Advertisement

High BMI is related to higher incidence of asthma, while a fish and fruit diet is related to a lower–

Results from a long-term follow-up study of three age groups in Sweden
Open ArchivePublished:February 19, 2010DOI:https://doi.org/10.1016/j.rmed.2009.12.013

      Summary

      The causes of the worldwide increase in asthma seen during the last decades remain largely unexplained, but lifestyle and diet are suggested to play important roles. In this follow up of a large-scale population sample in Sweden, we wanted to identify modifiable risk factors for the cumulative incidence over a 13-year follow-up period.
      In 1990, a self-administered questionnaire was completed by 12,560 individuals from three age groups (16, 30–39 and 60–69 years of age) in two counties of Sweden. In 2003, the eligible subjects (n = 11,282) were sent a new postal questionnaire.
      In total 8150 (response rate 73%) answered the questionnaire. The prevalence of asthma in 2003 had increased in all ages. In the young adults, the asthma prevalence rose from 11.3% in 1990 to 25.0% in 2003. Adult asthma onset was identified in 791 of the participants. Smoking [RR (95% CI) = 1.37 (1.12–1.68)], BMI [1.49 (1.25–1.77 per inter quartile range)], and nocturnal gastro-oesophageal reflux (GOR) [2.16 (1.72–2.72)] were significant independent risk factors for the cumulative incidence of asthma. The impact of risk factors differed between the age groups where BMI and GOR had a significantly higher impact in the middle aged and the elderly (p < 0.05). High consumption of fruit and fish was protective especially in the elderly [0.52 (0.35–0.77)]. No significant difference was found in the impact of risk factors between men and women.
      Weight loss, smoking cessation and a diet rich in fruit and fish may be of importance in preventing onset of adult asthma.

      Keywords

      Introduction

      The prevalence of asthma has increased dramatically during the last decades, especially among children.
      • Beasley R.
      • Crane J.
      • Lai C.K.
      • Pearce N.
      Prevalence and etiology of asthma.
      In Sweden, the reported prevalence has increased from 1.9% 1971, 2.8% 1981,
      • Åberg N.
      Asthma and allergic rhinitis in Swedish conscripts.
      3.5% 1985
      • Stjernberg N.
      • Eklund A.
      • Nyström L.
      • Rosenhall L.
      • Emmelin A.
      • Strömqvist L.H.
      Prevalence of bronchial asthma and chronic bronchitis in a community in northern Sweden; relation to environmental and occupational exposure to sulphur dioxide.
      and 5.0% 1991.
      • Lundbäck B.
      • Nyström L.
      • Rosenhall L.
      • Stjernberg N.
      Obstructive lung disease in northern Sweden: respiratory symptoms assessed in a postal survey.
      A recent study,
      • Ekerljung L.
      • Rönmark E.
      • Larsson K.
      • Sundblad B.M.
      • Bjerg A.
      • Ahlstedt S.
      • et al.
      No further increase of incidence of asthma: incidence, remission and relapse of adult asthma in Sweden.
      has concluded that the incidence of asthma among adults has been stable in Sweden for the past two decades, which might indicate that a plateau is reached. This finding is also consistent with a tendency of decreasing childhood asthma in Sweden.
      • Almqvist C.
      • Pershagen G.
      • Wickman M.
      Low socioeconomic status as a risk factor for asthma, rhinitis and sensitization at 4 years in a birth cohort.
      Great efforts have been made to investigate risk factors for asthma development in childhood, but more than half of the cohort of young adults in the European Community Respiratory Health Survey (ECRHS)
      • Sunyer J.
      • Anto J.M.
      • Tobias A.
      • Burney P.
      Generational increase of self-reported first attack of asthma in fifteen industrialized countries. European Community Respiratory Health Study (ECRHS).
      reported a debut of initial asthma-like symptoms above the age of 15 years. This finding might suggest that the majority of adult asthmatics have a disease starting in adulthood, but it could as well reflect temporary remissions of asthma during the teen ages.
      Our knowledge and understanding of long-term changes in asthma prevalence within populations as they age is still limited. Investigation of risk factors for asthma starting in adult age is important. Previous studies indicate that smoking,
      • Toren K.
      • Hermansson B.A.
      Incidence rate of adult-onset asthma in relation to age, sex, atopy and smoking: a Swedish population-based study of 15,813 adults.
      • Piipari R.
      • Jaakkola J.J.
      • Jaakkola N.
      • Jaakkola M.S.
      Smoking and asthma in adults.
      occupational exposure,
      • Kogevinas M.
      • Anto J.M.
      • Sunyer J.
      • Tobias A.
      • Kromhout H.
      • Burney P.
      Occupational asthma in Europe and other industrialised areas: a population-based study. European community respiratory health survey study group.
      • Kogevinas M.
      • Zock J.P.
      • Jarvis D.
      • Kromhout H.
      • Lillienberg L.
      • Plana E.
      • et al.
      Exposure to substances in the workplace and new-onset asthma: an international prospective population-based study (ECRHS-II).
      obesity
      • Camargo Jr., C.A.
      • Weiss S.T.
      • Zhang S.
      • Willett W.C.
      • Speizer F.E.
      Prospective study of body mass index, weight change, and risk of adult-onset asthma in women.
      • Shaheen S.O.
      • Sterne J.A.
      • Montgomery S.M.
      • Azima H.
      Birth weight, body mass index and asthma in young adults.
      and nocturnal gastro-oesophageal reflux
      • Gunnbjörnsdottir M.I.
      • Omenaas E.
      • Gislason T.
      • Norrman E.
      • Olin A.C.
      • Jogi R.
      • et al.
      Obesity and nocturnal gastro-oesophageal reflux are related to onset of asthma and respiratory symptoms.
      • Ruigomez A.
      • Rodriguez L.A.
      • Wallander M.A.
      • Johansson S.
      • Thomas M.
      • Price D.
      Gastroesophageal reflux disease and asthma: a longitudinal study in UK general practice.
      increase the risk of asthma in adulthood. Data collected by longitudinal studies gives a more confident evaluation of the importance of different risk factors. Consequently it is important to follow-up large cross-sectional studies like the ECRHS II
      • Chinn S.
      • Jarvis D.
      • Burney P.
      • Luczynska C.
      • Ackermann-Liebrich U.
      • Anto J.M.
      • et al.
      Increase in diagnosed asthma but not in symptoms in the European community respiratory health survey.
      which included respondents from many different geographical areas in more than 15 countries. Longitudinal studies of large cohorts from more limited geographical areas will also add important information. The purpose of this longitudinal study of a large cohort collected in 1990 from a defined geographical area in Sweden
      • Larsson L.
      • Boethius G.
      • Uddenfeldt M.
      Differences in utilization of asthma drugs between two neighbouring Swedish provinces: relation to symptom reporting.
      was to investigate risk factors for asthma development in adult age in three age groups.

