American Journal of Respiratory and Critical Care Medicine

Rationale: The temporal stability of adult asthma phenotypes identified using clustering methods has never been addressed. Longitudinal cluster–based methods may provide novel insights in the study of the natural history of asthma.

Objectives: To compare the stability of cluster-based asthma phenotype structures a decade apart in adults and to address the individuals’ phenotypic transition across these asthma phenotypes.

Methods: The latent transition analysis was applied on longitudinal data (twice, 10 yr apart) from 3,320 adults with asthma who took part in the European Community Respiratory Health Survey, the Swiss Cohort Study on Air Pollution and Lung and Heart Diseases in Adults, or the Epidemiological Study on Genetics and Environment of Asthma. Nine variables covering personal and phenotypic characteristics measured twice, 10 years apart, were simultaneously considered.

Measurements and Main Results: Latent transition analysis identifies seven asthma phenotypes (prevalence range, 8.4–20.8%), mainly characterized by the level of asthma symptoms (low, moderate, high), the allergic status, and pulmonary function. Phenotypes observed 10 years apart showed strong similarities. The probability of membership in the same asthma phenotype at both times varied across phenotypes from 54 to 88%. Different transition patterns were observed across phenotypes. Transitions toward increased asthma symptoms were more frequently observed among nonallergic phenotypes as compared with allergic phenotypes. Results showed a strong stability of the allergic status over time.

Conclusions: Adult asthma phenotypes identified by a clustering approach, 10 years apart, were highly consistent. This study is the first to model the probabilities of transitioning over time between comprehensive asthma phenotypes.

Scientific Knowledge on the Subject

Clustering approaches have been used in the respiratory epidemiology field to disentangle the heterogeneity of asthma, but the temporal stability of such asthma phenotypes and the long-term phenotypic transition at the individual level have never been addressed.

What This Study Adds to the Field

Modeling shows strong similarities in cluster-based adult asthma phenotypes studied 10 years apart. Transition toward increased asthma symptoms is more common in nonallergic phenotypes than in allergic ones.

Asthma is a complex disease characterized by a strong clinical heterogeneity and possible phenotypic variability over time (13). Accurate asthma phenotypes are needed for better asthma management and for better identification of phenotype-specific risk factors. Disentangling asthma phenotypes, or more generally IgE-related phenotypes, is a current challenge (4, 5). Beside the “candidate” approach that identifies a priori phenotypes on the basis of one or few disease characteristics, unsupervised or data-driven approaches have been proposed to unravel the heterogeneity of asthma by means of a clustering approach integrating multiple disease features (6, 7).

To date, clustering analyses have been performed in a cross-sectional manner, by integrating several domains of the disease measured at one point in time (811), or longitudinally, by using a single disease characteristic assessed at several time points to define trajectories (12, 13). Latent class analysis (LCA), a model-based clustering method, has previously been conducted in two large epidemiological studies, and revealed four distinct asthma phenotypes in each sample (11). These phenotypes clearly discriminated populations in terms of quality of life and blood eosinophil and neutrophil counts. The validity of cluster-derived asthma phenotypes is supported by similarities in phenotypes across different populations and by discriminative properties and association with clinical prognosis or risk factors (8, 9, 11, 1316). Although a previous study provided first results indicating a stability of cluster-based asthma phenotypes over 1 year (15), none applied such cluster analysis several years apart in the same population and compared the structure of the phenotypes obtained. Furthermore, the transitions between cluster-based asthma phenotypes at the individual level have never been addressed in large epidemiological studies (5). In addition, no studies have attempted to identify adult asthma phenotypes using a more comprehensive approach, simultaneously integrating several domains of the disease, repeatedly measured over time. Because asthma is a variable disease, such an approach may lead to the identification of more specific phenotypes, which could facilitate the identification of risk factors. Cohorts with long-term follow-up, such as the European Community Respiratory Health Survey (ECRHS) (17), the Epidemiological Study on Genetics and Environment of Asthma (EGEA) (18), and the Swiss Cohort Study on Air Pollution and Lung and Heart Disease in Adults (SAPALDIA) (19), can provide this additional information.

Using a large sample of subjects with asthma, recruited in the frame of three 10-year follow-up epidemiological studies (ECRHS, EGEA, and SAPALDIA), the present study aimed to compare the stability of cluster-based adult asthma phenotype structures a decade apart and to address the individuals’ phenotypic transition across these asthma phenotypes.

Additional details on methods are provided in the online supplement.

Participants

We analyzed longitudinal data from three large epidemiological cohorts recruited in the early 1990s and followed 10–12 years later: ECRHS, an international population-based study of asthma (ECRHSI, 1991–1993; ECRHSII, 1999–2002) (17), SAPALDIA, a Swiss population-based study (SAPALDIA1, 1991; SAPALDIA2, 2001–2003) (19), and EGEA, a French case–control and family study of adults and children at baseline (Epidemiological Study on Genetics and Environment of Asthma [EGEA1] 1, 1991–1995; EGEA2, 2003–2007) (20).

Phenotypic Characteristics

In the three studies, all subjects were extensively characterized using standardized protocols and questionnaires. Examination procedures included a detailed respiratory questionnaire, with questions on asthma, respiratory symptoms frequency, and treatments for asthma problems and allergic rhinitis. Lung function and bronchial hyperresponsiveness (BHR) test results were measured according to standardized procedures. The questionnaires are available on the Web sites of the studies (http://www.ecrhs.org/, http://www.sapaldia.net, and http://egeanet.vjf.inserm.fr). Previous analysis combining these three epidemiological studies has already been conducted for other purposes (21).

“Ever asthma” was defined in ECRHS and SAPALDIA by a positive answer to: “Have you ever had asthma?” and in EGEA either by a positive answer to one of the two standardized questions: “Have you ever had attacks of breathlessness at rest with wheezing?” or “Have you ever had asthma attacks?” or being recruited as an asthma case. Current asthma was defined by subjects with ever asthma who reported asthma attacks or asthma treatment or at least one asthma symptom (Table 1) over the previous 12 months.

