Obstructive and restrictive lung disease and functional limitation: data from the Third National Health and Nutrition Examination

Authors

  • D. M. Mannino,

    1. From the Air Pollution and Respiratory Health Branch, Division of Environmental Hazards and Health Effects, National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, GA, USA
    Search for more papers by this author
  • E. S. Ford,

    1. From the Air Pollution and Respiratory Health Branch, Division of Environmental Hazards and Health Effects, National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, GA, USA
    Search for more papers by this author
  • S. C. Redd

    1. From the Air Pollution and Respiratory Health Branch, Division of Environmental Hazards and Health Effects, National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, GA, USA
    Search for more papers by this author

David M. Mannino, MD, National Center for Environmental Health, Centers for Disease Control and Prevention 1600 Clifton Road, MS E-17 Atlanta, GA 30333, USA (fax: +1 404-498-1088; e-mail: dmannino@cdc.gov).

Abstract.

Objective.  To determine functional limitations in adults with obstructive or restrictive lung disease or respiratory symptoms.

Design.  Cross-sectional study.

Subjects.  Adult participants in phase 2 of the Third National Health and Nutrition Examination Survey, 1991–94.

Methods.  We classified subjects using spirometric criteria into the following mutually exclusive categories using the forced expiratory volume in 1 s (FEV1), the forced vital capacity (FVC), the FEV1/FVC ratio and the presence of respiratory symptoms: severe obstruction, moderate obstruction, mild obstruction, respiratory symptoms only, restrictive lung disease and no lung disease. We developed regression models to predict functional limitations (unable to walk a quarter of a mile, unable to lift 10 pounds, needs help with daily activities) that controlled for age, race, sex, education, smoking status, body mass index and comorbid conditions.

Results.  Severe and moderate obstruction were associated with an increased risk of being unable to walk a quarter of a mile [odds ratio (OR) 8.4, 95% confidence interval (CI) 3.6, 19.9 and OR 2.4, 95% CI 1.4, 4.0]. Restrictive lung disease and the presence of respiratory symptoms in the absence of lung function impairment were also associated with an increased risk of this outcome (OR 2.8, 95% CI 1.4, 5.6 and OR 2.8, 95% CI 2.0, 3.9). Similar results were obtained for the outcomes of being unable to lift 10 pounds or needing help with daily activities.

Conclusions.  The presence of obstructive or restrictive lung disease, or respiratory symptoms in the absence of lung function impairment is associated with increased functional impairment.

Introduction

Both chronic obstructive pulmonary disease (COPD) and restrictive lung diseases are important causes of morbidity and mortality in the US [1, 2]. Morbidity from chronic respiratory disease includes not only hospitalizations and emergency department visits, but also restricted activity and functional limitations [1].

Both diagnosed and undiagnosed obstructive lung disease are known to be associated with increased functional limitation [3]. In a recently published analysis of data from the Third National Health and Nutrition Examination Survey (NHANES III), Coultas et al. found that undiagnosed airflow obstruction was associated with impaired health and functional status. However, that analysis did not look at individuals with restrictive lung disease or respiratory symptoms, and these subjects were included in the referent group that did not have obstructive lung disease. Restrictive lung disease can have a variety of aetiologies, including diseases such as muscular weakness, congestive heart failure, interstitial lung disease, diabetes mellitus and obesity, all of which can also be associated with increased limitation and morbidity [4–8]. In addition, respiratory symptoms such as chronic cough or sputum production may have their own significant morbidities [9, 10].

Spirometric criteria were applied to define obstructive and restrictive lung disease in adults who had pulmonary function measurements obtained as part of phase 2 of NHANES III. In addition, a group of subjects who reported respiratory symptoms in the absence of lung function impairment were assessed. Then the relationship between impaired lung function and respiratory symptoms with reported functional limitations and self-reported health status, controlling for potential covariates that may also influence these measures, was assessed.

