Airways hyperresponsiveness and atopy : A comparison of Inuit and Montreal schoolchildren

OBJE(TIVE: To compare the prevalence ol exercise induced bronchial hyp<'rrespons iveness in Inuit child ren wit h that or chi ldren in Montreal. and to identify poss ible genl'lic and envi ronmental determ inants or the cl il lcn.:nces observed. DEs11; : ross-sectional survey. SETl'IN<;: Salluit. an isulat<.:d lnuil L'Olllmun ity in northern Quebec, and Montreal. POPULATION STllDIEU: All children attcndin~ school in Salluil in grades 2 lo 6 were eligible to participate. f-01:-the Montreal study. 18 schools were selec te I and frum each of thesL\ one class from each of grades I. 3 and "i we re chosen. MEASllREMElliTS: Data rnlkction lor both locati ons included an exe rcise challenge tes t lo assess exncisc induced bronchia l hyperrcsponsivL·ness (ElBHl. allergy sk in testi ng, a questi onnaire for parenh regard ing de tails or the home L'n virnn mcnl as well as the ch ild ·s hi story of' res piratory sv111ptu111s. ,111d collec tion of dust samples Imm the bedroom rtc)l)r and matl rcss for the presrncc· ul house dust mite . RESULTS : The prevalence or EIB H (ddinccl as a dc'cline or I 5'k or more bL·tw,T n pre-exercise forced exp iratory volume in I s I FE Y I I and that al 5 or IO rnins postexercisc ) was 19.5% (23 uf 118) among the Inuit children. rnmparcd wi th 8.8% (87 of 989) among the Montreal children. In contrast. nnly 8.60'r o r the In uit L'l1ild rcn h:nl :1 po,itivc :illcr'.,'.y sl,_i11 tc',t compared with 3-1-% i11 Mont real. CONC LUSIONS: A higher prevalence ul' ElB H was fou nd in lnuil schoo lchildren compared with chi! lrrn or similar age in Montrc'ai. altlll.>ugh the prcva le1ll'c' or atopy was cons iderably lower.

The purpose or this study was ro compare the prevalence of exe rcise inducL•cl bronchi,il hyperrespons ivene ss in In uit childre n with that ol' d1ildren in Montreal.Simi lar met hodolngy used in the two locations allowed valid cornparisons or the two populati1111 s to be m;:ide.and also enabl ed an exploration of' differences in the distribution of l' nvironmenral risk fac tors.This paper focuses on the res ults obtai ned fro m the first of several Inuit communities currently being studied.