TABLE 1. DEFINITION OF THE VARIABLES SELECTED FOR THE ANALYSIS

CriteriaDefinition Used in ECRHSDefinition Used in SAPALDIADefinition Used in EGEA
Asthma symptom score, 12 moSum of five symptoms:IdemIdem
 1. “Have you had wheezing or whistling in your chest at any time in the last 12 months?” and “Have you been at all breathless when the wheezing noise was present?”
 2. “Have you woken up with a feeling of tightness in your chest at any time in the last 12 months?”
 3. “Have you had an attack of shortness of breath that came on during the day when you were at rest at any time in the last 12 months?”
 4. “Have you had an attack of shortness of breath that came on following strenuous activity at any time in the last 12 months?”
 5. “Have you been woken by an attack of shortness of breath at any time in the last 12 months?”
Woken up by attack of coughing, 12 mo“Have you been woken by an attack of coughing at any time in the last 12 months?”IdemIdem
Chronic cough or phlegm“Do you usually cough during the day, or at night, in the winter, on most days for as much as three months each year” or “Do you usually bring up any phlegm from your chest during the day, or at night, in the winter, on most days for as much as three months each year?”IdemIdem
Asthma attack, 12 mo“Have you had an attacks of asthma in the past 12 months”IdemIdem
Asthma treatment*“Are you currently taking any medicines including inhalers, aerosols or tablets for asthma?”IdemDid you take any medicines for your asthma attacks in the past 12 mo?
Allergic sensitizationSpecific IgE to one allergen among cat, Dermatophagoides pteronyssinus, Cladosporium, and timothy grass at ECRHSI and ECRHSIIPositive Phadiatop test at SAPALDIA1 and SAPALDIA2Skin-prick test to any of the 11 allergens (cat, Dermatophagoides pteronyssinus, Blattela germanica, olive, birch, Parieteria judaica, timothy grass, ragweed pollen, Aspergillus, Cladosporium herbarum, Alternaria tenuis) at EGEA1 and EGEA2
Total IgE<100 IU/ml vs. ≥100 IU/mlIdemIdem
FEV1Assessed using the best of five expiratory curvesAssessed using the best curve from spirometry with up to eight maneuvers (conforming to ATS/ERS quality criteria). Using Quanjer equations from 1993 for FEV1 % predictedAssessed using the best of five expiratory curves
 Two classes: <80% predicted vs. ≥80%Two classes: <80% predicted vs. ≥80%Two classes: <80% predicted vs. ≥80%
Bronchial hyperresponsivenessFor subjects with FEV1 > 70%, PD20 ≤ 1 mg methacholineIdemFor subjects with FEV1 > 80%, PD20 ≤ 1 mg methacholine

Definition of abbreviations: ATS = American Thoracic Society; ECRHS = European Community Respiratory Health Survey; EGEA = Epidemiological Study on Genetics and Environment of Asthma; ERS = European Respiratory Society; PD20 = provocative dose causing 20% decrease in FEV1; SAPALDIA = Swiss Cohort Study on Air Pollution and Lung and Heart Disease in Adults.

* The question “Are you currently taking any medicines, including inhalers, aerosols, or tablets for asthma?” was not available at EGEA1, so, for longitudinal consistency, we retained the following question “Did you take any medicines for your asthma attacks in the past 12 months?” in EGEA. Strong agreement was observed between answers to these two questions at EGEA2 (overall agreement was 81.5%; κ = 0.63).

Biases

A standardized protocol and questionnaire were used in the three studies, allowing minimization of the risk for measurement biases. Analyses of follow-up bias have previously been addressed within each study, and showed that the follow-up participants did not strongly differ from the nonparticipants in regard to the main sociodemographic and clinical characteristics (18, 19, 22). A sensitivity analysis was conducted, which was restricted to the population-based cohorts (ECRHS and SAPALDIA) to address the role of potential selection bias introduced by the study design.

Strategy of Analysis

The main analysis was conducted on the pooled dataset. The latent transition analysis (LTA) is a cluster-based model developed for longitudinal data (23). Such models allow the integration of repeated data and the identification of discrete latent classes of individuals based on their shared characteristics across a set of observed categorical variables, and allow model change over time between classes. Parameters estimated in LTA include class membership probabilities, which represent the phenotype prevalence and the item response probabilities, which are used to characterize the phenotype structures. LTA also produces a transition probability matrix, estimating probabilities of membership in the same phenotype at each time point (entries along the diagonal) and probabilities of transitioning to a different phenotype over time (entries off the diagonal) (24, 25).

This longitudinal analysis was conducted based on the nine variables covering personal and phenotypic characteristics (Table 1). Variables included in our previous LCA analysis (11) that were constant/almost constant over time were not retained in the present analysis, because LTA is aimed at characterizing transition over time and requires variability in all variables. Interestingly, we replicated the four LCA-derived asthma phenotypes previously identified at follow-up (11) when applying the LCA model on this restricted number of variables (see Table E1 in the online supplement). Age, age at asthma onset, and sex were considered as covariates in the LTA model. The number of classes was statistically determined using the Bayesian information criterion (BIC). When different models showed BICs of similar magnitude, phenotype prevalences were also considered to avoid a solution leading to low prevalent phenotypes (<5%). To assess the stability of the structure of the asthma phenotypes 10 years apart, we tested the goodness of fit of the “constrained” LTA model in which identical item–response probabilities across time are estimated (meaning that the same cluster structure is forced in the two time points) as compared with the “unconstrained” LTA model in which item–response probabilities are estimated at each time point. Furthermore, to address sex differences in asthma phenotypes, models were also fitted by sex, and results observed in men and women were statistically compared.

We investigated how the LTA-derived asthma phenotypes defined at baseline were related to the subsequent risk for asthma exacerbation assessed at follow-up by the report of hospitalization for asthma or the use of oral steroids for breathing problems in the previous 12 months in ECRHS and EGEA (data not available in SAPALDIA).

Asthma Participants

The current study was based on 3,320 adults (2,031, 863, and 426 participants in ECRHS, SAPALDIA, and EGEA, respectively) reporting ever asthma at baseline or follow-up (mean [SD] follow-up time was 8.5 [±1.4], 11.0 [±0.3], and 11.7 [±1.1] yr for ECRHS, SAPALDIA, and EGEA respectively; Figure 1). A total of 2,072 subjects reported ever asthma at both points in time, 921 reported ever asthma at follow-up only, and 327 reported ever asthma at baseline only. The population includes 1,918 (58.5%) and 2,350 (72.5%) subjects with current asthma at baseline and follow-up, respectively.

Descriptive Data

Mean (±SD) age at baseline was 35.8 (±9.8) years, and 44.0% were men (Table 2). The prevalence of allergic characteristics (allergic sensitization, total IgE > 100 IU/ml) was stable over time. Low FEV1 (FEV1 < 80% predicted value) was observed in 12.0 and 14.0% of the subjects at baseline and follow-up, respectively. The prevalence of BHR was 44.8 and 40.6% at baseline and follow-up, respectively. About one-third of this population with asthma reported asthma attacks in the previous 12 months at both time points, and 54.7% reported adult-onset asthma. Because of the different study designs used, subjects with asthma recruited in EGEA had more asthma attacks, asthma symptoms, and asthma treatment at baseline, but these differences were lower at follow-up.

TABLE 2. DESCRIPTION OF THE ASTHMA POPULATION IN ECRHS-SAPALDIA-EGEA

 Baseline10-Year Follow-up
ECRHSI (n = 2,031)SAPALDIA1 (n = 863)EGEA1 (n = 426)ESE1 (n = 3,320)ECRHSII (n = 2,031)SAPALDIA2 (n = 863)EGEA2 (n = 426)ESE2 (n = 3,320)
Age, %*24.851.843.434.218.853.847.431.6
Sex, men, %41.547.648.844.0
Age at asthma onset, >16 yr, %54.257.851.154.7
Allergic sensitization, %63.059.270.863.165.459.171.164.4
Total IgE ≥ 100 IU/ml, %45.644.062.447.544.939.653.144.5
Woken by cough 12m, %45.237.242.642.844.237.639.641.9
Chronic cough/phlegm, %23.814.717.020.622.616.716.720.3
Asthma attack 12m, %42.323.053.738.740.424.132.535.1
Asthma symptoms score 12m: 0, %26.842.022.330.226.544.122.130.5
1 or 2, %35.437.226.234.739.837.245.139.8
≥3, %37.820.851.535.133.618.632.829.6
FEV1 < 80% predicted, %10.910.818.712.012.614.119.714.0
Methacholine test, n1,4235742372,2341,0974072041,708
BHR (PD20 ≤ 1 mg), %49.631.248.544.845.924.145.140.6
Asthma treatment, %35.016.173.733.743.525.953.539.7

Definition of abbreviations: BHR = bronchial hyperresponsiveness; ECRHS = European Community Respiratory Health Survey; EGEA = Epidemiological study on Genetics and Environment of Asthma; ESE = ECRHS-SAPALDIA-EGEA; PD20 = provocative dose causing 20% decrease in FEV1; SAPALDIA = Swiss Cohort Study on Air Pollution and Lung and Heart Disease in Adults.