      Methods

      The study design and target population

      In 1990 and 2003, a total of 12,500 postal questionnaires were distributed to all individuals born in 1974 and a random sample of individuals born 1951–1960 and 1921–1931 in the counties of Gästrikland and Jämtland in the central part of Sweden. The overall response rate was 90%. In 2003, 2190 young adults, 3557 middle aged and 2319 elderly of the 11,230 eligible subjects responded to the questionnaire. Subjects not responding to the first mailing received two reminders. A total of 8150 (73%) subjects, comprising 3817 men and 4333 women, answered the questionnaire 2003, but the analyses in this study are based on the 67% (n = 7563) who participated in both surveys (1990 + 2003).

      Questionnaire

      The first part of the questionnaire 2003 contained 22 items identical to those in 1990, including questions on respiratory symptoms, asthma, hay fever, heredity, and smoking habits. This questionnaire was a modification of a questionnaire used previously in studies in northern Sweden,
      • Lundbäck B.
      • Nyström L.
      • Rosenhall L.
      • Stjernberg N.
      Obstructive lung disease in northern Sweden: respiratory symptoms assessed in a postal survey.
      based on the British Medical Research Council questionnaire. The second part of the questionnaire 2003 included 92 new items covering e.g. weight, height, physical activity, dietary factors, gastro-oesophageal reflux, sleep disorders, snoring, occupational exposure and indoor environment: most of those previously used in ECRHS.
      • Chinn S.
      • Jarvis D.
      • Burney P.
      • Luczynska C.
      • Ackermann-Liebrich U.
      • Anto J.M.
      • et al.
      Increase in diagnosed asthma but not in symptoms in the European community respiratory health survey.

      Case definition

      An asthma case was defined by answering “yes” to one or more of the questions; “Do you have or have you had asthma?”, “Have you ever been diagnosed with asthma by a doctor?”, “Do you use asthma medication?” and “Do you have or have you had symptoms of asthma?” Cumulative incidence of asthma was defined as answering “no” to all four of these questions in 1990 and “yes” to at least one of these questions in 2003.

      Asthma heredity

      Family history of asthma was defined as answering “yes” to the question “Do your parents, siblings or children have asthma?” in 1990.

      Body mass index

      Body mass index (BMI) was calculated for each subject as self-reported weight in kilograms in 2003 divided by the squared height in meters (kg/m2).

      Gastro-oesophageal reflux and snoring

      The question asked in 2003 regarding gastro-oesophageal reflux (GOR) was: “Do you have heartburn or belching when you have gone to bed?” Subjects reporting these symptoms 1–2 nights per week were, in this study, referred to as reporting nocturnal GOR.
      • Gunnbjörnsdottir M.I.
      • Omenaas E.
      • Gislason T.
      • Norrman E.
      • Olin A.C.
      • Jogi R.
      • et al.
      Obesity and nocturnal gastro-oesophageal reflux are related to onset of asthma and respiratory symptoms.
      The question asked regarding snoring was: “Do you snore loudly and disturbingly?” Subjects reporting snoring 3–5 times per week were referred to as reporting habitual snoring.
      • Gunnbjörnsdottir M.I.
      • Omenaas E.
      • Gislason T.
      • Norrman E.
      • Olin A.C.
      • Jogi R.
      • et al.
      Obesity and nocturnal gastro-oesophageal reflux are related to onset of asthma and respiratory symptoms.

      Smoking

      The questions asked were: “Are you a smoker?” and “Are you an ex-smoker?” New categories of independent variables were created for smoking habits based on self-reported smoking status in 1990 and 2003. Subjects were divided into three groups: smokers, ex-smokers and never-smokers.

      Socioeconomic grouping

      A socioeconomic index was created using information on current occupation in 2003. On the basis of this, subjects were divided into five socioeconomic groups
      • Gunnbjörnsdottir M.I.
      • Franklin K.A.
      • Norbäck D.
      • Björnsson E.
      • Gislason D.
      • Lindberg E.
      • et al.
      Prevalence and incidence of respiratory symptoms in relation to indoor dampness: the RHINE study.
      : (1) managers and professionals (i.e. legislators, senior officials, managers and professionals); (2) other non-manual workers (i.e. technicians and associate professionals, clerks, service workers and market sales workers); (3) skilled manual workers (i.e. skilled agricultural and fishery workers and craft and related trades workers); (4) semi-skilled or unskilled manual workers (i.e. plant and machine operators and assemblers and elementary occupations); and (5) unclassifiable or unknown (i.e. housewife, student, not classifiable job, unemployed, not working because of poor health, and retired).

      Physical activity

      The level of physical activity during leisure time 2003 was categorized into three groups. An ordinary level of daily activities was defined as basic physical activity including cycling or walking to work. An intermediate level included regular physical activity like swimming, jogging, tennis, or aerobic exercise for at least 3 h a week. More vigorous activities on a weekly basis were defined as hard training several times a week. The categorization was adopted from a large population-based, prospective study on physical activity and mortality in women.
      • Camargo Jr., C.A.
      • Weiss S.T.
      • Zhang S.
      • Willett W.C.
      • Speizer F.E.
      Prospective study of body mass index, weight change, and risk of adult-onset asthma in women.

