Low income as a determinant of exercise capacity in COPD

Abstract

Exercise capacity (EC) is a critical outcome in chronic obstructive lung disease (chronic obstructive pulmonary disease (COPD)). It measures the impact of the disease and the effect of specific interventions like pulmonary rehabilitation (PR). EC determines COPD prognosis and is associated with health-care utilization and quality of life. Field walking tests and cardiopulmonary exercise test (CPET) are two ways to measure EC. The 6-minute walking test (6MWT) is the commonest and easiest field test. CPET has the advantage of assessing maximal aerobic capacity. Determinants of EC include age, gender, breathlessness, and lung function. Previous research suggests that socioeconomic status (SES), a meaningful factor in COPD, may also be associated with EC. However, those findings have not been replicated. We aimed to determine whether SES is an independent factor associated with EC in COPD. For this analysis, we used the National Emphysema Treatment Trial (NETT) database. NETT was a multicenter clinical trial where severe COPD patients were randomized to lung volume reduction surgery or medical therapy. Measures used were taken at baseline, postrehabilitation. Patients self-reported their income and were divided in two groups whether it was less or above US$30,000. Patients with a lower income had worse results in 6MWT (p < 0.0001). We found an independent association between income and the 6MWT in patients with severe COPD after adjusting for age, gender, lung function, dyspnea, and living conditions (p < 0.0007). One previous publication stated the relationship between income and EC. Our research confirms and extends previous publications associating EC with income by studying a large and well characterized cohort of severe COPD patients, also addressing EC by two different methods (maximal watts and 6MWT). Our results highlight the importance of addressing social determinants of health such as income when assessing COPD patients.

Introduction

Exercise capacity (EC) is a critical outcome in chronic obstructive lung disease (chronic obstructive pulmonary disease (COPD)). ¹ It measures the impact of the disease and tests the effect of specific interventions. ^{1 –3} EC determines COPD prognosis due to its relation with hospitalizations for exacerbations and all-cause mortality. ^4,5 Importantly, from a patient-centered focus, it is related to health-care utilization, quality of life (QoL), and symptoms. ³

Field walking tests and cardiopulmonary exercise test (CPET) are two ways to measure EC. Among those, the 6-minute walking test (6MWT) is the commonest and easiest. ^1,2,6 CPET has the advantage of assessing maximal aerobic capacity and contributes better to the differential diagnosis. ⁷ EC determinants are many, ^6,8 including age, breathlessness, and lung function.

Socioeconomic status (SES), a meaningful factor in COPD related to QoL, ⁹ physical activity, ¹⁰ completion of pulmonary rehabilitation (PR), and all-cause mortality, ¹¹ may also predict EC. Recent data show that SES also influences EC, with lower income patients having poorer results ¹² ; however, that finding has never been replicated.

Our aim is to determine whether income is an independent factor associated with EC in COPD, which would raise its importance as a meaningful factor to be assessed.

Methods

The present analysis used the National Emphysema Treatment Trial (NETT) database. ¹³ This multicenter trial randomized severe COPD patients with no significant comorbidities to lung volume reduction surgery or medical therapy, after 6–10-week PR. ^13,14 Smoking cessation, biochemically validated, was required at least 6 months before randomization. Patients were divided in two groups whether their annual reported income was less or above US$30,000, a limit chosen following data of US census bureau and Pew Research center. ^15,16

EC was measured by the 6MWT and maximal watts during CPET, following the American Thoracic Society Guidelines (ATS). ^6,17 Pulmonary function tests followed ATS. ^18,19

Statistical analyses, associations between income and categorical variables, were tested between income groups using χ²tests. Associations between income level and continuous variables were tested between groups using Wilcoxon tests. The association between income and 6MWT as well as income and CPET was explored using linear models adjusted for marital status, loneliness, age, gender, FEV1% predicted, modified medical research council scale (MMRC), smoking pack years, and education. All analyses were tested with 5% type I error rates with no adjustments for multiple testing. Analyses were carried out using SAS version 9.4.

Results

In total, 1218 patients completed PR and were analyzed using baseline data. Table 1 shows population’s characteristics and Tables 2 and 3 show the models designed to determine the independent association between income and EC.

Table 1.

Characteristics of all 1206 patients at baseline^a.

Characteristics	Income ≥US$30,000(N = 563)	Income ≤US$30,000(N = 643)	p Value
Medical treatment	52.0%	49.1%	0.3153
Sex, male	65.4%	57.5%	0.0054
Age, mean (SD) (years)	67.2 (5.5)	65.6 (6.6)	0.0001
Married	83.3%	48.5%	<0.0001
Lonelyb	5.5%	10.1%	0.0031
Smoking pack years	65.4 (29.3)	63.6 (32.6)	0.0826
Post BD FEV1% pred, mean (SD)	26.8 (7.1)	26.6 (7.2)	0.3947
6MWD in feet, mean (SD)	1250.7 (314.8)	1171.7 (306.8)	<0.0001
Maximum load (watts), mean (SD)c	41.6 (22.7)	36.8 (20.4)	0.0001
Total days in hospital, mean (SD)	3.0 (7.8)	3.3 (7.6)	0.3840
Any days in hospital, yes	31.7%	33.8%	0.4620
Total visits to ER, mean (SD)	0.6 (1.1)	0.7 (1.3)	0.5015
Any visit to ER, yes	37.5%	39.7%	0.4659
Follow-up status, deadd	39.4%	41.7%	0.4278
MRC Dyspnea Scale			0.0058
0	1.4%	1.6%
1	0.2%	0.5%
2	29.2%	22.1%
3	23.5%	19.9%
4	45.7%	55.9%
Education beyond high school	62.9%	35.1%	<0.0001

N: sample; SD: standard deviation; Post BD FEV1% pred: postbronchodilator forced expiratory volume in 1 second % predicted; ER: emergency department; MRC scale: Medical Research Council Dyspnea Scale; 6MWD: 6-minute walk distance. Every value in italics <0.05 denotes significance (standard practice).