* Older than 40 yr at baseline and >50 yr at follow-up.

Skin-prick tests (11 allergens) in EGEA1 and EGEA2; specific IgE (four allergens) at ECRHSI and ECRHS II, positive Phadiatop test at SAPALDIA1 and SAPALDIA2.

Phenotypes Using LTA

LTA analyses with seven, eight, and nine classes led to BIC of similar magnitude (Figure E1). The model with eight classes identified a low-prevalence phenotype (<5%); therefore, the seven-class model was retained. Strong similarities between asthma phenotypes identified at baseline and follow-up are evidenced by the “unconstrained” seven-class LTA analysis, in which specific asthma phenotype structures were freely estimated at each time point (Figure 2). Therefore, the “constrained” model in which asthma phenotypes are forced to be identical across time was retained as the main model to facilitate the interpretation of results. Phenotypes identified in the main model were largely characterized by the level of respiratory symptoms, the allergic status, and the pulmonary function (Figure 3 and Table 3):

Phenotype A (prevalence: 21 and 19% at baseline and follow-up, respectively), labeled “allergic, few symptoms, no treatment,” was predominantly composed of subjects with asthma with no or few respiratory symptoms (i.e., 1% had asthma attacks in the previous 12 mo) and with allergic sensitization.

Phenotype B (prevalence: 17 and 16% at baseline and follow-up, respectively), labeled “nonallergic, few symptoms, no treatment” was also composed of subjects with asthma with no or few respiratory symptoms, but compared with phenotype A, subjects belonging to this group did not show allergic sensitization (4% were sensitized to aeroallergens).

Phenotype C (prevalence: 8 and 12% at baseline and follow-up, respectively), labeled “nonallergic, high symptoms, treatment” was mainly composed of nonallergic subjects with high levels of respiratory symptoms.

Phenotype D (prevalence: 18 and 14% at baseline and follow-up, respectively), labeled “allergic, high symptoms, treatment, BHR” showed similarities with phenotype C, but, compared with phenotype C, subjects belonging to this group were more frequently sensitized to aeroallergens.

Subjects belonging to phenotypes E, F, and G (prevalence: between 9 and 16%) showed moderate levels of respiratory symptoms (between phenotypes A/B and C/D), but were distinguished by their allergic status, BHR status, and FEV1 % predicted, and were therefore labeled “allergic, moderate symptoms, BHR,” “allergic, moderate symptoms, normal lung function,” and “nonallergic, moderate symptoms, no treatment,” respectively.

TABLE 3. PARAMETER ESTIMATES BY THE “CONSTRAINED” SEVEN-CLASS LATENT TRANSITION ANALYSIS MODEL IN ECRHS-SAPALDIA-EGEA (N = 3,320)

 Phenotype A: “Allergic, Few Symptoms, No Treatment”Phenotype B: “Nonallergic, Few Symptoms, No Treatment”Phenotype C: “Nonallergic, High Symptoms, Treatment”Phenotype D: “Allergic, High Symptoms, Treatment, BHR”Phenotype E: “Allergic, Moderate Symptoms, BHR”Phenotype F: “Allergic, Moderate Symptoms, Normal Lung Function”Phenotype G: “Nonallergic, Moderate Symptoms, No Treatment”
Asthma phenotype prevalences (latent class membership probabilities)
 Baseline0.210.170.080.180.100.140.11
 Follow-up0.190.160.120.140.160.140.09
Item–response probabilities
 Woken by cough 12 mo0.250.250.710.530.210.540.72
 Asthma symptom score: 00.710.610.050.000.270.070.08
 1 or 20.270.360.290.140.600.550.49
 ≥30.020.020.660.860.130.380.42
 Chronic cough/phlegm0.050.070.480.280.090.230.47
 Asthma attack 12 mo0.010.050.790.880.260.540.26
 Asthma treatment0.030.060.950.910.480.340.00
 Allergic sensitization0.950.040.130.950.910.970.12
 Total IgE ≥100 IU/ml0.530.070.190.800.760.550.18
 FEV1 <80% predicted0.020.060.260.290.250.000.09
 BHR, PD20 ≤1mg0.250.150.520.900.840.350.32
Transition from baseline (rows) to follow-up (columns)
 Phenotype A0.730.000.010.040.000.200.01
 Phenotype B0.000.780.130.000.000.010.08
 Phenotype C0.000.150.730.000.020.000.10
 Phenotype D0.000.000.000.610.380.010.00
 Phenotype E0.000.000.000.110.880.000.00
 Phenotype F0.290.000.000.030.000.670.00
 Phenotype G0.010.140.290.010.000.010.54

Definition of abbreviations: BHR = bronchial hyperresponsiveness; PD20 = provocative dose causing 20% decrease in FEV1; ECRHS = European Community Respiratory Health Survey; EGEA = Epidemiological study on Genetics and Environment of Asthma; SAPALDIA = Swiss Cohort Study on Air Pollution and Lung and Heart Disease in Adults.

Bold indicates the entries along the diagonal of the probability matrix and displays for each phenotype the estimated probabilities of membership in the same phenotype at each time point. Italics indicate a probability of ≥5% of transitioning to a different phenotype over time.

Table 4 shows odds ratios corresponding to the association between covariates (sex, age, and age at asthma onset) and latent class membership. Compared with the phenotype B (nonallergic, few symptoms, no treatment; the group with the least at risk), men were more likely to belong to the allergic phenotypes (A, D, E, and F). Older subjects were more likely to belong to phenotypes C (nonallergic, high symptoms, treatment) and E (allergic, moderate symptoms, BHR) compared with phenotype B, and subjects with adult-onset asthma were more likely to belong to nonallergic phenotypes (C and G).

TABLE 4. ODDS RATIOS* FOR SEX, AGE, AND AGE AT ASTHMA ONSET ASSOCIATED WITH ASTHMA PHENOTYPES AT BASELINE

 Phenotype A: “Allergic, Few Symptoms, No Treatment”Phenotype B: “Nonallergic, Few Symptoms, No Treatment”Phenotype C: “Nonallergic, High Symptoms, Treatment”Phenotype D: “Allergic, High Symptoms, Treatment, BHR”Phenotype E: “Allergic, Moderate Symptoms, BHR”Phenotype F: “Allergic, Moderate Symptoms, Normal Lung Function”Phenotype G: “Nonallergic, Moderate Symptoms, No Treatment”P Value
Sex (women vs. men)0.4411.180.600.220.511.15<0.0001
Age (older vs. younger)0.7511.760.632.660.380.72<0.0001
Age at asthma onset (adult vs. childhood onset asthma)0.7512.520.471.040.832.56<0.0001

Definition of abbreviation: BHR = bronchial hyperresponsiveness.

* Example: Women are less likely (OR = 0.44) to belong to phenotype A as compared to phenotype B.