      Signs of building dampness in home

      Exposure to indoor dampness was considered if any of the following four damage types had been observed during the last 12 months in 2003: (1) water leakage or water damage indoors on walls, floor or ceilings; (2) bubbles or yellow discoloration on plastic floor covering or black discoloration of parquet floor; (3) visible mould growth indoors on walls, floor or ceilings; or (4) smell of mould indoors.
      • Norbäck D.
      • Björnsson E.
      • Janson C.
      • Palmgren U.
      • Boman G.
      Current asthma and biochemical signs of inflammation in relation to building dampness in dwellings.
      Subjects reporting one or more signs of dampness were referred to as reporting signs of indoor dampness in the last 12 months.

      Dietary habits

      There were seven questions in 2003 about current consumption of fish, meat, game meat, fruit, milk, fermented milk products and fast food such as hamburgers. There were five alternatives given for each type of food item, intended to measure the frequency of consumption. In the statistical calculations, a consumption score was used as follows: 0 = never consumed, 1 = less than once a week, 2 = once a week, 3 = more than once a week, 4 = daily consumption.
      • Takaoka M.
      • Norbäck D.
      Diet among Japanese female university students and asthmatic symptoms, infections, pollen and furry pet allergy.
      A summary score for fruit and fish was then calculated following the published recommendations from Svensson.
      • Svensson E.
      Construction of a single global scale for multi-item assessments of the same variable.

      Statistical methods

      Associations between asthma onset and the risk factors were estimated using logistic regression. To model the possible nonlinear relationship between BMI and asthma onset we used a restricted cubic spline function. The restricted cubic spline consists of piecewise polynomials of degree 3 joined together at specific points in the data called knots which are chosen by the analyst. We chose four knots placed at the 5th, 35th, 65th and 95th percentiles to obtain a smooth function that would not overfit the relationship. The spline function is linear beyond the 5th and 95th percentiles and has attractive properties which allow effect estimates and confidence intervals to be estimated along the smooth function. Another important feature is that the spline function is a function of the data and the knots only and thus unaffected by the response variable.
      • Norbäck D.
      • Björnsson E.
      • Janson C.
      • Palmgren U.
      • Boman G.
      Current asthma and biochemical signs of inflammation in relation to building dampness in dwellings.
      In order to minimize overfitting we limited ourselves to modelling no more than 53 degrees of freedom, following recommendations by Harrell.
      • Norbäck D.
      • Björnsson E.
      • Janson C.
      • Palmgren U.
      • Boman G.
      Current asthma and biochemical signs of inflammation in relation to building dampness in dwellings.
      Two models were fitted to assess effect modification by age group: one simple model with main effects only and one model where age group was allowed to interact with all other variables. The hypothesis of no effect modification was tested using a Wald-test. A p-value of 0.05 or less was considered statistically significant and evidence of at least one interaction term not equal to zero. All analyses were made using R v2.7.2
      • Takaoka M.
      • Norbäck D.
      Diet among Japanese female university students and asthmatic symptoms, infections, pollen and furry pet allergy.
      with functions from the Design library.
      • Svensson E.
      Construction of a single global scale for multi-item assessments of the same variable.

      Ethical approval

      The study was approved by Ethics Committee at the University of Umeå (§ 222, 1989-12-12) and the Ethics Committee, Faculty of Medicine, Uppsala (Dnr 01-313).