Methods

Study population

The NHANES III was conducted in two nationally representative phases from 1988 to 1994 by the National Center for Health Statistics of the Centers for Disease Control and Prevention, Atlanta, GA [11]. In this study a stratified multistage clustered probability design was used to select a representative sample of the US population, yielding results which can be extrapolated to the noninstitutionalized civilian US population. Study participants completed extensive questionnaires in the household and a comprehensive physical examination, including pulmonary function testing, either in the household or at a specially equipped mobile examination centre. The study was approved by the National Center for Health Statistics’ Institutional Review Board.

Subjects and demographics

The study sample was limited to adults aged 17 years and older who participated in phase 2 of NHANES III (1991–94) and who had pulmonary function testing performed in either the household or the mobile examination centre, and had complete data on their race, smoking status, body mass index, functional status and presence of respiratory symptoms. Functional status was not available for all adults in phase 1 of NHANES III, so the analysis was limited to phase 2 participants. Of the 9930 adult survey participants in phase 2 of NHANES III, 802 did not undergo medical examination, 687 did not undergo pulmonary function testing, 155 were pregnant, 659 were missing data on function limitations and 124 were missing data on other variables. After the exclusions, data from 7503 (75.6%) subjects were available for the main analysis.

Variable definition

The race of the participants was classified as ‘White’, ‘African-American’, ‘Mexican-American’ or ‘Other’ and was determined by self-report on the questionnaire. Other demographic covariates included sex, education (≤11, 12 or ≥13 years) and age. The subjects were stratified into four age strata: 17–44, 45–64, 65–74 and 75 years and older. The body mass index was calculated by dividing the weight in kilograms by the square of the height in metres and the following classifications were used: <18.5, ≥18.5–24, 25–29 and ≥30 kg m−2. The subjects were defined as being current smokers, former smokers, pipe or cigar smokers, or never smokers, based on their responses to series of questions. Current pipe or cigar smokers were considered ‘current smokers’. One had to have smoked at least 100 cigarettes, 50 cigars, or 3 ounces of pipe tobacco to qualify as a former or current smoker.

The subjects were classified as having a respiratory symptom when they gave a positive response to any one of the following questions involving specific symptoms (cough, phlegm, wheeze and dyspnoea): ‘Do you usually cough on most days for 3 consecutive months or more during the year?’; ‘Do you bring up phlegm on most days for 3 consecutive months or more during the year?’; ‘Have you had wheezing or whistling in your chest at any time in the past 12 months?’ and ‘Are you troubled by shortness of breath when hurrying on level ground or walking up a slight hill?’ The subjects were classified as having or not having: cardiovascular disease (positive response to physician diagnosed myocardial infarction, stroke or congestive heart failure); diabetes mellitus (positive response to physician diagnosis of diabetes mellitus) or arthritis (positive response to physician diagnosis of osteoarthritis or rheumatoid arthritis).

Pulmonary function data

Using either a dry rolling seal spirometer in the mobile examination centre or a portable spirometer in the household examination, spirometry was conducted on the examinees. Procedures for testing were based on the 1987 American Thoracic Society recommendations [12]. In order to obtain curves acceptable according to the protocol, examinees performed five to eight forced expirations. Subjects were excluded from this analysis if they either did not perform spirometry or had results that were not reliable. Values used in this analysis included the forced vital capacity (FVC), the forced expiratory volume in 1 s (FEV1) and the FEV1/FVC ratio. The predicted values of FEV1 and FVC were determined using previously published prediction equations [13]. The sex-specific model for white participants was applied to people of ‘Other’ race. The subjects were classifed using a modification of the global initiative for chronic obstructive lung disease (GOLD) criteria for COPD into the following mutually exclusive categories using the FEV1, the FVC, the FEV1/FVC ratio, and the presence of respiratory symptoms: severe COPD (FEV1/FVC < 0.70 and FEV1 < 50% predicted), moderate COPD (FEV1/FVC < 0.70 and FEV1 ≥ 50 to <80% predicted), mild COPD (FEV1/FVC < 0.70 and FEV1 ≥ 80%), symptoms only (presence of respiratory symptoms in the absence of any lung function abnormality), restrictive lung disease (FEV1/FVC ≥ 0.70 and FVC < 80% predicted) and no lung disease [14].