The sensitivity analysis restricted to population-based cohorts, ECRHS and SAPALDIA, showed similar patterns as those observed in the main analysis (Table E2). The LTA analyses conducted separately in men and women showed strong similarities in phenotype structures for the first four phenotypes (labeled A″, B″, C″, and D″ in these subpopulations) although nonallergic phenotypes (B″ and C″) were more prevalent in women than in men (Figure E2 and Table E3).

Phenotypic Transition over Time

The probabilities of membership in the same phenotype at each time point varied from 54 to 88% across phenotypes (Figure 3 and Table 3). For example, subjects belonging to phenotype G (nonallergic, moderate symptoms, no treatment) at baseline had a 54% probability of being in the same phenotype at follow-up, whereas subjects belonging to phenotype E (allergic, moderate symptoms, BHR) at baseline had an 88% probability of being in that same phenotype at follow-up.

Many of the transitions between phenotypes occurred rarely, as shown by the transition probabilities matrix in which 65% of the probabilities had values lower than 5% (Figure 3 and Table 3). The transition probabilities between phenotypes varied from 0 to 38%. Although subjects belonging to nonallergic phenotypes (B, C, and G) do not show a stronger probability of changing phenotypes across time as compared with subjects belonging to the allergic phenotypes (A, D, E, and F), different pattern of transition across allergic phenotypes and across nonallergic phenotypes were observed. Among nonallergic phenotypes (B, C, and G), each probability of transitioning to any other nonallergic phenotype was greater than 8%. In contrast, among the allergic phenotypes (A, D, E, and F), only a few probabilities of transitioning were greater than 5%, indicating more specific transition patterns among the allergic phenotypes. Transitions toward increased asthma symptoms were more frequently observed in nonallergic phenotypes as compared with allergic phenotypes. In particular, among subjects with no or few respiratory symptoms at baseline, the chance of being in the asthma phenotypes characterized by high level of symptoms at follow-up was higher for the nonallergic phenotypes as compared with the allergic phenotypes (13 vs. <5%, respectively). Subjects belonging to the allergic phenotypes were generally more prone to move toward improvement than worsening as compared with the nonallergic phenotypes. All probabilities of transitioning from an allergic to a nonallergic phenotype, and vice versa, were below 2%.

Phenotypes A″, B″, C″, and D″ observed in men and women showed similar phenotypic structure as compared with phenotypes A, B, C, and D observed in the whole population. For these phenotypes, the probability of remaining in the same phenotype did not vary between sexes (Table E3).

Predictive Ability of the Clusters

The risk of subsequent asthma exacerbation significantly differed between the LTA-derived asthma phenotypes, phenotypes C and D showing the greater risks, phenotypes E, F, and G showing intermediate risks between phenotypes C/D and A/B (low risk; Table 5).

TABLE 5. ASSOCIATION BETWEEN ASTHMA PHENOTYPES IDENTIFIED AT BASELINE AND EXACERBATION REPORTED AT FOLLOW-UP IN ECRHS AND EGEA (N = 2,457)

 Asthma Phenotypes Identified at Baseline*
Phenotype A: “Allergic, Few Symptoms, No Treatment”Phenotype B: “Nonallergic, Few Symptoms, No Treatment”Phenotype C: “Nonallergic, High Symptoms, Treatment”Phenotype D: “Allergic, High Symptoms, Treatment, BHR”Phenotype E: “Allergic, Moderate Symptoms, BHR”Phenotype F: “Allergic, Moderate Symptoms, Normal Lung Function”Phenotype G: “Nonallergic, Moderate Symptoms, No Treatment”
Exacerbation reported at follow-up
 % (n)4.9 (20)5.0 (16)20.0 (40)17.5 (85)9.5 (21)9.2 (28)14.0 (37)
 OR (95% CI)0.98 (0.5–1.93)1 (—)4.75 (2.58–8.74)4.02 (2.31–7.01)2.04 (1.02–3.93)1.93 (1.02–3.65)3.10 (1.68–5.7)

Definition of abbreviations: BHR = bronchial hyperresponsiveness; CI = confidence interval; ECRHS = European Community Respiratory Health Survey; EGEA = Epidemiological study on Genetics and Environment of Asthma; OR = odds ratio.

* We assigned individuals to the latent classes that they had the highest probability of belonging to (based on the posterior probability provided by the model at baseline).

Overall Chi-square test <0.0001.

On a large sample of well characterized European adults with asthma (n = 3 320), a longitudinal cluster–based analysis (LTA) led to the identification of seven similar phenotypes over time (prevalence range, 8.4–20.8%), mainly characterized by level of respiratory symptoms, allergic status, and pulmonary function. These phenotypes defined at baseline were related to the subsequent risk for asthma exacerbation assessed at follow-up, which supports the predictive ability of the classification. The probability of remaining in the same phenotype at 10-year follow-up varied from 54 to 88% across phenotypes, indicating that some asthma phenotypes are more stable over time than others. Transition toward increased asthma symptoms were more frequently observed in nonallergic phenotypes as compared with allergic phenotypes.

One of the main strengths of our study relates to the populations under study. The study is based on a large sample of well characterized adults with asthma, examined in the context of three 10-year follow-up studies—ECRHS, SAPALDIA, and EGEA. The studies used standardized protocols and questionnaires. A further strength is in regard to the novelty of our approach to unraveling the asthma heterogeneity at the population level, resting on a longitudinal clustering approach, allowing us to simultaneously account for several domains of the disease repeatedly measured over time. We recognize that such an approach applied to a longitudinal dataset is exploratory and hypothesis generating. However, the phenotypes that we observed are clinically relevant. The different study designs, with two population-based studies and a case–control and family-based survey, may be seen as a limitation of the study. However, similar results were observed in the sensitivity analysis restricted to the population-based sample, suggesting that the study design does not strongly affect our findings. As a result of the study designs, the population includes both persistent and remittent asthma, and the prevalence of severe asthma in this population is low. By addressing the evolution of asthma at the population level, including subjects reporting “asthma ever,” and not only current asthma, as in clinical settings, our study may provide complementary insights into the understanding of persistent versus remittent asthma. Our study, based on two observations in a 10-year period, is an important contribution to the understanding of the long-term variability, but does not account for the short-term variability of the disease. Other studies looking to the short-term variability and its integration with long-term patterns are needed. The identified asthma phenotypes can only account for the disease features represented by the variables included in the model, and therefore do not include, for example, the inflammatory components of the disease, as such information was not available in the three studies. We did not include environmental and genetic factors in the models, as our aim was to define asthma phenotypes on the basis of the clinical characteristics. Detailed information on asthma treatment was not available at baseline; therefore, we cannot provide information on the evolution of the treatment requirements. The selection of the number of classes is not straightforward in such models. LTA models with six to eight classes led to comparable BICs, and we retained the seven-class model where the prevalence of all classes was over 5%.

