INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The prevalence of asthma has been increasing in most industrialized countries over the past 2 decades and this does not appear to be explained by diagnostic transfer (1). The most likely cause is a change in our environment including diet (2). The role of viral respiratory infections in early life is unclear; these may be causally related to asthma or rather unmask subjects with a susceptibility to asthma (3). Infections in infancy have also been proposed to have a protective role against the occurrence of allergic diseases by promoting Th-1 as opposed to Th-2 responses by the developing immune system (4).

It has been postulated that patterns of environmental exposures among minority groups, as influenced by residence location, housing characteristics, occupation, and lifestyle, may affect overall respiratory health. One such minority group, the Inuit living in the far north above the 55th parallel, are subject to many of these factors. Results of a recent survey (5) indicate that the Inuit share with many other minority groups poor and crowded living conditions, with over 40% of households containing six or more persons, a low level of formal education, with one-third having either no schooling or less than a primary school education, high rates of unemployment, high rates of tobacco use, with two-thirds smoking cigarettes regularly, and frequent respiratory tract infections (5). However, given the extreme cold climate, exposure to allergens closely associated with allergic disease is uncommon among this population.

Asthma has also been reported to be quite unusual in this population (6) although no systematic evaluation has been carried out. We examined the respiratory health of primary school age Inuit children in Northern Quebec in the hope of shedding further light on the environmental determinants of asthma. We also wished to determine whether the very high rates of chronic respiratory disease in adults from this population (5) might have its origin in childhood respiratory ill health.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Subjects

The Inuit region considered for this study is a vast expanse of sparsely populated land stretching over an area of 563,515 km² located between 55° and 63° north latitude. The region is comprised of 14 communities scattered along its costal boundaries, with a total population of 7,078 (5). There are no roads in the region, and all communities are accessible by air only. Eligible study subjects were children in grades two to six inclusive, currently attending school in one of four Inuit communities in Northern Quebec. We have previously reported the results of one community, Salluit, undertaken as a pilot project in February 1994 (7). This community was retested two years later and these later results are included in the overall study presented here. The communities, which were tested over a 1-yr period from October 1994 to September 1995, included Kuujjuaq (population 1,405), Inukjuak (population 1,044), Povungnituk (population 1,091), and Salluit (population 848). These communities were chosen because they are the largest Inuit settlements in Northern Quebec and constitute over 60% of the Quebec population living above 55° north latitude.

Residents of the community were made aware of the study through the school and local radio broadcasts. Written consent was obtained from a parent or guardian of each child prior to their involvement in the study. Ethical approval was obtained from the Department of Epidemiology and Biostatistics Ethics Committee of McGill University.

Data Collection

The primary outcome of interest, exercise induced bronchial hyperresponsiveness (EIB), was determined using an exercise challenge test. Following the assessment of normal resting spirometric function, children underwent a 6 min free running exercise test in the school gymnasium (8). The level of exercise was assessed by both the distance covered in the 6 min interval and the heart rate achieved at maximal exercise, as measured by a pulsemeter. Post exercise spirometry was measured at 5 and 10 min following exercise. A positive response of EIB was defined as a decline of 15% or more between pre-exercise FEV₁ and that at 5 or 10 min post-exercise. A second outcome measure was that of airflow limitation, defined as an FEV₁/FVC ratio less than 0.75. All spirometry measurements were performed according to standardized methodology as described by the American Thoracic Society (9). Testing procedures for the exercise challenge test were standardized for all four communities, with the same personnel conducting the tests in all locations. Meteorologic parameters (temperature and humidity) were also obtained at the time of testing.

The presence of atopy was assessed by skin prick tests to common inhaled allergens using methods described by Bernstein (10). Solutions for testing comprised histamine, normal saline, Dermatophagoides pteronyssinus; Dermatophagoides farinae; mixed grass pollens; tree pollens; ragweed, mixed molds Aspergillus, Alternaria, Cladosporium, and Penicillium species, cockroach, feathers, cat epithelium, and dog (supplied by Omega Laboratories, Montreal). Allergens chosen were similar to those used in our study of Montreal schoolchildren (7), to enable a comparison of populations, with the addition of two potential allergens to which the Inuit are commonly exposed, namely feathers and dog.

A drop of each allergen was placed on the child's forearm and the skin was pricked through the droplet. Measurements of skin wheals were recorded after 15 min. The development of one or more wheals at least 3 mm greater than the negative control was regarded as a positive skin test. A subject was considered atopic if he or she had at least one such positive test.

Children were asked to collect saliva in the mouth and to spit into a clean test tube until they produced approximately 2 ml of saliva. Samples were frozen within 3 h of collection for later assays. The salivary sample was analyzed at a single hospital laboratory, using a double antibody radioimmunoassay (11), and adapted for the determination of cotinine from saliva according to Coutlas (12).