      Results

      In total 8150 of the eligible subjects in the remaining cohort (n = 11,282) answered the questionnaire, giving a response rate of 73%. Questionnaires with reported age and sex not consistent with data from the population register were excluded (n = 84). The study population was 8066 subjects, 2190 young adults, 3557 middle aged and 2319 elderly. The characteristics of the study population are given in Table 1.
      Table 1Characteristics of the study population.
      Age groupAll subjects (n = 8066)Young (n = 2190)Middle aged (n = 3557)Elderly (n = 2319)p-Value
      Females535552540.161
      Never smoker53744150<0.001
      Ex-smokers31143540
      Current smoker17122410
      Asthma heredity18161819<0.001
      Hay fever1314159<0.001
      BMI (SD) kg/m25 (4)24 (4)26 (4)25 (4)<0.001
      BMI ≥ 301181310<0.001
      Basic physical activity76617889<0.001
      Phys exercise >3 h/week19281910
      Hard physical training41031
      Nocturnal GOR > 1/week96815<0.001
      Snoring > 3/week20102521<0.001
      Damp dwelling81195<0.001
      Fruit less than once/week1116118<0.001
      Fruit daily50405254
      Fish less than once/week31492919<0.001
      Fish more than once/week85910
      Asthma 19906.5%6.56.76.10.465
      Asthma 200312.717.211.410.3<0.001
      Dr diagnosed asthma 19905.45.65.25.50.798
      Dr diagnosed asthma 200311.315.79.510.0<0.001
      Asthma medication 19905.86.35.45.70.211
      Asthma medication 200310.813.59.510.2<0.001
      Asthma symptoms 199011.710.211.313.4<0.001
      Asthma symptoms 200318.923.217.617.0<0.001
      Asthma (broad def)
      Defined as “yes” to one or more of the questions; “Do you have or have you had asthma?”, “Have you ever been diagnosed with asthma by a doctor?”, “Do you use asthma medication?” and “Do you have or have you had symptoms of asthma ?”
      1990
      12.811.312.514.5<0.001
      Asthma (broad def)
      Defined as “yes” to one or more of the questions; “Do you have or have you had asthma?”, “Have you ever been diagnosed with asthma by a doctor?”, “Do you use asthma medication?” and “Do you have or have you had symptoms of asthma ?”
      2003
      21.225.019.820.0<0.001
      Asthma onset10.714.79.39.0<0.001
      Unpaired t-test and Chi-squared test were used to test differences between age groups.
      a Defined as “yes” to one or more of the questions; “Do you have or have you had asthma?”, “Have you ever been diagnosed with asthma by a doctor?”, “Do you use asthma medication?” and “Do you have or have you had symptoms of asthma ?”
      There were 791 new cases of adult asthma onset. The prevalence of asthma in 2003 had increased in all age groups (Table 1), but this was expected since cumulative variables are measured. As presented in Table 2, female sex, young age, asthma heredity, hay fever, smoking, BMI, hard physical training, nocturnal GOR and habitual snoring were significant independent risk factors for onset asthma. High consumption of fruit and fish was protective. The cumulative incidence of asthma was not related to socioeconomic group (data not shown) and building dampness.
      Table 2Adjusted relative risks (RR) for onset of asthma.
      Onset of asthmap-Value
      Females1.45 (1.23–1.70)<0.001
      Young1.70 (1.41–2.04)<0.001
      Middle aged1
      Elderly1.07 (0.86–1.33)
      Asthma heredity1.47 (1.24–1.75)<0.001
      Hay fever2.34 (1.93–2.84)<0.001
      Non-smokers10.008
      Ex-smokers1.16 (0.96–1.40)
      Smokers1.37 (1.12–1.68)
      BMI per IQR1.49 (1.25–1.77)<0.001
      Basic physical activity10.005
      Training 3 times a week1.15 (0.94–1.39)
      Hard training several times a week1.60 (1.19–2.15)
      Nocturnal GOR > 1 week2.16 (1.72–2.72)<0.001
      Snoring > 3 week1.23 (1.01–1.49)0.038
      Building dampness in home1.10 (0.89–1.36)0.376
      Managers and professionals1.17 (0.89–1.54)0.501
      Other non-manual1
      Skilled manual0.91 (0.70–1.19)
      Semi-skilled or unskilled manual0.98 (0.81–1.19)
      Unclassifiable or unknown0.86 (0.65–1.15)
      Fish and fruit consumption0.78 (0.66–0.92)0.003
      GOR = gastro-oesophageal reflux. Data are presented as RR (95% CI) and are adjusted for all variables in the table.
      Next, the analyses were extended to adjustment for confounding and effect modification by age groups (young, middle aged and elderly). The impact of risk factors differed between the age groups (Table 3). BMI had a significantly higher impact in the middle aged and the elderly (p < 0.05). Also, nocturnal GOR had a significantly higher impact in the middle aged and the young adults compared to the elderly (p < 0.05). No significant difference in the impact of risk factors between men and women was found (data not shown).
      Table 3Adjusted relative risks (RR) separated age groups for onset of asthma including test for effect modification.
      Onset of asthmaTest for effect modification
      YoungMiddle agedElderlyp-Value
      Females1.56 (1.22–1.98)1.48 (1.11–1.97)1.24 (0.81–1.90)0.658
      Asthma heredity1.31 (0.97–1.76)1.56 (1.17–2.09)1.91 (1.23–2.98)0.388
      Hay fever2.65 (2.03–3.47)2.24 (1.62–3.09)1.50 (0.71–3.19)0.316
      Non-smokers1110.407
      Ex-smokers1.40 (1.05–1.88)0.94 (0.71–1.25)0.19 (0.81–1.73)
      Smokers1.54 (1.15–2.07)1.16 (0.85–1.59)1.22 (0.66–2.27)
      BMI per IQR1.22 (0.91–1.64)1.74 (1.32–2.29)1.79 (1.20–2.67)0.037
      Basic physical activity1110.564
      Training 3 times a week1.14 (0.96–1.34)1.33 (1.02–1.74)1.39 (0.87–2.21)
      Hard training several times a week1.29 (0.93–1.80)1.77 (1.03–3.04)1.93 (0.77–4.88)
      Nocturnal GOR > 1 week2.06 (1.47–2.90)2.86 (2.11–3.87)1.03 (0.56–1.90)0.013
      Snoring > 3 week1.35 (0.96–1.88)1.12 (0.85–1.49)1.62 (1.06–2.45)0.317
      Building dampness in home1.26 (0.94–1.69)1.01 (0.69–1.51)0.72 (0.07–6.94)0.606
      Managers and professionals1.01 (0.65–1.57)1.40 (0.96–2.03)1.05 (0.41–2.66)0.565
      Other non-manual111
      Skilled manual0.92 (0.60–1.39)0.90 (0.54–1.50)0.93 (0.47–1.83)
      Semi-skilled or unskilled manual1.13 (0.84–1.51)0.77 (0.54–1.09)1.20 (0.76–1.92)
      Unclassifiable or unknown0.84 (0.51–1.38)0.96 (0.55–1.65)0.94 (0.51–1.77)
      Fish and fruit consumption0.87 (0.69–1.10)0.78 (0.58–1.04)0.52 (0.35–0.77)0.106
      GOR = gastro-oesophageal reflux. Data are presented as RR (95% CI) and are adjusted for all variables in the table.
      The annual incidence of asthma onset per 1000 person years related to BMI is shown in Fig. 1 for the population at risk separated for age group. We found a higher cumulative incidence of asthma in both the group with the lowest BMI and, in particular, the obese category. The BMI range in the figure is limited to BMIs between the 5th and 95th percentiles, BMI 21 and 32, because of small numbers in the extremes. A U-shaped association was found for the middle aged and elderly, but in the young adult the cumulative incidence of asthma increased linearly with increasing BMI (Fig. 1). The association between asthma risk and low BMI persisted only in current smokers and ex-smokers. In the never-smokers there was a more linear relationship between BMI and the risk for asthma, with a higher cumulative incidence of asthma with increasing BMI.
      Figure thumbnail gr1
      Figure 1Effect of BMI by age categories on the annual incidence of asthma onset per 1000 person years adjusted for sex, heredity, hay fever, smoking, gastro-oesophageal reflux, snoring, physical activity, building dampness, socioeconomic group, and fish and fruit consumption. The x-axis is truncated at the 5th and 95th percentiles of the BMI distribution due to the sparse amount of data in the tails of the spline functions.
      The annual incidence of asthma onset per 1000 person years related to fish and fruit score is presented in Fig. 2, which shows a linear decrease of asthma onset with increasing consumption of fish and fruit. The impact of fish and fruit consumption differed between the age groups with the strongest protective effect in the elderly high consumers of fish and fruit. The effect was significantly weaker in the young adults.
      Figure thumbnail gr2
      Figure 2Effect of fruit and fish consumption score by age categories on the annual incidence of asthma onset per 1000 person years adjusted for sex, heredity, hay fever, smoking, gastro-oesophageal reflux, snoring, physical activity, building dampness, socioeconomic group, and BMI.