PATIENTS AND METHODS
Study subjects: For the Inuit study.cligibll' subjects were all children in grades 2 to C). inclusive.attending school in Salluit.Quebec.an isolated Inuit community (population 848) located at 62° nort:1 latitude.Res idents or the com munity were rnade awarL• or the study through tilt' school and local radi o broadcas ts .and written consent was obtai ned l'rom a parent or guardian of' each child he f'ore enrol ment in the study.All written inlon11:1tio11 to parents and guardia n. , was translated into lnu ktitut by an experienced tra nslator and. to ensure comprl'lll'ns11lll and accurac y. these documents were also tran slated back into their origina l form by another independent translator.
School lists in Salluit were used to idrntify 122 children attending school.01 121 el igib le subjects (one child was cxcluclecl from the study populati on due to a chroni c health condition) 120 part ici pated in the study .for an overa ll parti cipation rate of 1 J 1 )%.Two of the 120 children tested did not success fully L '<Hnplete the exe rcise c hallenge test: the rdmc.the linal study sample ,1v;1ilablc !'or analysi s was 118.
For the Montreal study.18 school., were selected rn1 the island of Montreal to represent a broad range• or soci oeconomic status.From eal'11 sc lmol one class of each o l grades I. .1 and 5 were se lec ted .for a tot al of 1274 children .Each child \\'aS prov ided with a letter.a short lJUL'st11m naire ,111d a crn1 -,ent 1' 01111 to be completed at home by a parent.
or 1274 ch ildrL'n se lected in Montrea l, 7:'i (.'i.\)'/i) did not r<'lurn the q ues tionnai re and consent form.while a fur ther 130 ( 10.2% ) of t he p,1rents refused participation lor their children.Forty-two of these paren ts did.however.complete the q uestionna ire: therefore.questionnaire data were available for 1111 of the children.0 1' 1069 chil dre n with consent to participate .23 were sick o r absent.28 of the exe rci se tes ts were not an:eptable.In the Montreal study there were 110 differe nces between part icipants and non participants as to age of the child.sex or race.Among fa milies who refused permission for th,:ir ch ild to participate in testing at school.hut who did retu rn the questio nnaire.mothers were more likel y to be smokers (17.9 % versus I X. 81'/0 ) and fathers less li ke ly to have a history or asth ma (J.9'k vers u, 17.2 % ).
In the Montrea l study ,1 subsample of children was chose n for more de tai led investi gat ions in their homes.The subsample Cllntained child ren wi th a physician • s diagnosi., of ast hma at any time in the past.as detem1i necl from the parent questionnaire.and those whose forced e xpiratory volu me in I s ( FEY 1) fe l I by I 0'7t• or more art er the exerc ise cha! lengc test in the school.For each of these •cases• the next child on the alphabetical class list of the same se x was chosen as a •cuntrnl' .
L::th ical approval for both studies was obtained from the Department of Epidemi ology and Biostatistics Ethic, Committee o r McGi ll Univnsit y.Data collection: In groups from both Salluit and Mo11treal the outc ome of in te rest .e xerci se indu ce d bn111chi;1I hypnresponsiven ss (EIBI I).,vas determi ned usi ng an exerci se cha llenge tes t.Fo llowing the assessme nt uf normal resting spirometric functi on.ch ildren undc rn•L'nt a 6 min free runni ng e xerc ise test in the school gy mnasium ( I (1).The level ol exercise was assessed by both the d ista nce covered in the 6 min interval and the hea rt ra te achieved at ma xirnal exercise as measured by a pulsemctcr.Postexerci se ,pirnmet ry was meas ured at 5 and 10 mins following exe rcise.A positive response of EIBH was defined as a dec line ol I ."i'kor num• between pre-e xerci se FEY 1 and that at 5 or 10 rnins pos t• exe rci se.All spi rometry measurements we re perro1111ed according to standard ized methodology as described by the American Thoracic Society ( 17 ).Test ing procedures ro r the exerci se chall enge t st we re standard ized for the two com111u-11i tics.with the same personne l cond ucting the tests in both !neat ions.No forma l assessment or the reliabilit y of' the FE\/ 1 meas urt'ments was undertaken.
The prese nce or atopy was asse ssed by sk in prick tes ts tu common inhaled all ergens using 111L'thnds descrihl'll by Bern stein ( 18).This w :1s carried out in all the S,illuit ch ild ren .but on ly among the subsamp le of Montrea l children chm;en fo r furt her study at home.Soluti ons for testi ng comprised hi stam ine.normal sa Ii nc.Den11uto11/iL1goidcs />lcro11_ 1•ssi11Hs .Uermatoplwgoidcs ji1ri11ae.mi xed g rass pollen,.tree pollens .ragweed.mixed mold s.Aspcrgi//11.1.,\//crnut'ia.Clodo-''fh>ri,1111 and PcHicilli11111 species.L•at ep ithdium and cockroach (su pplied by Omega LaboratoriL:s ).A drop of' each allergen was placed 011 the child ' s f'oremm :111d the ski n was pricked through the droplet.Meas urements of' the sk in wh al we re recorded after 15 111i11s.The devel opment o r one ()f' 1norL' whea ls at least 3 111 111 or greate r than that res ulti ng from the negat ive control was rega rded as a posi tive ski n test indicating atopy.A subject was considered a topic if he or she had at least one positive test.In Salluit.trainL'J intnvicwers from the community administered a quL'stionnairc in lnuktitut lo a parent or guardian (usually the mothe r) of each of the study subjects in their homes .The questionnaire had been translated into lnuktitut :ind back-translated. in thL' same manner as the other documents discussed above.In Montreal, the questionnaires.in English or French.were distributed and collected through the schools and self-admini stered by the parents.Questions included details about the home environment, including smoking history of the parents, and the child's hi •tory of resp iratory symptoms using questions from the IS AAC ( I lJ).IU ATL D (20) and American Thoracic Society (21) respiratory symptom and asthn,a quL'stionnaires.llousc dust mite samplcs were col IL-ctcd by a standard procedure (22) from the mattress and bedroom floor for each or the Salluit study subjects and among the subsample of Montrea l children chosen for further study at home.Results arc expressed as Fg of!.> ptcm11v.,si1111.1 1/g of sicvcJ Just.;\ kVL'l or 10 ,Ltg/g of dust has been associated with acute attacks and increased risk or future asthma ( 14 ).Data a nalysis: The intent of this paper is to provide descriptive data.consisting principally of percentage prevalence, rather than to test specific hypotheses.For this reason statistical testing has been limited to that comparing the difference betwee n the propo11ion of EIBH observed in the two populations 11sing the x 2 test.Eventu:1lly a mult ivariate analysis will be used to identify determinants of EIBH for each or the two populations separately, as well ;is together.while contrnlling fo r potential confounding variabiL's.
Because allergen skin tests and house dust mites were only measured in the chosen subsample of children in the Montreal study.the res ults from this subsample were extrapolated to the full sample of Montreal children to estimate the prevalence of these two vari ab les anJ to allow for a comparison with the prevalence obtained among the Salluit child ren .The rates of allergen skin tl'st positivity and the prevalence of house dust mite concentrations were extrapo-1.ill'daccorJing to the sampling weights from their observed frequency in the comparison group among the subsample visited at home.The observed frequency among children with EIBH or a history of asthma were thL'n directly com-94 bincd with the extrapolated frequencies to provide frequencies in the complete group studied in the schools.I.The mean age of the Inuit children was IO years, slightly older than that of the Montreal children.Inuit children were much more likely to be exposed to tobacco smoke, both in utero and currentl y, than the Montreal children.Over two-thirJs of the Inuit children were exposed to tobacco smoke in utero.compared with 11:: .ssthan one-third of the Montreal children.The Inuit children were also more likely to have a mother who was a cu1Tcn1 smoker (69.6%) than were the Montreal children (37.2%), while the percentage of children whose fathers were current smokers was approximately equal for the two populations.