A questionnaire regarding details of the home environment, smoking history of the parents, and the child's history of respiratory symptoms using questions from the ISAAC (13), IUATLD (14) and the American Thoracic Society (15) respiratory symptom and asthma questionnaires was administered to a parent (usually the mother) by a trained interviewer from the community. Questionnaires were translated into Inuktitut by an experienced translator and, to ensure comprehension and accuracy, were also translated back into their original form independently by a different translator.

House dust mite samples were collected by a standard procedure (16) from both the mattress and bedroom floor of study subjects. Given the almost complete absence of house dust mite obtained during our pilot study in which over 200 mattress and floor samples were collected, we elected to undertake further collection of house dust mite for only a subsample of 15 to 20 homes in each community. This collection would enable us to confirm the absence of house dust mite in each community studied. Results are expressed as µg of D. pteronyssinus I/g of sieved dust.

The medical records located in the local community health center or hospital were reviewed for each subject and details regarding birth information, including number of weeks gestation and birth weight, as well as all out-patient and hospital admissions for respiratory related conditions, were abstracted. An index of lower respiratory illness, indicating a history of any of bronchitis, bronchiolitis, asthma, asthmatic bronchitis, pneumonia, or other lower respiratory tract infection, was created. Given the isolation of the communities and the limited mobility of the population, the medical records for the subjects located in each community were both extensive and complete. Outpatient medical care is available in all four communities, with a small local hospital located in two of the communities, Kuujjuak and Povungnituk.

Data Analysis

The outcome variables in this analysis were two measures of airway function, EIB and airflow limitation, as defined above. The determinants of the outcome (independent variables) consisted of the child's age and gender, presence of atopy, exposure to cigarette smoke including current smoking status of the child and parents as well as saliva cotinine concentrations, pregnancy related factors including low birth weight (< 2,500 g), prematurity ( 37 wk gestation), and maternal smoking during pregnancy, as well as lower respiratory infection prior to two years of age.

Saliva cotinine concentration was analyzed as ranks due to their nonnormal distribution. We categorized the resultant ranks of saliva cotinine concentrations as low (bottom tertile), medium (middle tertile) and high (top tertile).

The prevalence of EIB and FEV₁/FVC < 0.75, as well as determinants of these outcomes, were initially examined for each community separately. Given the limited number of study subjects, and in particular the low prevalence of the outcome variables of interest, data from the four communities were combined for the multivariate analyses.

We used the odds ratio as an estimator of the degree of association between the markers of airway dysfunction and the study factors. Odds ratios and 95% confidence intervals were determined using unconditional logistic regression (17).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

School lists were used to identify 594 children currently enrolled in school. Of the 590 eligible subjects (four children were excluded due to chronic health conditions), the parents of 10 (1.7%) refused participation for their children. A further 4 (0.7%) did not return a consent form and 26 of the children (4.4%) were absent from school at the time of testing. The overall participation rate was therefore 93.2%. Of the 550 subjects tested, 41 non-Inuit children were excluded from the present analyses; the final study sample therefore consisted of 509 Inuit children.

Table 1 provides descriptive characteristics for our study subjects, including history of respiratory symptoms and prevalence of airway dysfunction, by community and for all communities combined. Approximately half of the subjects were female; the average age was 9.5 yr. Six percent of subjects reported a history of wheezing in the prior 12 mo, and approximately 20% reported a usual cough. We were concerned about the lack of standardization resulting from having different translators administering the questionnaire to the children as well as the difficulty of translating respiratory symptoms, especially wheezing, into another language. We therefore chose not to analyze the determinants of respiratory symptoms further.

The overall prevalence of EIB was 5.9%, ranging from a low of 3.4% in Povungnituk to a high of 11.9% in Salluit. The prevalence of FEV₁/FVC less than 0.75 was consistent across communities, with an overall prevalence of 7.7%.

Determinants of respiratory health, by community and for all communities combined, are provided in Table 2. The percentage of children who reported they were currently smoking was high at 31.9% overall, with a low of 9.8% reported in Kuujjuak and a high of 51% in Povungnituk. The majority of parents were also smokers, with 74.9% of mothers and 68.9% of fathers reported to be current regular smokers. The high exposure to tobacco smoke is confirmed by the saliva cotinine concentrations, with an overall mean of almost 50 ng/ml.

The prevalence of pregnancy related determinants of respiratory health were relatively consistent across communities, with 5.3% of subjects of low birth weight (< 2, 500 g), 17.8% born prematurely ( 37 wk gestation), and 79.5% exposed to tobacco smoke in utero.