      Discussion

      The results of this 13-year follow-up study of three age groups in Sweden add evidence to an independent relationship between several risk factors and the cumulative incidence of asthma in adults illustrating the multi-factorial genesis of asthma. The main finding is that lifestyle factors such as smoking, BMI, physical training and low consumption of fruit and fish are significant independent risk factors for the cumulative incidence of asthma after adjusting for socioeconomic group. The impact of BMI and GOR differed significantly between age groups. BMI had a greater impact in the middle aged and the elderly, whilst GOR had an impact in the middle aged and the young adults. No significant difference was found in the impact of risk factors between men and women. As previously reported, family history of asthma, hay fever, female sex and young age were strong risk factors in all age groups.
      The three age groups studied could roughly be categorized at baseline in 1990 as teenagers, parents and grandparents. At follow up, they had turned into 29-year-old young adults, middle aged 43–52 years old, and elderly, 73–82 years old. The asthma prevalence in 2003 had increased in all age groups, most strikingly in the young adults independently of how the asthma outcome was measured (Table 2). This is expected since cumulative variables are used, but the prevalence of the current use of asthma medication – a non-cumulative variable – had increased in the same amount. In the youngest age group, prevalence of doctor diagnosed asthma was high, 16%, about the same as in Swedish teenagers 10 years ago.
      • Vesterinen E.
      • Kaprio J.
      • Koskenvuo M.
      Prospective study of asthma in relation to smoking habits among 14,729 adults.
      In another Swedish study of conscripts, the prevalence of asthma had only increased slightly.
      • Bråbäck L.
      • Hjern A.
      • Rasmussen F.
      Body mass index, asthma and allergic rhinoconjunctivitis in Swedish conscripts – a national cohort study over three decades.
      However, asthma was diagnosed after an examination of a physician, but the prevalence of allergic rhinitis had doubled in a 10-year period. Still, the prevalence of asthma and respiratory had levelled off in a recent study from Western Sweden.
      • Lotvall J.
      • Ekerljung L.
      • Rönmark E.P.
      • Wennergren G.
      • Linden A.
      • Rönmark E.
      • et al.
      West Sweden asthma study: prevalence trends over the last 18 years argues no recent increase in asthma.
      Smoking habits have changed considerably in one generation. It is worth noting that 74% of the young adults in this study, parents or parents-to-be, are never-smokers, compared with only 41% in the middle aged who had left the period of bringing up children behind them. Our results support the hypothesis that smoking is a risk factor for asthma onset in adult age. Studies on the effect on smoking on asthma are however still controversial: some demonstrate more asthma with smoking,
      • Piipari R.
      • Jaakkola J.J.
      • Jaakkola N.
      • Jaakkola M.S.
      Smoking and asthma in adults.
      • Plaschke P.P.
      • Janson C.
      • Norrman E.
      • Björnsson E.
      • Ellbjär S.
      • Järvholm B.
      Onset and remission of allergic rhinitis and asthma and the relationship with atopic sensitization and smoking.
      whilst others have failed to find an increased risk.
      • Troisi R.J.
      • Speizer F.E.
      • Rosner B.
      • Trichopoulos D.
      • Willett W.C.
      Cigarette smoking and incidence of chronic bronchitis and asthma in women.
      • Vesterinen E.
      • Kaprio J.
      • Koskenvuo M.
      Prospective study of asthma in relation to smoking habits among 14,729 adults.
      An increased remission rate of asthma was observed among subjects who quit smoking during a 10-year observation period
      • Holm M.
      • Ome naas E.
      • Gislason T.
      • Svanes C.
      • Jögi R.
      • Norrman E.
      • et al.
      Remission of asthma: a prospective longitudinal study from northern Europe (RHINE study).
      and recent intervention studies with smoking cessation demonstrated improvement of asthma.
      • Tönnesen P.
      • Pisinger C.
      • Hvidberg S.
      • Wennike P.
      • Bremann L.
      • Westin A.
      • et al.
      Effects of smoking cessation and reduction in asthmatics.
      Our study adds evidence to an independent relationship between increased BMI and onset of adult asthma. On the physiological level, adipose tissue might be actively involved in inflammatory processes and there might be a causal link between obesity and chronic inflammatory airway disease.
      • Loerbroks A.
      • Apfelbacher C.J.
      • Amelang M.
      • Sturmer T.
      Obesity and adult asthma: potential effect modification by gender, but not by hay fever.
      A higher cumulative incidence of asthma with increasing BMI was found in all age groups and in both men and women. In other studies, the association between increasing BMI and risk of asthma has been most consistent in the women.
      • Loerbroks A.
      • Apfelbacher C.J.
      • Amelang M.
      • Sturmer T.
      Obesity and adult asthma: potential effect modification by gender, but not by hay fever.
      • Beuther D.A.
      • Sutherland E.R.
      Overweight, obesity, and incident asthma: a meta-analysis of prospective epidemiologic studies.
      But, previous studies have also linked underweight to a higher risk for developing asthma.
      • Luder E.
      • Ehrlich R.I.
      • Lou W.Y.
      • Melnik T.A.
      • Kattan M.
      Body mass index and the risk of asthma in adults.
      In the present study, the relationship between the cumulative incidence of asthma and BMI was U-shaped for the men in the present study, but not for the women. An analysis restricted to the non-smokers showed a linear relationship between BMI and the cumulative incidence of asthma. The high cumulative incidence of asthma in the subjects with lowest BMI could partly be explained by that some subjects with low BMI had been misclassified with asthma rather than COPD and/or emphysema. This is supported by the fact that men in the two oldest age groups had been significantly heavier smokers than the women. The opposite was found in the young adults, but the young women had not smoked long enough to get COPD.
      In a recent review
      • Eneli I.U.
      • Skybo T.
      • Camargo Jr., C.A.
      Weight loss and asthma: a systematic review.
      the authors also found a fairly consistent association between weight loss and improved asthma, which is an important epidemiological criterion of causality. This is indeed an interesting finding in the light of the present study. However, it also highlights one practical issue of causality in epidemiology: the induction of disease versus provocation of symptoms. It is possible that overweight causes asthma, but overweight might also provoke more symptoms in mild asthmatics. The same applies to some extent to smoking, GOR, physical exercise, housing conditions and work exposures.