Characteristics of stuJy subjects and exposure to tobacco smoke arc presented in Table
Among the 118 Inuit children who success fully completed the exercise challenge test.the overall prevalence of EIBH was 19.5 %.The prevalence was highest among children seven to nine years old (25 %) and decreased to I I 'It in children 13 to 15 years old.As indicated in Table 2, EIBH in the Inuit was more common in boys ( 13 of 55 , 23.6%) than in girls (IO of 63, 15.9%).In Montreal the overall prevalence was 8.8% and was lower in boys (39 of 497.7.8 cfo ) than in girls (48 of 492, 9.7%), though the sex difference in Montreal children was proportionately less than among Inuit children.In a comparison of the two populations, the prevalence or EIBH was significantly higher among the Inuit boys than among the Montreal boys (P<0.00I), but there was no signifi cant difference in the prevalence of EIBH between Inuit and Montreal girls (P>O.I 0).
Rates of atopy , or allergy skin test positivity.were much lower among the Inuit children.Overall, 8.6'/t of the Inuit children had a positive skin test compared with 34%, of the Montreal children (P<0.00I) .The pre valence of atopy was higher among boys in both the Inuit and Montreal children (Table 2) .
The prcv:..tlcncc of the child 's respiratory symptoms, as reported by parents, was consistently higher among the Inuit population (Table 3).
House dust mite was virtually nonexistent among thL' Inuit.Only one of 93 bedroom floor samples and none or the 93 matt ress samples contained I g or mon: of total dust mite/g of dust.The corresponding levels in Montreal, when extrapolated to the full sample, were 15.6% for bedroom tloor samples and 2J.6% for mattress samples.