Functional limitation and reported health status

Functional limitations were determined based on self-response to questions asking whether they had ‘no difficulty, some difficulty, much difficulty or are unable to do certain activities’. Limitations included walking for a quarter of a mile and lifting or carrying something as heavy as 10 pounds and considered any reported difficulty as a limitation. Subjects were also asked ‘Because of any impairment or health problem, do you need the help of other persons in handling routine needs, such as everyday household chores, doing necessary business, shopping or getting around for other purposes?’ and a positive response was considered as a limitation. Subjects categorized their general health status as ‘excellent, very good, good, fair, or poor’ and we stratified the responses into fair or poor versus good or better.

Analysis

All estimates were calculated using the sampling weight to represent adults aged 17 years and older in the US. The purpose of these sampling weights was to adjust for unequal probabilities of selection and to account for nonresponse. All percentages shown are weighted and age-adjusted to the distribution of participants in the final analytic sample.

Univariate and multivariate logistic regression models were developed to describe the relationship between impaired lung function and respiratory symptoms with functional limitation or health status. Multivariate models adjusted for age, sex, race, smoking status, body mass index and comorbid chronic conditions including cardiovascular disease, models were evaluated for interaction by the covariates. Both SAS and SUDAAN, a program that adjusts for the complex sample design when calculating variance estimates [15, 16], were used in this analysis.

Results

The final data set comprised 7503 subjects representing an estimated 165 million adults in the US. The 2427 adults not included in the final analytic sample because of missing data were older than the participants (mean age 48.8 years vs. 42.2 years, P < 0.05) and more likely to be female (55.0% vs. 51.5%, P < 0.05) or former smokers (28.3% vs. 23.4%, P < 0.05).

The pulmonary function criteria classified 1.3% of subjects as having severe COPD, 5.1% as having moderate COPD, 6.8% as having mild COPD, 6.6% as having restrictive lung disease, and 23.4% having respiratory symptoms in the absence of lung function impairment (Table 1). The proportion of subjects with moderate or severe COPD was higher in older age groups and current or former smokers (Table 1). The proportion of subjects with restrictive lung disease was higher in older age groups and people with diabetes mellitus, but not in current or former smokers (Table 1). The proportion of subjects with moderate or severe COPD in this analysis (6.4%) was similar to the proportion (6.8%) found in a previous analysis of the larger NHANES II phase 1 and 2 data sets [17].

Table 1.  Participants stratified by age, race, sex, education level, smoking status, body mass index, cardiovascular disease, diabetes mellitus, and arthritis and the age-adjusted, weighted proportion (standard error in parenthesis) within each subgroup with each class of chronic obstructive pulmonary disease (COPD) or restrictive lung disease. From the Third National Health and Nutrition Examination Survey, phase 2, 1991–94
 n Severe COPDaModerate COPDMild COPDRespiratory symptomsRestrictive diseaseNormal
  1. aSevere COPD – forced expiratory volume in 1 s (FEV1)/the forced vital capacity (FVC) < 0.70 and FEV1 < 50% predicted; moderate COPD – FEV1/FVC < 0.70 and FEV1 ≥ 50 to <80% predicted; mild COPD – FEV1/FVC < 0.70 and FEV1 ≥ 80%; respiratory symptoms – presence of respiratory symptoms in the absence of any lung function abnormality; restrictive lung disease – FEV1/FVC ≥ 0.70 and FVC < 80% predicted.