      An independent relationship between obesity, nocturnal gastro-oesophageal reflux and habitual snoring and the onset of asthma was reported in a 5–10 years follow-up study by the ECRHS.
      • Gunnbjörnsdottir M.I.
      • Omenaas E.
      • Gislason T.
      • Norrman E.
      • Olin A.C.
      • Jogi R.
      • et al.
      Obesity and nocturnal gastro-oesophageal reflux are related to onset of asthma and respiratory symptoms.
      In the previously mentioned review
      • Eneli I.U.
      • Skybo T.
      • Camargo Jr., C.A.
      Weight loss and asthma: a systematic review.
      the authors commented on GOR as a condition which may mediate or confound an association between weight loss and change in asthma symptoms.
      Some aspects of diet have been recently recognized as potential risk factors for asthma, but the evidence is conflicting. Our study was not designed to address dietary factors in detail, but seven questions on dietary intake were included. After controlling for possible confounders, the cumulative incidence of asthma was significantly increased for subjects with low consumption of both fruit and fish. Other dietary factors did not affect the risk for asthma (data not shown). Woods and co-workers have previously reported apples and pears to be a protective factor for current asthma.
      • Woods R.K.
      • Walters E.H.
      • Raven J.M.
      • Wolfe R.
      • Ireland P.D.
      • Thien F.C.
      • et al.
      Food and nutrient intakes and asthma risk in young adults.
      Romieu and co-workers investigated whether fruit and vegetable intakes predicted the prevalence of adult asthma among French women participating in the E3N study.
      • Romieu I.
      • Varraso R.
      • Avenel V.
      • Leynaert B.
      • Kauffmann F.
      • Clavel-Chapelon F.
      Fruit and vegetable intakes and asthma in the E3N study.
      They concluded that some fruit and vegetables may decrease the prevalence of adult asthma. Antioxidants have been suggested to be of importance, but results from antioxidant/vitamin supplementation have been disappointing.
      • Kaur B.
      • Rowe B.H.
      • Arnold E.
      Vitamin C supplementation for asthma.
      Studying dietary patterns instead of specific foods and nutrients was a new approach in the French E3N study.
      • Varraso R.
      • Kauffmann F.
      • Leynaert B.
      • Le Moual N.
      • Boutron-Ruault M.C.
      • Clavel-Chapelon F.
      • et al.
      Dietary patterns and asthma in the E3N study.
      An overall Western diet was associated with an increased risk of frequent asthma attacks, while a nut and wine diet seemed to be protective. In the present study, we used a fruit and fish index because the intake of these dietary factors was correlated. A high intake of fruit and fish may represent a dietary pattern in Sweden in the same way as a nut and wine diet in France.
      Physical training in asthmatics is beneficial as it improves cardiopulmonary fitness. Exercise-induced asthma and other airway symptoms are, however, a common problem among athletes.
      • Helenius I.
      • Lumme A.
      • Haahtela T.
      Asthma, airway inflammation and treatment in elite athletes.
      Airway cooling and airway drying during physical exercise have been investigated, but the mechanisms are still unclear.
      • Anderson S.D.
      • Kippelen P.
      Airway injury as a mechanism for exercise-induced bronchoconstriction in elite athletes.
      In our study hard physical training increased the risk of asthma. In an earlier study of cross country skier, in the same region, asthma and bronchial hyperresponsiveness were much more common in cross country skiers than in the general population and non-skiers.
      • Larsson K.
      • Ohlsen P.
      • Larsson L.
      • Malmberg P.
      • Rydström P.O.
      • Ulriksen H.
      High prevalence of asthma in cross country skiers.
      Strenuous exercise at low temperatures entailing breathing large volumes of cold air was suggested as the probable explanation.
      Asthma is a syndrome with common features that arise from very different pathways. The impact of BMI was higher in the middle aged and elderly when the impact of risk factors between the three age groups in this study was compared. For diet, a trend was seen for a more asthma protective effect with increased age (p = 0.06). The impact of gastro-oesophageal reflux was higher in the young and the middle aged. These findings support that asthma phenotypes differ over age groups as outlined in a recent review discussing phenotypes of early/childhood asthma and late/adult onset asthma.
      • Wenzel S.E.
      Asthma: defining of the persistent adult phenotypes.
      There are few longitudinal studies of risk factors and onset of adult asthma. At follow up in ECRHS II, 8 years on the average had passed, and data was collected from 15 countries with different lifestyle and diagnostic traditions. The present study investigated a limited geographical area with a more homogenous lifestyle. The strength of this study is also that three different age groups were followed up after 13 years. The detailed information of the respondents' entire working life was a solid base for socioeconomic grouping. Smoking habits were reported both at baseline and follow up, but other lifestyle factors were only reported on follow up and not available at baseline. This is a weakness, as lifestyle may have changed during the observation period. Also, the associations found may reflect both cause and consequence of the asthma disease.
      Another methodological problem is that the cumulative incidence of asthma is based on self-reported data with no objective measurements. The questions used in this study have however been evaluated in other studies and have been found to have high specificity for asthma.
      • Toren K.
      • Brisman J.
      • Järvholm B.
      Asthma and asthma-like symptoms in adults assessed by questionnaires. A literature review.
      Since the primary purpose of the present study was to investigate risk factors for asthma development not prevalence, we decided to use a “broad definition” of asthma in order to get power. However, then testing different asthma outcomes in the model, kind of the same risk factors came up using different asthma definitions. Another problem is using asthma questions that have not been validated in elderly. There are few previous studies of the cumulative incidence of asthma in the elderly. It is a problem using asthma questions that have not been validated in elderly. Self-reported asthma in this age group could represent e.g. heart disease or COPD, but smoking was not a risk factor for asthma in this age group, which contradicted misclassification.
      In conclusion, this study adds evidence to an independent relationship between certain lifestyle factors and the cumulative incidence of asthma in adults. Weight loss, reduced smoking and a diet rich in fruit and fish may be of importance to prevent the onset of adult asthma. Randomized clinical trials with weight loss, smoking cessation and dietary regimes in established asthma cohorts are needed to be able to identify evidence-based preventive measures.

