INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A number of studies have addressed familial aggregation of asthma (1, 2). None of these studies has focused on lung function in these asthma families, choosing instead to focus on the diagnosis of asthma or its intermediate phenotypes (3).

A number of studies on familial aggregation of COPD have been conducted that have focused on lung function. Initial clinical reports have documented the familial occurrence of COPD (4), and subsequent reports described the similarity of pulmonary function levels among first degree relatives (5, 6). Larson and coworkers (5) studied familial aggregation of VC, FEV₁ and FEF₅₀ in 271 parental pairs and their 354 natural children with satisfactory pulmonary function data. All measures of pulmonary function aggregated in families, independent of family size, reported diagnosis of airflow obstruction and smoking. These investigators (5) subsequently studied 156 relatives of 61 patients with COPD and found a significantly higher prevalence of abnormalities in pulmonary function among relatives than among a control group of spouses. The level of pulmonary function and the degree of correlation were not described. In a series of studies conducted in the East Boston, Massachusetts population, the Respiratory Epidemiology Group at the Channing Laboratory demonstrated that familial aggregation of FEV₁ was significant when measured either as a percentage of a predicted value or as a score calculated from the data. Parent-child correlations were influenced by cigarette smoking, but controlling for cigarette smoking did not eliminate familial aggregation (6). A follow-up of 404 nuclear families in this same community, focusing on a random sample of children initially 5 to 9 yr of age and their parents, revealed familial aggregation of FEV₁ and FEF_25-75. Path analysis to determine the relative contributions of heredity and environment to individual levels of pulmonary function revealed that genetic heritability consistently contributed 41 to 47% in both parents and their children (7). Other population studies in the United States, including large studies in Baltimore (8) and in Tecumseh, Michigan (9), yielded results similar to those of East Boston.

More recently, Kauffmann and coworkers (10) performed a case-control study of genetic markers thought to be related to low levels of pulmonary function. They compared two groups of subjects: never smokers with low FEV₁ values and heavy smokers with high FEV₁ values. The subjects were men and women between 25 and 40 yr of age who had been selected from a larger study of outdoor air pollution in France (the PAARC Study). The investigators studied levels of ₁-antitrypsin, immunoglobins A, G, M, and E, and haptoglobin. They also assessed ABO blood group, Rh, ABH and Lewis secretor status and protease inhibitor, vitamin D binding protein, transferin, and HLA-A and HLA polymorphisms. A significantly higher proportion of nonsmokers with lower levels of lung function were ABH nonsecretors belonging to blood group O. A significantly lower proportion of non-O blood group nonsmokers with low pulmonary function were haptoglobin P1S carriers. This study suggested that specific genetic factors may be associated with low levels of pulmonary function. A follow-up study performed on these data again demonstrated familial aggregation of pulmonary function in the children involved in the PAARC study (11).

Taken together, the available data provided consistent evidence for familial aggregation of pulmonary function in both children and adults. Many important issues remain nonaddressed. First, the identified correlation among family members may be attributable to shared environmental factors, genetic factors, or both. Further studies are needed to dissect the genetic and environmental contributions to familial aggregation of pulmonary function, to identify responsible genes, and to test gene-environmental interactions. Second, studies thus far have generally focused on parent-child or sib-sib correlation. Few studies have simultaneously investigated the independent and combined relation of pulmonary function of parents and the first sibling to the pulmonary function of subsequent siblings, with adjustment for each person's important covariates such as age, sex, body size, and smoking status. It is well documented that asthmatics, as a result of increased airway responsiveness and ongoing airway inflammation, exhibit reduced growth of FEV₁ in childhood (12) and accelerated decline in FEV₁ in adulthood (13). Little information is available, however, on whether the pattern of familial aggregation of pulmonary function differs between asthma and random families.