Age (years)
  17–4443480.0 (0.0)1.6 (0.4)3.5 (0.4)24.5 (0.9)5.2 (0.7)65.1 (1.2)
  45–6417192.6 (0.6)7.6 (1.0)8.8 (1.1)22.0 (1.6)8.0 (1.3)50.9 (2.3)
  65–748474.7 (1.5)15.8 (2.3)17.6 (1.8)20.0 (2.0)9.6 (1.6)32.3 (2.0)
 ≥755893.2 (1.0)15.3 (1.8)17.1 (1.5)22.5 (1.4)11.1 (1.7)30.9 (2.1)
Race
 White28431.3 (0.2)5.1 (0.4)6.8 (0.5)23.4 (0.7)6.6 (0.6)56.9 (1.0)
 African-American22971.4 (0.3)5.1 (0.5)7.3 (0.5)23.4 (0.9)5.6 (0.6)57.2 (1.4)
 Mexican-American20140.8 (0.2)5.0 (0.6)5.3 (0.7)23.2 (0.9)8.9 (0.9)56.8 (1.2)
 Other3490.4 (0.2)2.9 (0.5)6.0 (1.0)25.2 (0.9)2.7 (0.2)62.8 (1.4)
Sex
 Men32911.2 (0.2)6.0 (0.7)9.2 (0.9)19.2 (0.8)5.9 (0.8)58.5 (1.5)
 Women42121.3 (0.3)4.3 (0.4)4.6 (0.4)27.4 (1.0)7.2 (0.7)55.2 (1.4)
Education (years)
 ≤1128791.2 (0.4)5.9 (0.7)4.8 (0.4)29.2 (1.8)9.2 (1.5)49.7 (1.8)
  1222921.2 (0.3)6.2 (1.0)6.5 (0.8)25.9 (1.9)6.9 (1.0)53.4 (1.7)
 ≥1323321.3 (0.3)3.8 (0.4)8.0 (0.7)19.3 (1.3)4.9 (0.7)62.6 (1.7)
Smoker
 Current18683.2 (0.8)10.7 (1.0)7.7 (1.1)33.3 (1.6)6.9 (1.0)38.2 (1.8)
 Former15711.0 (0.2)4.8 (0.9)9.4 (1.2)18.6 (1.5)5.2 (1.1)61.0 (2.3)
 Never40640.5 (0.2)2.7 (0.2)4.8 (0.4)19.3 (1.1)7.2 (0.6)65.5 (1.3)
Body mass index (kg m−2)
 <18.51667.5 (3.6)15.1 (4.8)1.6 (0.7)23.0 (3.3)17.1 (4.8)35.7 (5.5)
   18.5–2427991.8 (0.4)5.4 (0.7)7.9 (1.0)18.6 (1.2)5.3 (0.7)61.1 (1.4)
   25–2925280.9 (0.3)4.7 (0.7)7.8 (0.8)22.8 (1.3)4.8 (0.7)58.9 (1.6)
 ≥3020100.7 (0.3)4.6 (0.8)5.0 (1.0)32.4 (1.3)9.0 (1.5)48.3 (1.7)
Cardiovascular disease
 Yes4311.1 (0.4)8.2 (1.4)3.9 (0.9)49.4 (6.7)6.9 (1.5)30.6 (7.6)
 No70721.3 (0.3)4.7 (0.4)6.9 (0.5)22.9 (0.8)6.3 (0.7)57.8 (1.1)
Diabetes mellitus
 Yes5201.2 (0.5)5.3 (1.5)5.6 (2.6)37.0 (4.8)11.2 (2.4)39.7 (6.1)
 No69831.3 (0.2)5.0 (0.4)6.9 (0.5)23.0 (0.8)6.2 (0.6)57.6 (1.1)
Arthritis
 Yes13611.7 (0.6)5.6 (1.1)6.4 (1.1)33.9 (3.2)5.2 (0.8)47.1 (3.7)
 No61421.2 (0.2)5.1 (0.5)7.3 (0.7)21.8 (0.7)6.5 (0.7)58.1 (1.0)
 Total75031.3 (0.2)5.1 (0.4)6.8 (0.5)23.4 (0.7)6.6 (0.6)56.9 (1.0)

Prevalence of functional limitations and reported fair or poor health status are shown in Table 2. The reported prevalence of most of these indicators were higher in older age groups, in the presence of chronic comorbid conditions, amongst African-Americans and Mexican-Americans, amongst women and amongst current smokers.