Lower respiratory infections prior to the age of two years were very common among this group of subjects, with almost 75% of children diagnosed at least once in a medical center with a lower respiratory tract infection. In fact, almost 25% of children had more than three visits to a medical center for a lower respiratory tract infection prior to the age of 2 yr.

Atopy was uncommon, with only 5.3% of children positive to at least one allergen, and specific sensitization to dust mite was very unusual (present in only four children). This is in line with the almost complete absence of mite allergen in house dust (none of 50 dust samples taken from the mattress or bedroom floor contained one µg or more of allergen per gram of dust), and the usual absence of furred pets in the living space of our subjects.

The prevalence of airway dysfunction, by age group and gender, and their most important determinants considered singly are provided in Table 3. No difference was found in prevalence of EIB between boys (6.9%) and girls (4.8%) (chi-square 0.97; p = 0.32), or in the prevalence of FEV₁/FVC less than 0.75 between boys (9.2%) and girls (6.1%) (chi-square 1.18; p = 0.28). Of note is the fact that the prevalence of EIB decreased with age, while the prevalence of a low FEV₁/FVC increased with age.

Meteorological conditions were not associated with the prevalence of EIB. The average temperature and humidity in the school gymnasiums were 21.2° C (SD 1.75) and 42.6% (SD 12.6), respectively.

There was no significant association between pregnancy related factors and airway dysfunction. After adjusting for age and gender, only prematurity and maternal smoking during pregnancy were associated with a slightly increased, but nonsignificant risk of EIB (OR 1.26, 95% CI 0.4-4.0; OR 2.5, 95% CI 0.6-11.3, respectively).

In Table 4 we examine the independent contribution of potential major determinants of both EIB and airflow limitation obtained from a logistic regression analysis. Atopy was not found to be associated with either EIB or FEV₁/FVC < 0.75 and its presence did not change the results of the model, indicating the lack of a confounding effect; therefore it was not included in the final model presented here.

The only statistically significant predictor of EIB was age; exercise-induced bronchospasm was approximately 30% less common for each year increase in age. There was a tendency for airways responsiveness to be less frequent among girls and in children who reported smoking at the time of the study. After adjusting for these factors there was no association seen between EIB and cotinine rank in tertiles or number of lower respiratory tract infections (LRI's) noted in the medical record in the two first years of life. If the report of smoking by the child is not included in the model, the likelihood of having EIB is significantly less in the top tertile of salivary cotinine (odds ratio 0.33; 95% CI: 0.12-0.94).

The prevalence of airflow limitation (FEV₁/FVC 0.75) did not vary significantly by age or gender. After adjusting for reported smoking by the child, a serum cotinine in the top tertile of the observed distribution was independently associated with an almost four fold increase in the prevalence of a low FEV₁/ FVC. Subjects who had experienced four or more lower respiratory infections before the age of 2 yr were also three times more likely to demonstrate airflow limitation.

If the report of smoking by the child is not included in the model, the odds ratio between a low FEV₁/FVC and being in the top tertile of salivary cotinine concentrations was 2.99 (95% CI: 1.07-8.37) and for more than three LRIs, 2.93 (95% CI: 0.97-8.82).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In contrast to our previous report of the results from a pilot study, we did not find an excess of airways hyperresponsiveness (AHR) among Inuit schoolchildren who were mostly between 6 and 13 yr of age. If anything, the prevalence of exercise induced bronchospasm (EIB) was less than observed among a group of urban school children of similar age examined with identical methods: EIB was observed in 6.9% of Inuit boys in comparison to 7.8% of Montreal boys; for girls the prevalence was 4.8% among the Inuit and 9.7% in Montreal. Furthermore, the prevalence of EIB was below 5.5% in three of the four communities examined. AHR in the fourth community, Salluit, was present in 11.9% of the children. The only observable differences between the four Inuit communities were the greater isolation and apparent poverty of Salluit. There were no obvious differences among communities in smoking habits, exposure to environmental tobacco smoke (ETS), early life experiences or prevalence of positive allergy skin tests. Of note, a previous survey of indicators of respiratory health, including a symptom questionnaire and spirometric lung function, also found the population of Salluit to rate among the poorest of the Inuit communities in terms of quality of respiratory health (18). Salluit had been the site of our pilot study and the very high prevalence of EIB (23.6% in boys and 15.9% in girls) on our first visit (7) prompted us to further investigate the respiratory health of children in these communities. In retrospect, it appears that this very high prevalence of AHR was likely due to to a high rate of viral respiratory tract infection at the initial examination. Such viral infections are a common cause of AHR (19, 20) while the crowded living conditions (21) and the presence of a single school and community center in these communities likely promote a very high attack rate of viral respiratory infections.