Our study is the first to compare asthma phenotypes obtained in an unsupervised manner by means of LTA in the same population over a long period of time. The asthma phenotypes obtained at both time points were qualitatively very similar. By showing the consistency of such cluster-derived phenotypes over time, our study provides additional evidence to support the validity of these phenotypes. Two previous studies, applying similar clustering methods on two different adult populations, were able to show strong similarities between the derived phenotypes across European and Korean populations (11, 15). Between-study comparability of the asthma phenotypes showed similarities in the phenotypes, although populations, methods, and disease features considered differed between studies (5). Recently, Patrawalla and colleagues (26) applied a simple algorithm to cluster asthma as defined in the Severe Asthma Research Program population (considering baseline FEV1, maximal post-bronchodilator FEV1, and age at asthma onset) in an urban population, and revealed five groups phenotypically similar to those identified in Severe Asthma Research Program subjects. Further evidence of validity of cluster-derived asthma phenotypes is provided by the assessment of the discriminative properties and effect of clinical prognosis of these phenotypes (8, 9, 11, 13, 14). Further studies on long-term follow-up cohorts are needed to replicate our findings.

The longitudinal LTA approach revealed seven phenotypes, four of which were highly comparable to the four LCA-derived asthma phenotypes previously reported in ECRHS and EGEA (11). We previously showed that these phenotypes clearly discriminated populations in terms of quality of life and blood eosinophil and neutrophil counts, better than classical phenotypes (11). The three novel phenotypes, mainly characterized by subjects with a moderate level of symptoms, differed with regard to the allergic status, BHR, and their evolution over time. As compared with LCA, LTA allows accounting for additional information from other points in time, and therefore leads to a refinement of asthma phenotypes by the identification of additional clusters. Further analyses are needed to address the relevance of such phenotypes in epidemiology and in asthma management.

Our findings show strong stability of the asthma phenotypes with regard to allergic status over time, indicating small changes in allergy in this adult population over a 10-year period. Prospective studies on the long-term evolution of allergy among adults are rare. Some adult population–based studies consistently found a low prevalence and a high remission of allergic sensitization at a 10-year (27) and a 25-year follow-up (28), whereas only a small change in specific IgE levels were observed over 10 years in ECRHS (29). Such data from cohorts of subjects with asthma is missing from the literature.

This study is the first to model the probabilities of transitioning over time between comprehensive asthma phenotypes defined by a clustering approach. Our findings clearly indicate different patterns of transition over time across phenotypes, with some phenotypes showing higher stability than others and with transitions between phenotypes varying from implausible to very likely. Transitions toward increased asthma symptoms were more frequently observed in nonallergic phenotypes as compared with allergic phenotypes, indicating that allergic status may have an effect in the course of asthma activity in adulthood. Few studies have focused on the role of allergy in the natural history of asthma in adults. A recent Canadian study using health administrative databases did not suggest an effect of allergy in the course of asthma activity (30). Our results on the longitudinal analysis of phenotypic characteristics support the evidence derived from genetic investigations (31), that studies should be conducted to better understand the relationships of allergic characteristics with asthma expression.

Using this data-driven approach, strong similarities in phenotypic characteristics and evolution were observed between asthma phenotypes identified in men and women. Nevertheless, consistent with previous findings showing that women are at greater risk for nonallergic asthma (32), the prevalence of nonallergic asthma phenotypes was higher in women than in men.

From a methodological point of view, such a longitudinal approach may allow us to limit problems encountered in using incidence to analyze risk factors in follow-up studies (33). Indeed, as opposed to the analysis of incidence that assumes that asthma is a true dichotomous disease, which is unlikely to be true, LTA provides a novel modeling approach for longitudinal studies that does not make such an assumption. We therefore suggest that such modeling may help to discover the causes of asthma.

Although we emphasize the combined role of clinical characteristics routinely collected in the disease evaluation and provide novel insights into the natural history of asthma, it is too early to say to what extent this methodology will result in preventive and therapeutically relevant findings. However, our results provide the opportunity to investigate both the determinants and the outcomes of these phenotypes, which may eventually provide relevant, applicable findings.

In summary, this study adds new evidence to support the interest in cluster-derived asthma phenotypes for a better understanding of this complex disease and for asthma management. Although further studies with longer follow-up are warranted, clustering techniques applied to longitudinal data, allowing variability over time to be taken into account, may identify more accurate asthma phenotypes. This could facilitate the identification of risk factors, and may lead to the identification of specific phenotypes more prone to worsen over time.

The authors thank the Epidemiological Study on Genetics and Environment of Asthma (EGEA) cooperative group members:

Coordination: F. Kauffmann; V. Siroux (epidemiology); F. Demenais (genetics); I. Pin (clinical aspects); R. Nadif (biology). Respiratory epidemiology: Institut National de la Santé et de la Recherche Médicale (INSERM) U 700, Paris: M. Korobaeff (Epidemiological Study on Genetics and Environment of Asthma [EGEA] 1), F. Neukirch (EGEA1); INSERM U 707, Paris: I. Annesi-Maesano; INSERM Centre for Research in Epidemiology and Population Health (CESP)/U 1018, Villejuif: F. Kauffmann, N. Le Moual, R. Nadif, M. P. Oryszczyn; INSERM U 823, Grenoble: V. Siroux. Genetics: INSERM U 393, Paris: J. Feingold; INSERM U 946, Paris: E. Bouzigon, F. Demenais, M. H. Dizier; Centre National de Génotypage, Evry: I. Gut (now National Centre for Genomic Analysis), M. Lathrop (now Centre d'Etude du Polymorphisme Humain/McGill). Clinical Centers: Grenoble: I. Pin, C. Pison; Lyon: D. Ecochard (EGEA1), F. Gormand, Y. Pacheco; Marseille: D. Charpin (EGEA1), D. Vervloet; Montpellier: J. Bousquet; Paris Cochin: A. Lockhart (EGEA1), R. Matran (now in Lille); Paris Necker: E. Paty, P. Scheinmann; Paris-Trousseau: A. Grimfeld, J. Just. Data and quality management: INSERM ex-U155 (EGEA1): J. Hochez; INSERM CESP/U 1018, Villejuif: N. Le Moual, INSERM ex-U780: C. Ravault; INSERM ex-U794: N. Chateigner; Grenoble: J. Ferran.

The authors thank all those who participated in the setting of the study and on the various aspects of the examinations involved: interviewers, technicians for lung function testing, coders, those involved in quality control and data management, and all those who supervised the study in all centers. The authors are grateful to the three CIC-INSERM of Necker, Grenoble, and Marseille who supported the study and in which subjects were examined. They are indebted to all the subjects who participated, without whom the study would not have been possible.

The European Community Respiratory Health Survey (ECRHS) data incorporated in this analysis would not have been available without the collaboration of the following individuals and their research teams:

ECRHS Coordinating Center: P. Burney, D. Jarvis, S. Chinn, J. Knox (ECRHS II), C. Luczynska, J. Potts.

Steering Committee for ECRHS II: P. Burney, D. Jarvis, S. Chinn, U. Ackermann-Liebrich, J. M. Anto, I. Cerveri, R. de Marco, T. Gislason, J. Heinrich, C. Janson, N. Kunzli, B. Leynaert, F. Neukirch, J. Schouten, J. Sunyer; C. Svanes, P. Vermeire, M. Wjst.