Table 2.  Participants stratified by lung function, age, race, sex, education levels, smoking status, body mass index, cardiovascular disease, diabetes mellitus, and arthritis and the age-adjusted, weighted proportion (standard error in parenthesis) with functional limitations. From the Third National Health and Nutrition Examination Survey, phase 2, 1991–94
  nUnable to walk a quarter of a mileUnable to lift 10 poundsNeeds help with daily activitiesSelf-reported poor or fair health
  1. aSevere chronic obstructive pulmonary disease (COPD) – forced expiratory volume in 1 s (FEV1)/the forced vital capacity (FVC) < 0.70 and FEV1 < 50% predicted; moderate COPD – FEV1/FVC < 0.70 and FEV1 ≥ 50 to <80% predicted; mild COPD – FEV1/FVC < 0.70 and FEV1 ≥80%; respiratory symptoms – presence of respiratory symptoms in the absence of any lung function abnormality; restrictive lung disease – FEV1/FVC ≥0.70 and FVC < 80% predicted.

Severe COPDa9118.0 (3.1)13.7 (3.0)4.5 (1.2)15.4 (3.2)
Moderate COPD36612.4 (3.7)9.9 (3.7)3.4 (2.5)17.0 (4.6)
Mild COPD5148.4 (2.5)8.7 (1.8)0.7 (0.2)10.2 (3.9)
Respiratory symptoms180715.6 (1.2)13.5 (1.1)2.9 (0.5)23.8 (2.3)
Restrictive lung disease50815.7 (2.3)12.0 (1.9)5.4 (1.2)14.9 (2.0)
No lung disease42175.0 (0.7)4.9 (0.7)0.7 (0.2)7.8 (0.8)
Age (years)
 17–4443485.0 (0.7)5.7 (0.8)0.8 (0.2)9.6 (1.1)
 45–64171911.8 (1.3)10.4 (1.2)1.6 (0.3)15.5 (1.5)
 65–7484719.6 (1.9)13.6 (1.6)2.9 (0.6)23.5 (1.7)
 ≥7558934.9 (2.8)24.0 (3.1)11.6 (1.6)24.8 (3.4)
Race
 White28439.3 (0.7)7.8 (0.7)1.6 (0.2)9.8 (1.0)
 African-American229712.6 (0.9)12.2 (0.8)3.3 (0.4)23.5 (1.2)
 Mexican-American201410.9 (1.1)11.0 (1.1)2.3 (0.6)31.5 (2.0)
 Other3498.9 (1.5)10.9 (2.5)1.0 (0.5)19.5 (2.9)
Sex
 Men32917.4 (0.7)5.2 (0.6)1.2 (0.2)11.4 (1.0)
 Women421211.4 (0.9)11.6 (0.9)2.2 (0.3)14.8 (1.1)
Education (years)
 ≤11287912.7 (1.4)10.8 (1.1)2.3 (0.4)24.7 (2.6)
 12229211.2 (0.8)9.0 (0.8)1.5 (0.3)14.0 (1.2)
 ≥1323326.4 (0.7)6.8 (1.0)1.4 (0.3)6.6 (0.7)
Smoker
 Current186813.8 (1.2)10.0 (1.2)1.4 (0.5)17.7 (1.7)
 Former15717.5 (0.9)7.1 (0.9)1.5 (0.5)10.9 (1.4)
 Never40648.0 (1.0)8.5 (0.9)1.7 (0.3)11.2 (1.0)
Body mass index (kg m−2)
 <18.516614.6 (4.3)11.1 (3.8)3.8 (2.6)14.8 (2.9)
 18.5–2427997.7 (1.2)8.3 (0.9)1.5 (0.3)10.2 (0.9)
 25–2925288.1 (0.7)7.4 (0.7)1.6 (0.3)14.5 (1.3)
 ≥30201014.1 (1.1)11.1 (1.0)2.0 (0.5)15.8 (1.4)
Cardiovascular disease
 Yes43124.0 (7.3)10.6 (2.2)12.6 (6.3)33.4 (6.5)
 No70728.8 (0.7)8.1 (0.6)1.4 (0.2)11.9 (0.9)
Diabetes mellitus
 Yes52021.9 (4.6)16.5 (4.5)4.9 (1.0)30.1 (5.9)
 No69839.0 (0.6)8.3 (0.6)1.6 (0.2)12.4 (0.9)
Arthritis
 Yes136120.5 (2.7)17.0 (1.9)3.4 (0.8)23.3 (2.4)
 No61427.3 (0.6)6.6 (0.6)1.3 (0.2)10.9 (0.9)
 Total75039.5 (0.7)8.5 (0.6)1.7 (0.2)13.1 (0.9)