^a Base-line measurements were obtained after rehabilitation but before randomization.

^bLoneliness was measured by asking subjects “Do you feel lonely or socially isolated now or in the last 3 days?” An answer was either “yes” or “no.” ²⁰

^c Maximum load (watts) was chosen to measure exercise capacity during CPET.

^d The follow-up period was 2 years.

Table 2.

Income as a determinant of 6MWT, adjusted for age, sex, Post FEV1% predicted, MMRC, loneliness, married status, smoking pack years, and education beyond high school.

Variables	DF	Parameterestimates	Standard error	t Value	Pr > |t|
Intercept	1	1607.40	95.58	16.82	<0.0001
Married	1	27.47	18.45	1.49	0.1367
Lonely or Isolateda	1	−23.24	29.50	−0.79	0.4310
Income < US$30,000	1	−48.15	17.76	−2.71	0.0068
Age	1	−10.87	1.35	−8.05	<0.0001
Sex, male	1	180.85	17.82	10.15	<0.0001
Post BD FEV1% pred	1	16.10	1.17	13.75	<0.0001
MMRC	1	−67.59	8.61	−7.85	<0.0001
Education beyond high schoolb	1	32.89	16.78	1.96	0.0502
Smoking pack years	1	−0.11	0.26	−0.42	0.6720

DF: degrees of freedom; Pr > |t|: t-test significance level; Post BD FEV1% pred: post bronchodilator forced expiratory volume in 1 second; MMRC: Modified medical research council scale. Every value in italics <0.05 denotes significance (standard practice).

^aLoneliness and isolation was measured by asking subjects “Do you feel lonely or socially isolated now or in the last 3 days?” An answer was either “yes” or “no.” ²⁰

^bEducation was determined by dividing patients into two groups: those who had education levels beyond high school and those who did not.

Table 3.

Income as a determinant of watts in CPET, adjusted for age, sex, Post FEV1% predicted, mMRC, loneliness, married status, smoking pack years and education beyond high school.

Variables	DF	Parameter estimates	Standard error	t Value	Pr > |t|
Intercept	1	48.24	5.99	8.06	<0.0001
Married	1	1.79	1.16	1.55	0.1222
Lonely or isolateda	1	−0.48	1.85	−0.26	0.7957
Income < US$30,000	1	−2.27	1.11	−2.04	0.0419
Age	1	−0.77	0.08	−9.14	<.0001
Sex, male	1	22.14	1.12	19.83	<0.0001
Post BD FEV1% predicted	1	1.36	0.07	18.60	<0.0001
MMRC	1	−2.88	0.54	−5.34	<0.0001
Education beyond high schoolb	1	2.14	1.05	2.04	0.0417
Smoking pack years	1	0.01	0.02	0.47	0.6416

DF: degrees of freedom; Pr > |t|: t-test significance level; Post BD FEV1% pred: postbronchodilator forced expiratory volume in 1 second; MMRC: Modified medical research council scale. Every value in italics <0.05 denotes significance (standard practice).

^aLoneliness and isolation was measured by asking subjects “Do you feel lonely or socially isolated now or in the last 3 days?”. An answer was either “yes” or “no.” ²⁰

^bEducation was determined by dividing patients into two groups: those who had education levels beyond high school and those who did not.

The association between 6MWT and income (Table 2) remained significant after adjusting for meaningful confounders such as sex, age, Post BD FEV1% predicted, MMRC, marriage, loneliness, smoking pack years, and education beyond high school (yes or no). We also found a significant association between income and maximal load in watts on CPET (Table 3) validating the 6MWT results. In addition, our results showed individuals with low income found themselves lonelier (Tables 1 and 2).

Discussion

We found that low income is associated with worse results in the 6MWT after adjusting for meaningful confounders. The latter was further validated using another measure of EC: watts in CPET. Our findings confirm and extend a previous report ¹² but in a larger and more characterized cohort of severe COPD, using two different EC measures.

Given the 6MWT’s prognostic significance, ^{4 –6} our findings are important because usually income is not taken into account in COPD evaluation, despite being a factor that is easily obtainable, and conveys information about patients’ overall social condition. Plausible explanations for our results might include that financial constrains may limit patients’ access to programs like PR, as stated by previous publications. ^{21 –23} The fact of finding low-income participants more lonely suggests another social, yet critical repercussion of low income. Loneliness is independently related to poorer outcomes and higher health-care utilization in COPD (article in press).

Limitations

Given that NETT selected severe emphysema COPD patients without significant comorbidities, our results may limit generalizability.

Conclusions

In severe COPD patients, low income is independently associated with EC. Our results may be of importance to fully assess patients with severe COPD addressing social determinants of health.