Principal Investigators and Senior Scientific Teams for ECRHS II: Australia: Melbourne (M. Abramson, R. Woods, E. H. Walters, F. Thien); Belgium: South Antwerp & Antwerp City (P. Vermeire [deceased], J. Weyler, M. Van Sprundel, V. Nelen); Denmark: Aarhus (E. J. Jensen); Estonia: Tartu (R. Jogi, A. Soon); France: Paris (F. Neukirch, B. Leynaert, R. Liard, M. Zureik); Grenoble (I. Pin, J. Ferran-Quentin); Bordeaux (A. Taytard, C. Raherison); Montpellier (J. Bousquet, P. Demoly); Germany: Erfurt (J. Heinrich, M. Wjst, C. Frye, I. Meyer); Hamburg (K. Richter); Iceland: Reykjavik (T. Gislason, E. Bjornsson, D. Gislason, T. Blondal, A. Karlsdottir); Italy: Turin (M. Bugiani, P. Piccioni, E. Caria, A. Carosso, E. Migliore, G. Castiglioni); Verona (R. de Marco, G. Verlato, E. Zanolin, S. Accordini, A. Poli, V. Lo Cascio, M. Ferrari); Pavia (A. Marinoni, S. Villani, M. Ponzio, F. Frigerio, M. Comelli, M. Grassi, I. Cerveri, A. Corsico); The Netherlands: Groningen & Geleen (J. Schouten, M. Kerkhof); Norway: Bergen (A. Gulsvik, E. Omenaas, C. Svanes, B. Laerum); Spain: Barcelona (J. M. Anto, J. Sunyer, M. Kogevinas, J. P. Zock, X. Basagana, A. Jaen, F. Burgos); Huelva (J. Maldonado, A. Pereira, J. L. Sanchez); Albacete (J. Martinez-Moratalla Rovira, E. Almar); Galdakao (N. Muniozguren, I. Urritia); Oviedo (F. Payo); Sweden: Uppsala (C. Janson, G. Boman, D. Norback, M. Gunnbjornsdottir); Goteborg (K. Toren, L. Lillienberg, A. C. Olin, B. Balder, A. Pfeifer-Nilsson, R. Sundberg); Umea (E. Norrman, M. Soderberg, K. Franklin, B. Lundback, B. Forsberg, L. Nystrom); Switzerland: Basel (N. Kunzli, B. Dibbert, M. Hazenkamp, M. Brutsche, U. Ackermann-Liebrich); United Kingdom: Norwich (D. Jarvis, B. Harrison); Ipswich (D. Jarvis, R. Hall, D. Seaton); United States: Portland (M. Osborne, S. Buist, W. Vollmer, L. Johnson).

The authors thank the Swiss Cohort Study on Air Pollution and Lung and Heart Disease in Adults (SAPALDIA) study team members:

Study Directorate: T. Rochat (pneumology [p]), J. M. Gaspoz (cardiology [c]), N. Künzli (epidemiology [e]/exposure [exp]), N. M. Probst Hensch (e/genetic and molecular biology [g]), C. Schindler (statistics [s]).

Scientific Team: J. C. Barthélémy (c), W. Berger (g), R. Bettschart (p), A. Bircher (allergology [a]), G. Bolognini (p), O. Brändli (p), C. Brombach (nutrition [n]), M. Brutsche (p), L. Burdet (p), M. Frey (p), U. Frey (pediatrics [pd]), M. W. Gerbase (p), D. Gold (e/c/p), E. de Groot (c), W. Karrer (p), R. Keller (p), B. Knöpfli (p), B. Martin (physical activity [pa]), D. Miedinger (occupational health [o]), U. Neu (exp), L. Nicod (p), M. Pons (p), F. Roche (c), T. Rothe (p), E. Russi (p), P. Schmid-Grendelmeyer (a), A. Schmidt-Trucksäss (pa), A. Turk (p), J. Schwartz (e), D. Stolz (p), P. Straehl (exp), J. M. Tschopp (p), A. von Eckardstein (clinical chemistry [cc]), E. Zemp Stutz (e).

Scientific Team at Coordinating Centers: M. Adam (e/g), E. Boes (g), P. O. Bridevaux (p), D. Carballo (c), E. Corradi (e), I. Curjuric (e), J. Dratva (e), A. Di Pasquale (s), L. Grize (s), D. Keidel (s), S. Kriemler (pa), A. Kumar (g), M. Imboden (g), N. Maire (s), A. Mehta (e), F. Meier (e), H. Phuleria (exp), E. Schaffner (s), G. A. Thun (g) A. Ineichen (exp), M. Ragettli (e), M. Ritter (exp), T. Schikowski (e), G. Stern (pd), M. Tarantino (s), M. Tsai (e), M. Wanner (pa).

The study could not have been done without the help of the study participants, technical and administrative support, and the medical teams and field workers at the local study sites:

Local Fieldworkers: Aarau—S. Brun, G. Giger, M. Sperisen, M. Stahel; Basel—C. Bürli, C. Dahler, N. Oertli, I. Harreh, F. Karrer, G. Novicic, N. Wyttenbacher; Davos—A. Saner, P. Senn, R. Winzeler; Geneva—F. Bonfils, B. Blicharz, C. Landolt, J. Rochat; Lugano—S. Boccia, E. Gehrig, M. T. Mandia, G. Solari, B. Viscardi; Montana—A. P. Bieri, C. Darioly, M. Maire, Payerne—F. Ding, P. Danieli A. Vonnez; Wald—D. Bodmer, E. Hochstrasser, R. Kunz, C. Meier, J. Rakic, U. Schafroth, A. Walder.

Administrative Staff: C. Gabriel, R. Gutknecht.

The Epidemiological Study on Genetics and Environment of Asthma (EGEA) was supported, in part, by grants from: Merck Sharp and Dohme; the Hospital Program of Clinical Research (Paris, France); National Research Agency (ANR) Health Environment, Health-Work Program; ANR Biological Collections for Health Program; French Agency of Health Safety; Environment and Work; National Scientific Committee of the Medico-Technology Support at Home; and Isere Committee Against Respiratory Diseases.

The Swiss Study on Air Pollution and Lung and Heart Diseases in Adults (SAPALDIA) was supported by the Swiss National Science Foundation grants 33CS30_134276/1, 33CSCO-108796, 3247BO-104283, 3247BO-104288, 3247BO-104284, 3247-065896, 3100-059302, 3200-052720, 3200-042532, 4026-028099, 3233-054996, and PDFMP3-123171, the Federal Office for Forest, Environment and Landscape, the Federal Office of Public Health, the Federal Office of Roads and Transport, the canton’s government of Aargau, Basel-Stadt, Basel-Land, Geneva, Luzern, Ticino, Valais, Zurich, the Swiss Lung League, the canton’s Lung League of Basel Stadt/Basel Landschaft, Geneva, Ticino, Valais and Zurich, Schweizerische Unfallversicherungsanstalt (SUVA), Freiwillige Akademische Gesellschaft, UBS Wealth Foundation, Talecris Biotherapeutics GmbH, and Abbott Diagnostics.