We conducted a large genetic epidemiologic study in Anqing, China, to examine the contributions of environmental and genetic factors to asthma. Our primary goal was to assess familial aggregation of pulmonary function using the data from asthma and random families, including a mother, a father, a first sibling, and all subsequent siblings older than 8 yr of age. That is, each family contributed at least four subjects. We investigated the independent relation of pulmonary function among family members and assessed whether this relation was influenced by each person's characteristics, including sex, age, height, weight, education level, and smoking status. We were particularly interested in whether the pulmonary function of the mother, father, and first sibling each had a differential effect on the pulmonary function of subsequent siblings and whether there was an interactive relation of pulmonary function of the parents and the first sibling on the pulmonary function of subsequent siblings. We further examined whether the pattern of familial aggregation differed between asthma and random families.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Site

This study was conducted in collaboration with Anhui Medical University and Anqing Health Bureau in Anqing, China. Anqing stretches for about 80 km along the northern bank of the Yangtze River. The annual average temperature is 15.0° C. Anqing has three urban areas and eight rural counties, with a total area of 15,000 km². The total population in 1990 was 5.8 million (urban 9%; rural 91%), with 29% younger than 15 yr of age, 63% between 15 and 59 yr of age, and 8% older than 60 yr of age. The birth and mortality rates per thousand were 26.2 and 7.0, respectively, and average life expectancy was 65 yr. Anqing was selected for this genetic epidemiologic study for the following reasons: (1) the inhabitants are relatively homogeneous with respect to ethnic group, environment, occupation, and diet; (2) the villages have existed for several thousands years with a stable resident population; (3) the population is large enough to ensure the availability of a sufficient number of index case families. In addition, medical care in each county is administered through a three-tier (county, township, village) service network. Established a quarter of a century ago to provide medical services for all residents, this system consists of 28,000 physicians and offers a unique opportunity for efficient and uniform index case identification.

Identification of Asthma Index Families

Asthma was defined according to the following criteria: (1) The patient has a history of repeated onset of wheeze with dyspnea but is asymptotic between two events; (2) The obstructive symptoms can be significantly improved using a bronchodilator. The diagnosis of asthma was made by local physicians. COPD is defined according to history of respiratory disease, results of pulmonary function testing, and radiographs. Physician-diagnosed asthma in the present study was only used as a screening criteria. Further tests, including methacholine challenge tests, skin tests, bronchodilator tests, and IgE measurements were performed to define asthma and to distinguish from COPD. Asthma-index families were collected in the eight counties (Zongyang, Huining, Qianshan, Tongcheng, Taihu, Wangjiang, Susong, and Yuexi) through a multistage process. First, core investigators from Anhui Medical University and Anqing City Hospitals/Research Institutes held a three-day workshop in each township to train township and village doctors and to collect index case family information. The first day was used to explain the purpose, scope, and procedures of the study. The definition of an index case family was introduced and several examples were presented. Each village doctor was asked to go back to his or her own clinic to prepare a list of all families with asthma or bronchitis. The information on all index case families was collected from the village doctors in the next 2 d. Pulmonologists reviewed all index case family lists with the township/village doctors to eliminate ineligible families. After the workshop, the field team members and village doctors visited each index case family and used a short questionnaire to confirm the information provided by the village doctors. They also collected additional information on family size, pedigrees chart, health status, symptoms and medications. With this information, index case families were chosen on the basis of the following criteria: (1) the presence of an index asthma patient who was at least 8 yr of age; (2) the existence of one or more siblings, at least 8 yr of age; (3) the availability of both parents; (4) no more than one parent with a history of asthma.

Identification of Random Control Families

Random families were selected from the general population to serve as controls. In both asthma index families and random control families, asthma was defined as a "yes" response to "Have you ever had asthma?" and "Was the asthma diagnosed by a physician?" All control families in Anqing were selected from 1992 census records by a two-stage random sampling technique. The sample unit at the first stage was the administrative unit, i.e., the village. At the second stage, the unit was the household. Our inclusion criteria were: (1) a family size of at least four; (2) the availability of both parents; (3) two or more siblings in the family; and (4) an age of at least 8 yr for the youngest sibling.

Procedures

The survey was conducted between July 1, 1994 and January 26, 1998 by a team of locally hired interviewers who could use the dialect of the region and the faculty members from Anhui Medical University. A letter explaining the study was sent to each eligible family. Local officials and health centers arranged for the interviews and measurements to take place at the central office at times convenient for the participants.