DISCUSSION
In a cross-sectional study of 118 Inuit schoolchildren we observed a high prevaknce of airways hyperresponsiveness as shown by EIB H.With respect to the influence of age, resu lts of our data are similar to those reported by Bardagi (23), in that the prevalence of EIBH decreased progressively with age.This excess in a mt>asure of bronchial hyperrespon-~ivcness appears to be contrary to the commonl y held belief among physicians working with the population that asthma is an unusual occurrence (5).This contradiction may be more apparent than real.however.since the airway hyperresponsiveness may be of a different etiology than that commonly found among children with allergic asthma, and may manifest ifsclf differently clinically from the asthma Wt' usually recognize among children in more southern climates.
We chose to assess bronchial hyperresponsi veness using an exercise challenge test rather than another brunchoconstrictive stimulus such as hi stamine or methacholinc for two primary reasons.First, the validity of exercise as a means or assessing airway narrowing has been demonstrated (23), and second, exercise was felt to be a more practical and safe stimulus to uSL' in these remote isolated communities than phannacological induction of airway narrowing.We are not sure whether the excess bronchial hyperresponsiveness observed would also occur with mcthacholine or histamine.
While it has bee n demonstrated that responses to these different bronchoconstrictive stimuli correlate within individuals, this cotTelation is inconsistent and relatively weak (24 ), though exercise induced brnnchospasm may be more specific to asthma (25).
Skin tes t reactivit y to common inhaled allergens was present in less than I 0'/c, of the Inuit children.compared with a third or more of children in Montreal , suggesting that allergic sensitization was uncommon.Furthermore, the most important allergenic cause of asthma .the dust mite.could not be detected in most bedrooms of Inuit children and was never present at leve ls thought to be associated with sensitization or acute symptoms ( l 4 ).
On the other hand.exposure to tobacco smoke was extremely common in the Inuit children due to both a very high prevalence of smoking among parents and the amount of time spent indoors in these northern regions.The' relationship between exposure to second-hand smoke and an increase in bronchial responsiveness among children has previously been described (26).Cold air, proposed by some as a cause of airway disease. is another potential irritant that may partially account for the high prevalence of EIBH among the Inuit children (27 ,28).Howe ver, a more likely explanation may be the very high rate of both upper and lowt>r respiratory tract infections occurring among Inuit children during infancy and early childhood.Evidence from the literature suggests that respiratory infections may have long tenn consequences on airway !'unction (29).Meteorological conditions in the gymnasiums where the exercisL~ tests were conducted were similar in the two locations, and therefore were not likely to haw influenced the difference in EIBH observed.Averagg temperature and humidity were 22.0°C and 22.8 9;,, respectively.in Salluit and 19.3°C and 29.8 %, respectively.in Montreal.
It would be difficult to attribute the high prevalence of brond1ial hyperresponsiveness to social disadvantages alone.since the relationship between bronchial hypcrresponsiveness and socioeconomic status has been wL•ak and inconsistent in previou.~studies (30,31 ).It is also unlik ely that hL'ating methods accounted for the differences obsL' rved.All Inuit homes were heated with an oil furnace.while in Montreal the majority of homes used electric radiators.A genetic component may be present since respiratory diseases remain an important cause of mortality and morbidity among adult Inuit, which is evident despite the reduction in prevalence of tuberculosis (32).
Although these results arc preliminary and must be interpreted as such.they du suggest that the prevalence of airways hypcrresponsiveness, as shown by ElBH, is considerably higher among Inuit school children than among children of similar age in Montreal.However, the presence or atopy is remarkably lower (P<0.00I).We are currently examining several hundred more Inuit children and will be ex ploring the potential role of the indoor environment, as well as medical history of respi ratory illness, in bringing about this high rate of exerci se induced bronchial hyperrcsponsiveness.

Can Respir J Vol 2
No 2 Summer 1995 Airways hyperresponsiveness and atopy in Inuit children

TABLE 1
Characteristics of study subjects

TABLE 2
Prevalence of EIBH and atopy among children in Salluit and Montreal