast to Alternaria-positive asthma, Alternaria-negative asthma was found to be associated with the presence of maternal (but not paternal) asthma. Mechanistic bases have been found for gender-specific inheritance patterns that relate to processes of gene imprinting and/or transfer of mitochondrial DNA (31). However, environmental effects are also possible. Other than ruling out maternal smoking as an explanation of the maternal influence, we have no data at this time to identify the mechanisms of the maternal effects. These exclusively maternal effects might occur as in utero influences.

Further studies are needed to identify more precisely the underlying mechanisms responsible for the differences in parental influences on the two types of asthma described here. Whatever the mechanisms, at the present time, these differences in parental asthma influences add further to the discriminating pattern for these two subphenotypes of asthma. Whether differences in the influence of parental asthma provide similar subphenotype categorization for asthma in other locales (and with different asthma-associated allergens) awaits further studies, as does the elucidation of the biologic bases for those influences.

In summary, asthma in children who were not sensitized to Alternaria appears to be phenotypically distinct from asthma in children skin test positive for Alternaria. Alternaria-negative asthma was accompanied by lower levels of total IgE, independence from skin test reactivity to aeroallergens, greater prevalence of wheezing LRIs in the first year of life, earlier asthma diagnosis, and an increased asthma remittance rate by age 11. Nevertheless, both forms of asthma had higher IgE levels compared with nonasthma control subjects. Both maternal and paternal asthma increased the risk of developing Alternaria-positive asthma in the child. However, evidence that this parental contribution to asthma risk occurred via inherited factors regulating IgE levels or skin test reactivity could not be demonstrated, weakening the supposition that the genes regulating these characteristics in the general population might be significant contributors to asthma. For asthma among Alternaria-negative children, asthma in the mother but not the father provided increased risk, indicating either a non-Mendelian mode of inheritance or an environmental, possibly an in utero, effect.