Unless otherwise specified, the following procedures were carried out in accordance with the NIH Collaborative Agreement on Asthma Genetics: (1) standardized questionnaire (modified ATS-DLD) assessing respiratory history and symptoms, occupational and smoking histories, home environment, and family history of asthma, and other chronic or genetic diseases; (2) pulmonary function testing (spirometry); (3) methacholine challenge for all subjects with FEV₁ values greater than 60% of predicted value; (4) bronchodilator testing; (5) skin testing of reactivity to 10 specific antigens along with a positive and negative control; and (6) drawing of venous blood for serum IgE level, phadeatop, eosinophil, and leukocyte counts, and DNA extraction. In addition, weight and height were measured by standard methods; the subjects removed their shoes and outerwear before measurement. Height was measured to the nearest 0.1 cm on a portable stadiometer. Weight was measured to the nearest 0.1 kg with the subject standing motionless on the scale.

Pulmonary Function Tests

Standardized pulmonary function tests were conducted with ATS "Snowbird Guideline" approved equipment (Schiller, Switzerland). The maneuvers were performed with subjects seated and wearing noseclips. Each subject was encouraged to complete a minimum of five and as many as eight maneuvers to obtain three acceptable measures. The technician may have asked the subject to perform as many as eight attempts. On the basis of the Epidemiologic Standardization Criteria, a minimum duration of 6 s FVCs within the lesser of 5% or 200 ml, and technician judgment of an adequate maneuver constituted an acceptable test. The maximum of the three measurements was used for this analysis because it was believed to be more reproducible than the mean (14), and the "best test" was the simplest and most practical result to record (15).

Statistical Methods

The key outcome variable was FEV₁. Similar analysis was performed for FVC. Age, sex, and body size have been shown to be the major determinants of pulmonary function in both healthy and diseased subjects (12). Our analysis confirmed the importance of these variables in this population. Therefore, sex, age, height, weight, along with smoking status and educational level (known risk factor in adults) were included as controlling variables in all analyses.

Our analyses were conducted in several stages. First, we developed four predictive models of FEV₁ by sex and age (< 20 and  20 yr) in this population. The statistical modeling has been detailed elsewhere (16). In brief, subjects who reported any symptoms (including chronic cough, chronic phlegm, shortness of breath, and persistent wheeze), physician-diagnosed asthma, or cigarette smoking were excluded. The sex-specific models included age, height, and weight as predictors. In this analysis, we further expanded the model to include smoking, asthma status, and education level. These predictive models were used to calculate predicted value, residual, and percentage of predicted value (100 × [observed FEV₁/predicted FEV₁]) for each person. In the subsequent analysis, we used percent-predicted values of FEV₁ and FVC instead of raw values to take into account each subject's age, sex, height, weight, education level, smoking status, and asthma status.

Second, we investigated correlation of FEV₁ values among family members by computing adjusted correlation coefficients. The following pairwise correlation coefficients were obtained: (1) father-mother; (2) father-first offspring; (3) mother-first offspring; (4) father-other offspring; (5) mother-other offspring; (6) first offspring-other offspring. In the absence of environmental factors, the correlation should be around zero between the spouses; and approximately equivalent between parents-offspring pair and first offspring-other offspring pair. In the presence of environmental factors, the correlation may be more than zero between spouses and should be greater in the first offspring-other offspring pairs than in parents-offspring pairs.

We further estimated the independent relation of father's, mother's, and the first sibling's FEV₁ values to the FEV₁ of subsequent siblings by multiple regression analysis. Our analysis included multiple offspring from the same parents. This approach not only increased the statistical power of the study but also provided an opportunity to study a sequence of siblings. The standard errors were estimated by the generalized estimation equation (GEE) (17) for correction of inter-correlation in FEV₁ among siblings. We also used the GEE to test the difference in the association of father's, mother's, and the first sibling's FEV₁ values with other siblings' FEV₁ values.