      Acknowledgements

      We acknowledge the work with the data set performed by Mats Uddenfeldt and Mathias Rask-Andersen. This study was supported financially by the Swedish Heart and Lung Foundation, the Swedish Association against Asthma and Allergy, the Centre for Clinical Research, Uppsala University/County Council of Gävleborg, and the University Hospital in Uppsala.

      Conflict of interest statement

      No possible conflicts of interest (e.g. funding sources for consultancies or studies of products) exist in this study.

      References

        • Beasley R.
        • Crane J.
        • Lai C.K.
        • Pearce N.
        Prevalence and etiology of asthma.
        J Allergy Clin Immunol. 2000 Feb; 105: S466-S472
        • Åberg N.
        Asthma and allergic rhinitis in Swedish conscripts.
        Clin Exp Allergy. 1989 Jan; 19: 59-63
        • Stjernberg N.
        • Eklund A.
        • Nyström L.
        • Rosenhall L.
        • Emmelin A.
        • Strömqvist L.H.
        Prevalence of bronchial asthma and chronic bronchitis in a community in northern Sweden; relation to environmental and occupational exposure to sulphur dioxide.
        Eur J Respir Dis. 1985 Jul; 67: 41-49
        • Lundbäck B.
        • Nyström L.
        • Rosenhall L.
        • Stjernberg N.
        Obstructive lung disease in northern Sweden: respiratory symptoms assessed in a postal survey.
        Eur Respir J. 1991 Mar; 4: 257-266
        • Ekerljung L.
        • Rönmark E.
        • Larsson K.
        • Sundblad B.M.
        • Bjerg A.
        • Ahlstedt S.
        • et al.
        No further increase of incidence of asthma: incidence, remission and relapse of adult asthma in Sweden.
        Respir Med. 2008 Dec; 102: 1730-1736
        • Almqvist C.
        • Pershagen G.
        • Wickman M.
        Low socioeconomic status as a risk factor for asthma, rhinitis and sensitization at 4 years in a birth cohort.
        Clin Exp Allergy. 2005 May; 35: 612-618
        • Sunyer J.
        • Anto J.M.
        • Tobias A.
        • Burney P.
        Generational increase of self-reported first attack of asthma in fifteen industrialized countries. European Community Respiratory Health Study (ECRHS).
        Eur Respir J. 1999 Oct; 14: 885-891
        • Toren K.
        • Hermansson B.A.
        Incidence rate of adult-onset asthma in relation to age, sex, atopy and smoking: a Swedish population-based study of 15,813 adults.
        Int J Tuberc Lung Dis. 1999 Mar; 3: 192-197
        • Piipari R.
        • Jaakkola J.J.
        • Jaakkola N.
        • Jaakkola M.S.
        Smoking and asthma in adults.
        Eur Respir J. 2004 Nov; 24: 734-739
        • Kogevinas M.
        • Anto J.M.
        • Sunyer J.
        • Tobias A.
        • Kromhout H.
        • Burney P.
        Occupational asthma in Europe and other industrialised areas: a population-based study. European community respiratory health survey study group.
        Lancet. 1999 May 22; 353: 1750-1754
        • Kogevinas M.
        • Zock J.P.
        • Jarvis D.
        • Kromhout H.
        • Lillienberg L.
        • Plana E.
        • et al.
        Exposure to substances in the workplace and new-onset asthma: an international prospective population-based study (ECRHS-II).
        Lancet. 2007 Jul 28; 370: 336-341
        • Camargo Jr., C.A.
        • Weiss S.T.
        • Zhang S.
        • Willett W.C.
        • Speizer F.E.
        Prospective study of body mass index, weight change, and risk of adult-onset asthma in women.
        Arch Intern Med. 1999 Nov 22; 159: 2582-2588
        • Shaheen S.O.
        • Sterne J.A.
        • Montgomery S.M.
        • Azima H.
        Birth weight, body mass index and asthma in young adults.
        Thorax. 1999 May; 54: 396-402
        • Gunnbjörnsdottir M.I.
        • Omenaas E.
        • Gislason T.
        • Norrman E.
        • Olin A.C.
        • Jogi R.
        • et al.
        Obesity and nocturnal gastro-oesophageal reflux are related to onset of asthma and respiratory symptoms.
        Eur Respir J. 2004 Jul; 24: 116-121
        • Ruigomez A.
        • Rodriguez L.A.
        • Wallander M.A.
        • Johansson S.
        • Thomas M.
        • Price D.
        Gastroesophageal reflux disease and asthma: a longitudinal study in UK general practice.
        Chest. 2005 Jul; 128: 85-93
        • Chinn S.
        • Jarvis D.
        • Burney P.
        • Luczynska C.
        • Ackermann-Liebrich U.
        • Anto J.M.
        • et al.
        Increase in diagnosed asthma but not in symptoms in the European community respiratory health survey.
        Thorax. 2004 Aug; 59: 646-651
        • Larsson L.
        • Boethius G.
        • Uddenfeldt M.
        Differences in utilization of asthma drugs between two neighbouring Swedish provinces: relation to symptom reporting.
        Eur Respir J. 1993 Feb; 6: 198-203
        • Gunnbjörnsdottir M.I.
        • Franklin K.A.
        • Norbäck D.
        • Björnsson E.
        • Gislason D.
        • Lindberg E.
        • et al.
        Prevalence and incidence of respiratory symptoms in relation to indoor dampness: the RHINE study.