The only factor significantly related to EIB among Inuit children was age; AHR was progressively less common among older children as has been described in other populations (22, 23). The lack of a larger number of significant associations between EIB and either personal characteristics or environmental factors may be due in part to small sample size, especially the smaller than expected number with AHR. The four communities chosen for this study constitute over 60% of the total Inuit population in Northern Quebec. The inclusion of the other ten communities would have increased our sample size only marginally, and given the considerable expense associated with undertaking research in this isolated setting, this was not deemed to be practical. Given the participation rate of 93.2%, we have obtained the largest number of children feasible, and acknowledge that the study may have limited power to detect significant associations. There was a tendency for EIB to be less common in girls and among children who had actively taken up smoking. This latter trend suggests a certain health selection away from smoking and has been found elsewhere (24). Atopy, a common determinant of AHR in children (25), was quite uncommon in this population and this likely contributes to the relatively low prevalence of AHR found. Common causes of allergic AHR such as dust mites and cats (26), were not present in the environment of these children. The relative rarity of asthma and atopy among the Inuit may conceivably be related to a greater consumption of oily fish (27). Exposure to cigarette smoke was so widespread and intense that its effect on AHR may have been masked by the absence of a low or nonexposed group of sufficient size. Finally, the high rate of respiratory infections observed during the first few years of life in this population (5) may favor the development of Th1 as opposed to Th2 immune responses, the latter associated with an increased risk of allergic diseases including asthma (4).

We used exercise challenge as our measure of AHR because of its safety when carried out even without medical supervision and its acceptability to the communities involved. We also had used this method with success and its continued use allowed for comparison with results obtained among other groups (7, 28). The validity of this method for assessing AHR among children in the general population is widely accepted (13, 22).

Of concern is our finding that nearly 8% of the Inuit children examined had evidence of significant airflow limitation. Using similar methods, such evidence of airflow limitation was present in only four of 989 Montreal schoolchildren (28). Airflow limitation as evidenced by an FEV₁/FVC < 0.75, was more common among those with the highest levels of salivary cotinine and in children whose medical records provided confirmation of four or more episodes of lower respiratory infection in the first two years of life.

The pervasiveness of exposure to cigarette smoke is likely responsible in large part for the common occurrence of airflow limitation. In utero exposure was particularly common with almost eighty percent of mothers reporting having smoked during pregnancy. Such exposure has recently been confirmed as a cause of airflow limitation discernible very early in life (29). Smoking by the children themselves as well as by adults living in the same homes is also substantially more common than that reported, even among adolescents or adults respectively, in more southern parts of the province of Québec (30). The frequency of smoking and the number of cigarettes consumed may have increased among the Inuit as suggested by two surveys carried out ten years apart in Igloolik, North West Territories (31).

The impairment of lung function observed in this community appears to have been present in a previous study of this population carried out in the mid-60s (32) when assessed as the proportion of subjects whose spirometric measures of airflow were outside the expected range derived from the Inuit population without respiratory symptoms. Values of FEV₁ and FVC as well as mid-expiratory flows (MMF%) were above the average for caucasian children of the same age, however. We therefore selected the FEV₁/FVC ratio as our measure of airway function since it does not require adjustment for size using information derived from a different population.

The very high frequency of respiratory infections at a young age has long been recognized among the Inuit (33). Coultas and colleagues (34), in their thorough review of respiratory diseases in minorities of the United States also report a high incidence of severe lower respiratory tract infections in both American Indian children of the United States and Canada. Among our study population, the number of lower respiratory infections documented in the child's medical record during the first two years of life related adversely to lung function as assessed by the FEV₁/FVC measured after an interval of approximately 5 to 10 yr. The long-term consequences of these early life events varies according to reports published thus far but are associated with deficits in lung function in many studies (35). The occurrence of frank bronchiectasis may also be a contributing to the excess prevalence of airflow obstruction since this problem has been encountered frequently among Inuit children in the past (36).

In summary, airways hyperresponsiveness was relatively uncommon among school age Inuit children despite the high levels of exposure to tobacco smoke and the common occurrence of lower respiratory illnesses in the first two years of life. This is in keeping with the rarity of diagnosed asthma in this population (6) and may be explained, at least in part, by the absence of exposure to common indoor inhaled aeroallergens and the rarity of allergic sensitization to these allergens. Airflow obstruction, as evidenced by a reduced FEV₁/FVC, was not uncommon and was found more frequently among children with the highest levels of salivary cotinine, reflecting both active and passive exposure to cigarette smoke, and among children having experienced the greater number of lower respiratory illnesses in early life. From the public health point of view, measures to reduce crowding, and thus the spread of respiratory infections, and, more importantly, the prevention of smoking are likely to have the greatest beneficial impact on the respiratory health of these isolated communities.