Finally, we used multiple logistic regression analysis to assess the risk of low FEV₁ (defined as less than the 10th percentile of percent-predicted values) among the other siblings in relation to parental and first-sibling's FEV₁ status (defined in low, middle, and high tertiles of percent-predicted values). We presented odds ratios (ORs) and 95% confidence intervals (CIs) for each risk group. All p values were two-tailed.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In all, our study included 2,044 asthma families and 203 random families (a total of 10,584 subjects) with complete data on pulmonary function and major covariates. The mean and standard deviations of age, height, weight, FEV₁, and FVC, and the frequency distribution for sex, education level, smoking status, and physician-diagnosed asthma for father, mother, the first sibling and other siblings in asthma and random families are presented in Table 1. In addition to displaying a significant difference in the prevalence of asthma, the two types of families also differ in term of age, weight, levels of FVC, FEV₁, education level, and cigarette smoking status. The mean height was comparable for the two groups.

The adjusted correlation coefficients of FEV₁ among family members are presented in Table 2. In asthma families, mother and father were negatively correlated. The correlation coefficents of parent-first sibling and parent-other siblings were similar. The sib-sib correlation coefficient (0.22) was greater than the parent-child correlation coefficient. In addition, the correlation coefficient of mother-child (0.13-0.15) was somewhat greater than that for father-child (0.08). In random families, the correlation between mother and father was positive but small in magnitude. The father-child, mother-child, and sib-sib correlation coefficents were similar. Overall, the degree of correlation in FEV₁ among family members was greater for random families than for asthma families. We performed a stratified analysis by sex to see whether the correlation differed by the sex of the offspring and found that it did not (data not presented). Similar patterns were found for FVC (Table 2).

Using the GEE, we first examined the individual relation of the FEV₁ of the father, the mother, or the first sibling to that of the other siblings without adjustment (Table 3, Univariate). We then examined their independent relation by simultaneous inclusion of all three predictive variables in the model (Table 3, Multivariate). The regression coefficients from the simultaneous modeling of the mother, father, and the first sibling were slightly smaller than those from the separate modeling. Nevertheless, the data indicate that the FEV₁ of the mother, father, and the first sibling each had a significant and independent relation to the FEV₁ of the other siblings. Again, among asthma families, the regression coefficient of the first sibling was greater than those of the parents; and the regression coefficient of the mother was greater than that of the father. Among random families, similar regression coefficients were found for the father, the mother, and the first sibling.

The difference in association of the father, mother, and the first sibling's FEV₁ values with those of other siblings was tested using GEE (Table 4). In asthma families, the association differed significantly between father and mother, between father and the first sibling, and between mother and the first sibling. In random families, no significant difference was found.

We further evaluated the combined association of the FEV₁ values of the parents and the first-sibling with the risk of low FEV₁ (defined as less than the 10th percentile of percent-predicted value) in the other siblings in asthma families. We studied nine groups defined by parental (mean for father and mother) and first sibling's FEV₁ tertiles based on percent-predicted values. As shown in Figure 1, the rate of low FEV₁ was lowest (4.0%) when both the parents and the first sibling were in the highest tertile (high-high group), and it was highest (18.4%) when both the parents and the first sibling were in the lowest tertile (low-low group). The rates were intermediate if only the parents (7.0% for low-high group) or only the first sibling (11.5% for the high-low group) was in the lowest tertile.

View larger version (58K): [in this window] [in a new window]   Figure 1.   Percentage of low FEV1 (defined as less than 10th percentile of percent-predicted value) in other siblings in nine groups based on parental and first-sibling's FEV1 status (defined by the tertile of percentage-predicted value).