        Thorax. 2006 Mar; 61: 221-225
        • Norbäck D.
        • Björnsson E.
        • Janson C.
        • Palmgren U.
        • Boman G.
        Current asthma and biochemical signs of inflammation in relation to building dampness in dwellings.
        Int J Tuberc Lung Dis. 1999 May; 3: 368-376
        • Takaoka M.
        • Norbäck D.
        Diet among Japanese female university students and asthmatic symptoms, infections, pollen and furry pet allergy.
        Respir Med. 2008 Jul; 102: 1045-1054
        • Svensson E.
        Construction of a single global scale for multi-item assessments of the same variable.
        Stat Med. 2001 Dec 30; 20: 3831-3846
        • Bråbäck L.
        • Hjern A.
        • Rasmussen F.
        Body mass index, asthma and allergic rhinoconjunctivitis in Swedish conscripts – a national cohort study over three decades.
        Respir Med. 2005 Aug; 99: 1010-1014
        • Lotvall J.
        • Ekerljung L.
        • Rönmark E.P.
        • Wennergren G.
        • Linden A.
        • Rönmark E.
        • et al.
        West Sweden asthma study: prevalence trends over the last 18 years argues no recent increase in asthma.
        Respir Res. 2009; 10: 94
        • Plaschke P.P.
        • Janson C.
        • Norrman E.
        • Björnsson E.
        • Ellbjär S.
        • Järvholm B.
        Onset and remission of allergic rhinitis and asthma and the relationship with atopic sensitization and smoking.
        Am J Respir Crit Care Med. 2000 Sep; 162: 920-924
        • Troisi R.J.
        • Speizer F.E.
        • Rosner B.
        • Trichopoulos D.
        • Willett W.C.
        Cigarette smoking and incidence of chronic bronchitis and asthma in women.
        Chest. 1995 Dec; 108: 1557-1561
        • Vesterinen E.
        • Kaprio J.
        • Koskenvuo M.
        Prospective study of asthma in relation to smoking habits among 14,729 adults.
        Thorax. 1988 Jul; 43: 534-539
        • Holm M.
        • Ome naas E.
        • Gislason T.
        • Svanes C.
        • Jögi R.
        • Norrman E.
        • et al.
        Remission of asthma: a prospective longitudinal study from northern Europe (RHINE study).
        Eur Respir J. 2007 Jul; 30: 62-65
        • Tönnesen P.
        • Pisinger C.
        • Hvidberg S.
        • Wennike P.
        • Bremann L.
        • Westin A.
        • et al.
        Effects of smoking cessation and reduction in asthmatics.
        Nicotine Tob Res. 2005 Feb; 7: 139-148
        • Loerbroks A.
        • Apfelbacher C.J.
        • Amelang M.
        • Sturmer T.
        Obesity and adult asthma: potential effect modification by gender, but not by hay fever.
        Ann Epidemiol. 2008 Apr; 18: 283-289
        • Beuther D.A.
        • Sutherland E.R.
        Overweight, obesity, and incident asthma: a meta-analysis of prospective epidemiologic studies.
        Am J Respir Crit Care Med. 2007 Apr 1; 175: 661-666
        • Luder E.
        • Ehrlich R.I.
        • Lou W.Y.
        • Melnik T.A.
        • Kattan M.
        Body mass index and the risk of asthma in adults.
        Respir Med. 2004; 98: 29-37
        • Eneli I.U.
        • Skybo T.
        • Camargo Jr., C.A.
        Weight loss and asthma: a systematic review.
        Thorax. 2008 Aug; 63: 671-676
        • Woods R.K.
        • Walters E.H.
        • Raven J.M.
        • Wolfe R.
        • Ireland P.D.
        • Thien F.C.
        • et al.
        Food and nutrient intakes and asthma risk in young adults.
        Am J Clin Nutr. 2003 Sep; 78: 414-421
        • Romieu I.
        • Varraso R.
        • Avenel V.
        • Leynaert B.
        • Kauffmann F.
        • Clavel-Chapelon F.
        Fruit and vegetable intakes and asthma in the E3N study.
        Thorax. 2006 Mar; 61: 209-215
        • Kaur B.
        • Rowe B.H.
        • Arnold E.
        Vitamin C supplementation for asthma.
        Cochrane Database Syst Rev. 2009; 1: CD000993
        • Varraso R.
        • Kauffmann F.
        • Leynaert B.
        • Le Moual N.
        • Boutron-Ruault M.C.
        • Clavel-Chapelon F.
        • et al.
        Dietary patterns and asthma in the E3N study.
        Eur Respir J. 2008 Oct 1;
        • Helenius I.
        • Lumme A.
        • Haahtela T.
        Asthma, airway inflammation and treatment in elite athletes.
        Sports Med. 2005; 35: 565-574
        • Anderson S.D.
        • Kippelen P.
        Airway injury as a mechanism for exercise-induced bronchoconstriction in elite athletes.
        J Allergy Clin Immunol. 2008 Aug; 122 (quiz 36–37): 225-235
        • Larsson K.
        • Ohlsen P.
        • Larsson L.
        • Malmberg P.
        • Rydström P.O.
        • Ulriksen H.
        High prevalence of asthma in cross country skiers.
        BMJ. 1993 Nov 20; 307: 1326-1329
        • Wenzel S.E.
        Asthma: defining of the persistent adult phenotypes.
        Lancet. 2006 Aug 26; 368: 804-813
        • Toren K.
        • Brisman J.
        • Järvholm B.
        Asthma and asthma-like symptoms in adults assessed by questionnaires. A literature review.
        Chest. 1993 Aug; 104: 600-608