List of bodies that funded the local studies in the European Community Respiratory Health Survey (ECRHS) II: Albacete—Fondo de Investigaciones Santarias grants 97/0035-01, 13 99/0034-01, and 99/0034-02, Hospital Universitario de Albacete, Consejeria de Sanidad. Antwerp—Fund for Scientific Research, Flanders Belgium grant G.0402.00, University of Antwerp, Flemish Health Ministry. Barcelona—Fondo de Investigaciones Sanitarias grants 99/0034-01 and 99/0034-02, Red Respira grant RTIC 03/11 ISC IIF. CIBER of Epidemiology and Public Health has been established and founded by Instituto de Salud Carlos III. Basel—Swiss National Science Foundation, Swiss Federal Office for Education and Science, SUVA. Bergen—Norwegian Research Council; Norwegian Asthma and Allergy Association; Glaxo Wellcome AS, Norway Research Fund. Bordeaux—Institut Pneumologique d’Aquitaine. Erfurt—GSF-National Research Centre for Environment and Health, Deutsche Forschungsgemeinschaft (DFG) grant FR 1,526/1-1. Galdakao—Basque Health Department. Gothenburg—Swedish Heart Lung Foundation, Swedish Foundation for Health Care Sciences and Allergy Research, Swedish Asthma and Allergy Foundation, Swedish Cancer and Allergy Foundation. Grenoble—Program Hospitalier de Recherche Clinique-DRC de Grenoble 2,000 no. 2610, Ministry of Health, Direction de la Recherche Clinique, Ministere de l’Emploi et de la Solidarite, Direction Generale de la Sante, CHU de Grenoble, Comite des Maladies Respiratoires de l’Isere. Hamburg—GSF-National Research Centre for Environment and Health, DFG grant MA 711/4-1. Ipswich and Norwich—National Asthma Campaign (UK). Huelva—Fondo de Investigaciones Sanitarias (FIS) grants 97/0035-01, 99/0034-01, and 99/0034-02. Montpellier—Program Hospitalier de Recherche Clinique-DRC de Grenoble 2,000 no. 2610, Ministry of Health, Direction de la Recherche Clinique, CHU de Grenoble, Ministere de l’Emploi et de la Solidarite, Direction Generale de la Sante, Aventis (France), Direction Régionale des Affaires Sanitaires et Sociales Languedoc-Roussillon. Oviedo—FIS grants 97/0035-01, 99/0034-01, and 99/0034–02. Paris—Ministere de l’Emploi et de la Solidarite, Direction Generale de la Sante, UCBPharma (France), Aventis (France), Glaxo France, Program Hospitalier de Recherche Clinique-DRC de Grenoble 2,000 no. 2610, Ministry of Health, Direction de la Recherche Clinique, CHU de Grenoble. Pavia—Glaxo, Smith and Kline Italy, Italian Ministry of University and Scientific and Technological Research (MURST), local University Funding for Research 1998 and 1999 (Pavia, Italy). Portland—American Lung Association of Oregon, Northwest Health Foundation, Collins Foundation, Merck Pharmaceutical. Reykjavik—Icelandic Research Council, Icelandic University Hospital Fund. Tartu—Estonian Science Foundation. Turin—ASL 4 Regione Piemonte (Italy), AO CTO/ICORMA Regione Piemonte (Italy), Ministero dell’Università e della Ricerca Scientifica (Italy), Glaxo Wellcome spa (Verona, Italy). Umeå—Swedish Heart Lung Foundation, Swedish Foundation for Health Care Sciences and Allergy Research, Swedish Asthma and Allergy Foundation, Swedish Cancer and Allergy Foundation. Uppsala—Swedish Heart Lung Foundation, Swedish Foundation for Health Care Sciences and Allergy Research, Swedish Asthma and Allergy Foundation, Swedish Cancer and Allergy Foundation. Verona—University of Verona; Italian MURST; Glaxo, Smith and Kline Italy.

List of bodies that funded ECRHS I for centers in ECRHS II: Belgian Science Policy Office, National Fund for Scientific Research; Ministère de la Santé, Glaxo France, Institut Pneumologique d’Aquitaine, Contrat de Plan Etat-Région Languedoc-Rousillon, CNMATS, CNMRT (90MR/10, 91AF/6), Ministre delegué de la santé, RNSP, France; GSF, and the Bundesminister für Forschung und Technologie, Bonn, Germany; Ministero dell’Università e della Ricerca Scientifica e Tecnologica, CNR, Regione Veneto grant RSF n. 381/05.93, Italy; Norwegian Research Council project no. 101,422/310; Dutch Ministry of Wellbeing, Public Health and Culture, Netherlands; Ministerio de Sanidad y Consumo FIS grants 91/0016060/00E-05E and 93/0393, and grants from Hospital General de Albacete, Hospital General Juan Ramón Jiménez, Consejería de Sanidad, Principado de Asturias, Spain; The Swedish Medical Research Council, the Swedish Heart Lung Foundation, the Swedish Association against Asthma and Allergy; Swiss National Science Foundation grant 4026-28099; National Asthma Campaign, British Lung Foundation, Department of Health, South Thames Regional Health Authority, UK; United States Department of Health, Education, and Welfare Public Health Service grant 2 S07 RR05521-28.