The adjusted ORs and 95% CIs for low FEV₁ among the other siblings in each of the nine groups are presented in Table 5. With the OR of the high-high group as the reference value (1.0), the OR for low-high group was 1.8 (95% CI, 1.0 to 3.5); that for the high-low group was 3.2 (95% CI, 1.7 to 5.7); and that for the low-low group was 5.5 (95% CI, 3.3 to 9.0), which is slightly greater than the sum of the individual ORs. This finding suggests that the combined effects of the parents' FEV₁ and the first sibling's FEV₁ is at least additive. Because Table 4 suggests that a father's and mother's FEV₁ may have a differential effect on their offspring's FEV₁ in asthma families, we also examined the combined effect separately for the father and mother (Tables 5 and 6). The data again suggest that the mother and father each contributed to the risk of low FEV₁ in offspring, but the mother appeared to have a greater influence than the father. A similar pattern was found for FVC, but the magnitude of the ORs ratios was greater. Because of inadequate sample size, we were unable to assess the combined association of the FEV₁ of the parents and the first sibling with the FEV₁ values of the other siblings in random families.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our study is one of the first on familial aggregation of pulmonary function in a rural Chinese population. This population offers a unique opportunity to study genetic and environmental determinants of pulmonary function. In addition to characteristics such as stable resident population, homogeneity, and large family size, the rural environment, the abundance of physical activity, and the lack of medication use contrast sharply with the typical urban and suburban settings in the United States. The major limitations of our study are its cross-sectional design and its potential information bias with regard to dietary and environmental/occupational variables, given that data were collected by interview. Nevertheless, our sample size was fairly large, standard protocol and quality-control procedures were employed systematically and consistently, and pulmonary function, weight, and height were measured objectively.

Consistent with previous investigations, this study found significant parent-child and sib-sib correlation in pulmonary function. It also provided additional evidence on familial aggregation of pulmonary function. More importantly, this study yielded several new findings. First, the pulmonary function of the mother, of the father, and of the first sibling were each significantly and independently related to the pulmonary function of the other siblings, and the association was observed even after adjustment for age, sex, height, weight, cigarette smoking status, education level, and asthma status. Second, the combined effects of the pulmonary function of the parents and the first sibling on the pulmonary function of the other siblings appeared to be at least additive. Third, there are some difference in the pattern of familial aggregation of pulmonary function between asthma and random families. In asthma families, the sib-sib association was greater than the parent-child association, and mother-child association was greater than the father-child association. However, no such differences were found in random families.

The familial aggregation of pulmonary function found in this study may result from genetics, environmental factors, or both. Cigarette smoking has been documented as an important cause of asthma and COPD (18). Although smoking is common in these asthma families, it cannot account for the observed correlations as we adjusted for smoking in our analysis.

The differential pattern of familial aggregation of pulmonary function between asthma and random families is interesting. This difference may be attributable to environmental exposures unique to asthma families, or it may be due to different genetic susceptibility to common environmental exposures or gene-environmental interactions. Asthma, defined as increased airway responsiveness with airway inflammation and reversible airflow obstruction (19), is considered to be a genetically complex disease and may be a risk factor of COPD. The link between childhood asthma and later development of COPD is not well understood nor has its environmental or genetic basis been fully elucidated. In this rural community, the major indoor air pollutants are smoke from combustion of coal or wood/straw for cooking and cigarette smoke. Mothers are usually the primary care-givers of children. As the mother and her children spend a lot of time in the kitchen, they are heavily exposed to these indoor air pollutants, which have been linked to reduced pulmonary function in previous studies conducted in the urban city of Beijing (20). Further investigation of genetic and environmental factors responsible for the observed familial aggregation of pulmonary function in asthma and random families may provide new insight into the different patterns observed.

In summary, our data suggest a strong familial aggregation of pulmonary function in this rural Chinese population. This familial influence on pulmonary function is detected from childhood onwards and can be observed in both asthmatic and random families. Healthcare providers should recognize that a person's risk of low pulmonary function is related to both parental and sibling pulmonary function status. More important, children from families in which both the parents and the siblings have low pulmonary function are at disproportionally high risk. Such information may eventually help parents and healthcare providers decide on a course of clinical assessment and management consistent with each person's risk. Researchers should be encouraged to elucidate the biomedical, social, and environmental pathways that contribute to the familial aggregation of pulmonary function. Identification of these pathways may help us to better understand the etiology of COPD and asthma, and it may lead to better strategies for early prevention, identification, and treatment.