In univariate models, all strata of impaired lung function or respiratory symptoms were associated with a higher odds ratio (OR) of functional limitation, or fair or poor health status (Table 3). Significant interactions between measures of lung function impairment and other covariates were not detected. After adjustment for covariates, severe COPD, moderate COPD, restrictive lung disease and respiratory symptoms in the absence of impaired lung function remained associated with a higher OR of functional limitations and fair or poor health status.

Table 3.  Results from logistic regression models predicting functional limitations in both univariate models and multivariate models adjusted for age, sex, race, smoking status, body mass index and the presence of chronic disease. From the Third National Health and Nutrition Examination Survey, phase 2, 1991–94. Odds ratio and 95% confidence interval values, respectively, are given in parentheses
 Unable to walk a quarter of a mileUnable to lift 10 poundsNeeds help with daily activitiesSelf-reported poor or fair health
  1. aSevere chronic obstructive pulmonary disease (COPD) – forced expiratory volume in 1 s (FEV1)/the forced vital capacity (FVC) < 0.70 and FEV1 < 50% predicted; moderate COPD – FEV1/FVC < 0.70 and FEV1 ≥ 50 to <80% predicted; mild COPD – FEV1/FVC < 0.70 and FEV1 ≥ 80%; symptoms only – presence of respiratory symptoms in the absence of any lung function abnormality; restrictive lung disease – FEV1/FVC ≥ 0.70 and FVC < 80% predicted.

Univariate
 Severe COPDa19.5 (9.4, 40.2)11.6 (6.3, 21.3)28.4 (10.9, 74.4)8.1 (3.8, 17.4)
 Moderate COPD5.7 (3.4, 9.6)3.2 (1.8, 5.5)6.9 (2.2, 21.9)3.7 (2.3, 6.0)
 Mild COPD3.0 (1.5, 5.7)2.6 (1.5, 4.3)3.1 (1.4, 6.6)1.8 (1.1, 2.9)
 Respiratory symptoms4.1 (2.8, 6.0)3.2 (2.3, 4.3)6.5 (4.3, 9.8)4.1 (3.2, 5.2)
 Restrictive lung disease5.1 (3.0, 8.7)3.4 (2.3, 5.0)14.5 (6.7, 31.1)2.8 (2.0, 3.9)
 No lung disease 1.00 1.00 1.001.00
Multivariate
 Severe COPD8.4 (3.6, 19.9)7.7 (3.8, 15.4)16.2 (4.9, 53.5)5.3 (2.0, 13.6)
 Moderate COPD2.4 (1.4, 4.0)2.1 (1.1, 4.0)3.1 (0.7, 14.9)2.1 (1.1, 3.8)
 Mild COPD1.9 (0.9, 3.7)2.2 (1.3, 4.0)1.7 (0.7, 4.5)1.4 (0.8, 2.6)
 Respiratory symptoms2.8 (2.0, 3.9)2.5 (1.7, 3.5)5.0 (2.8, 9.2)3.2 (2.5, 4.1)
 Restrictive lung disease2.8 (1.4, 5.6)2.3 (1.5, 3.6)9.0 (3.1, 26.6)1.7 (1.2, 2.4)
 No lung disease 1.00 1.00 1.001.00

Discussion

Analysis of cross-sectional, nationally representative data from the US showed that the presence of severe COPD was associated with a higher OR of three measures of functional impairment and a fair or poor health status. Whilst severe COPD had the highest OR of limitations, moderate COPD, restrictive lung disease and respiratory symptoms in the absence of impaired lung function were all associated with a higher OR of functional limitations and a fair or poor health status to a similar extent.