1. Panettieri RA Jr, Covar R, Grant E, Hillyer EV, Bacharier L. Natural history of asthma: persistence versus progression—does the beginning predict the end? J Allergy Clin Immunol 2008;121:607613.
2. Spahn JD, Covar R. Clinical assessment of asthma progression in children and adults. J Allergy Clin Immunol 2008;121:548557, quiz 558–559.
3. Green RH, Pavord I. Stability of inflammatory phenotypes in asthma. Thorax 2012;67:665667.
4. Antó JM, Pinart M, Akdis M, Auffray C, Bachert C, Basagaña X, Carlsen KH, Guerra S, von Hertzen L, Illi S, et al.; WHO Collaborating Centre on Asthma and Rhinitis (Montpellier). Understanding the complexity of IgE-related phenotypes from childhood to young adulthood: a Mechanisms of the Development of Allergy (MeDALL) seminar. J Allergy Clin Immunol 2012;129:943954, e4.
5. Wenzel SE. Asthma phenotypes: the evolution from clinical to molecular approaches. Nat Med 2012;18:716725.
6. Siroux V, Garcia-Aymerich J. The investigation of asthma phenotypes. Curr Opin Allergy Clin Immunol 2011;11:393399.
7. Wenzel SE. Asthma: defining of the persistent adult phenotypes. Lancet 2006;368:804813.
8. Moore WC, Meyers DA, Wenzel SE, Teague WG, Li H, Li X, D’Agostino R Jr, Castro M, Curran-Everett D, Fitzpatrick AM, et al.; National Heart, Lung, and Blood Institute’s Severe Asthma Research Program. Identification of asthma phenotypes using cluster analysis in the Severe Asthma Research Program. Am J Respir Crit Care Med 2010;181:315323.
9. Haldar P, Pavord ID, Shaw DE, Berry MA, Thomas M, Brightling CE, Wardlaw AJ, Green RH. Cluster analysis and clinical asthma phenotypes. Am J Respir Crit Care Med 2008;178:218224.
10. Fitzpatrick AM, Teague WG, Meyers DA, Peters SP, Li X, Li H, Wenzel SE, Aujla S, Castro M, Bacharier LB, et al.; National Institutes of Health/National Heart, Lung, and Blood Institute Severe Asthma Research Program. Heterogeneity of severe asthma in childhood: confirmation by cluster analysis of children in the National Institutes of Health/National Heart, Lung, and Blood Institute Severe Asthma Research Program. J Allergy Clin Immunol 2011;127:382389, e1–e13.
11. Siroux V, Basagaña X, Boudier A, Pin I, Garcia-Aymerich J, Vesin A, Slama R, Jarvis D, Anto JM, Kauffmann F, et al. Identifying adult asthma phenotypes using a clustering approach. Eur Respir J 2011;38:310317.
12. Henderson J, Granell R, Heron J, Sherriff A, Simpson A, Woodcock A, Strachan DP, Shaheen SO, Sterne JA. Associations of wheezing phenotypes in the first 6 years of life with atopy, lung function and airway responsiveness in mid-childhood. Thorax 2008;63:974980.
13. Savenije OE, Granell R, Caudri D, Koppelman GH, Smit HA, Wijga A, de Jongste JC, Brunekreef B, Sterne JA, Postma DS, et al. Comparison of childhood wheezing phenotypes in 2 birth cohorts: ALSPAC and PIAMA. J Allergy Clin Immunol 2011;127:15051512, e14.
14. Spycher BD, Silverman M, Kuehni CE. Phenotypes of childhood asthma: are they real? Clin Exp Allergy 2010;40:11301141.
15. Kim TB, Jang AS, Kwon HS, Park JS, Chang YS, Cho SH, Choi BW, Park JW, Nam DH, Yoon HJ, et al. Identification of asthma clusters in two independent korean adult asthma cohorts. Eur Respir J 2013;41:13081314.
16. Just J, Gouvis-Echraghi R, Couderc R, Guillemot-Lambert N, Saint-Pierre P. Novel severe wheezy young children phenotypes: boys atopic multiple-trigger and girls nonatopic uncontrolled wheeze. J Allergy Clin Immunol 2012;130:103110, e8.
17. European Community Respiratory Health Survey II Steering Committee. The European Community Respiratory Health Survey II. Eur Respir J 2002;20:10711079.
18. Siroux V, Boudier A, Bousquet J, Bresson JL, Cracowski JL, Ferran J, Gormand F, Just J, Le Moual N, Morange S, et al.; Epidemiological Study on the Genetics and Environment of Asthma. Phenotypic determinants of uncontrolled asthma. J Allergy Clin Immunol 2009;124:681687, e3.
19. Ackermann-Liebrich U, Kuna-Dibbert B, Probst-Hensch NM, Schindler C, Felber Dietrich D, Stutz EZ, Bayer-Oglesby L, Baum F, Brändli O, Brutsche M, et al.; SAPALDIA Team. Follow-up of the Swiss Cohort Study on Air Pollution and Lung Diseases in Adults (Sapaldia 2) 1991–2003: methods and characterization of participants. Soz Praventivmed 2005;50:245263.
20. Kauffmann F, Dizier MH, Pin I, Paty E, Gormand F, Vervloet D, Bousquet J, Neukirch F, Annesi I, Oryszczyn MP, et al. Epidemiological study of the genetics and environment of asthma, bronchial hyperresponsiveness, and atopy: phenotype issues. Am J Respir Crit Care Med 1997;156:S123S129.
21. Imboden M, Bouzigon E, Curjuric I, Ramasamy A, Kumar A, Hancock DB, Wilk JB, Vonk JM, Thun GA, Siroux V, et al. Genome-wide association study of lung function decline in adults with and without asthma. J Allergy Clin Immunol 2012;129:12181228.
22. de Marco R, Marcon A, Jarvis D, Accordini S, Almar E, Bugiani M, Carolei A, Cazzoletti L, Corsico A, Gislason D, et al.; European Community Respiratory Health Survey Therapy Group. Prognostic factors of asthma severity: a 9-year international prospective cohort study. J Allergy Clin Immunol 2006;117:12491256.
23. Lanza ST, Patrick ME, Maggs JL. Latent transition analysis: benefits of a latent variable approach to modeling transitions in substance use. J Drug Issues 2010;40:93120.
24. Lanza ST, Collins LM. A new SAS procedure for latent transition analysis: transitions in dating and sexual risk behavior. Dev Psychol 2008;44:446456.
25. Lanza ST, Bray BC. Transitions in drug use among high-risk women: an application of latent class and latent transition analysis. Adv Appl Stat Sci 2010;3:203235.
26. Patrawalla P, Kazeros A, Rogers L, Shao Y, Liu M, Fernandez-Beros ME, Shang S, Reibman J. Application of the asthma phenotype algorithm from the Severe Asthma Research Program to an urban population. PLoS ONE 2012;7:e44540.
27. Warm K, Backman H, Lindberg A, Lundbäck B, Rönmark E. Low incidence and high remission of allergic sensitization among adults. J Allergy Clin Immunol 2012;129:136142.
28. Linneberg A, Gislum M, Johansen N, Husemoen LL, Jørgensen T. Temporal trends of aeroallergen sensitization over twenty-five years. Clin Exp Allergy 2007;37:11371142.
29. Jarvis D, Luczynska C, Chinn S, Potts J, Sunyer J, Janson C, Svanes C, Künzli N, Leynaert B, Heinrich J, et al. Change in prevalence of IgE sensitization and mean total IgE with age and cohort. J Allergy Clin Immunol 2005;116:675682.
30. Gershon A, Guan J, Victor JC, Wang C, To T. The course of asthma activity: a population study. J Allergy Clin Immunol 2012;129:679686.
31. Moffatt MF, Gut IG, Demenais F, Strachan DP, Bouzigon E, Heath S, von Mutius E, Farrall M, Lathrop M, Cookson WO; GABRIEL Consortium. A large-scale, consortium-based genomewide association study of asthma. N Engl J Med 2010;363:12111221.
32. Leynaert B, Sunyer J, Garcia-Esteban R, Svanes C, Jarvis D, Cerveri I, Dratva J, Gislason T, Heinrich J, Janson C, et al. Gender differences in prevalence, diagnosis and incidence of allergic and non-allergic asthma: a population-based cohort. Thorax 2012;67:625631.
33. Pekkanen J, Sunyer J. Problems in using incidence to analyze risk factors in follow-up studies. Eur J Epidemiol 2008;23:581584.
Correspondence and requests for reprints should be addressed to Valérie Siroux, INSERM U823, Institut Albert Bonniot—INSERM–University J. Fourier (U823) Research Center, F-38042 Grenoble Cedex 9, France. E-mail:

Supported by the “AGIR à Dom” scientific committee. A list of support of The Epidemiological Study on Genetics and Environment of Asthma, The Swiss Study on Air Pollution and Lung and Heart Diseases in Adults, the local studies in the European Community Respiratory Health Survey II, and ECRHS I for centers in ECRHS II may be found before the beginning of the References.

Author Contributions: A.B., V.S., and F.K. were involved in the conception, hypotheses delineation and design of the study; J.M.A., J.B., P.O.B., J.H., C.J., N.K., B.L., R.d.M., T.R., C.S., I.P., N.P.-H., J.S., F.K., and V.S. participated in the acquisition of data; X.B., J.G.-A., and I.C. were involved as experts in clustering modeling approaches; A.B., V.S., H.H., and I.P. participated in the data analysis and interpretation; A.B. and V.S. wrote the article; I.C., X.B., H.H., J.M.A., J.B., P.O.B., E.D.-L., J.G.-A., J.H., C.J., N.K., B.L., R.d.M., T.R., C.S., R.V., I.P., N.P.-H., J.S., and F.K. reviewed the manuscript and revised it critically. All authors approved the final version of the manuscript.

This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org

Originally Published in Press as DOI: 10.1164/rccm.201301-0156OC on June 18, 2013

Author disclosures are available with the text of this article at www.atsjournals.org.

Related

No related items
American Journal of Respiratory and Critical Care Medicine
188
5

Click to see any corrections or updates and to confirm this is the authentic version of record