The prior analysis of this database restricted the analysis to white participants ≥45-years old and focused only on obstructive lung disease as the outcome of interest [3]. In that analysis the comparison group of ‘normal’ subjects included people with both restrictive lung disease and respiratory symptoms, which could have resulted in underestimating the effect of COPD on functional limitations or health status. For example, in the prior analysis the multivariate regression model for fair or poor health status amongst subjects with Stage II or III COPD (roughly corresponding to severe COPD category of the present study) yielded an OR of 2.7 [95% confidence interval (CI) 1.1, 5.8] whereas the present analysis yielded an OR of 5.3 (95% CI 2.0, 13.6), with similar findings for the other end-points [3]. Similarly, none of the models in the prior analysis amongst subjects with Stage 1b COPD (roughly corresponding to moderate COPD category of the present study) were associated with an increased risk of limitations, whereas all the present models, with the exception of ‘Needs help with daily activities’ were.

The finding that restrictive lung disease is associated with a higher OR of functional limitations and fair or poor health status is expected. Restriction on spirometry can be related to a variety of aetiologies, including interstitial lung disease, obesity, muscular weakness and small airways disease [4, 18, 19]. The precise determination of restriction requires the measurement of total lung capacity [20] although others have shown that most patients with restriction on spirometry have this confirmed when the total lung capacity is measured [21]. In addition, it is possible that subjects with ‘normal’ lung function who did not complete the test were included in the ‘restrictive’ category. This bias could have the potential to underestimate functional limitation amongst subjects with restrictive disease [22].

A surprising finding was that people reporting respiratory symptoms in the absence of lung function impairment also reported fair or poor health status, or functional limitations at a level similar to that seen in moderate COPD. In the GOLD criteria for COPD the presence of respiratory symptoms in the absence of lung function impairment is considered Stage 0 COPD, with the recommended intervention being ‘information and advice about reducing risk factors’ [14]. The data suggest that some of these people may have significant limitations related to these symptoms, with the possibility that they might benefit from interventions for chronic cough or sputum production or aimed at the underlying aetiology of their symptoms [23]. ‘Respiratory’ symptoms, particularly dyspnoea on exertion, may also be related to diagnosed or undiagnosed cardiovascular disease. Although diagnosed cardiovascular disease was controlled for in this analysis (and a higher proportion of subjects with diagnosed disease had respiratory symptoms in the absence of lung function impairment, Table 1), it is possible that undiagnosed cardiovascular disease may be related to both the presence of respiratory symptoms and the reporting of functional limitations [10].

Another somewhat surprising finding was the high proportion of subjects with lung function impairment who did not report functional limitations. For example, only 18% of subjects with severe COPD reported they were unable to walk a quarter of a mile (Table 2). This paradox can be explained, in part, by age-adjustment, which weighted the results towards the younger participants. In non age-adjusted results, 45.5% of subjects with severe COPD reported being unable to walk a quarter of a mile, but these were over represented amongst the older subjects. It is still likely, however, that people underreport functional limitations.

This analysis has some limitations. As a cross-sectional study, it could not determine whether the lung function impairment or respiratory symptoms actually led to the higher level of functional limitations. This is particularly true with regard to respiratory symptoms; it is possible that people reporting more respiratory symptoms report many things, including functional limitations, more frequently. Over 20% of the eligible sample was missing data on either covariates or outcomes and was excluded from the analysis. Those excluded tended to be older and, presumably, sicker, than those included, so findings may actually underestimate the extent of limitation associated with lung function impairment or respiratory symptoms.

In conclusion, this analysis finds that obstructive lung disease, restrictive lung disease and respiratory symptoms are all associated with a significantly higher risk of functional impairment and fair or poor health status. A large proportion of lung disease is undiagnosed in the US, which has led to initiatives such as the National Lung Health Education Program that promote more widespread use of spirometry in primary care settings [24, 25]. Although the main objective of this programme is the early detection of COPD, spirometry would also detect restrictive lung disease, which would require further clinical evaluation and treatment. In addition, this analysis highlights the importance of chronic respiratory symptoms that may be amenable to clinical interventions or considered in studies looking at functional limitations as end-points.

Conflict of interest statement

No conflict of interest declared.

Ancillary