COPD IN CONSTRUCTION WORKERS:

RESULTS FROM TWO MEDICAL

MONITORING PROGRAMS

John Dement Ph.D., CIH1, Laura Welch, M.D.2,

Knut Ringen Dr. P.H.3

1Division of Occupational and Environmental Medicine, Duke University Medical Center 2The Center for Construction Research and Training (CPWR) 3 Stoneturn Consultants

SUPPORT: DOE and NIOSH

john.dement@dm.duke.edu

Presentation Overview

• Briefly review COPD and known risk factors.

• COPD prevalence among DOE construction trade

workers participating in a medical examination

program.

• COPD prevalence among sheet metal workers

participating in a medical examination program.

• Trade-specific COPD risks estimates after

adjustment for confounders such as smoking.

• Exploratory analyses of COPD risks associated

with specific exposures.

What is COPD?

• American Thoracic Society [ATS, 1995] defines COPD as:

– Presence of airflow obstruction due to chronic bronchitis and emphysema.

– COPD airflow limitation is irreversible.

• Chronic airflow limitation in COPD is caused by a combination of:

– Small airway disease (obstructive bronchiolitis) .

– Lung parenchymal destruction (emphysema).

COPD vs. ASTHMA

Asthma – Reversible airflow

limitation

– Onset early in life, often

childhood

– Accompanied by allergies,

rhinitis

– May be symptom free

between attacks

COPD

– Irreversible airflow

limitation

– FEV1 declines with age

greater than normal

– Onset in midlife, typically

after age 50

– Strongly linked to smoking

history

– Symptoms include chronic

cough, sputum production,

and dyspnea.

Clinical COPD Diagnosis

• Symptoms:

– Chronic cough, sputum production, dyspnea

on exertion.

• Spirometry:

– Post bronchodialator FEV1 < 80% predicted.

and

– FEV1/FVC ratio <0.70.

COPD Disease Burden

• NHLBI estimates that 12 million people in U.S. have diagnosed COPD and an additional 12 million have COPD but undiagnosed.

• 130,993 U.S. deaths due to COPD in 2005. Fourth leading cause of death in U.S. and sixth worldwide.

• COPD cost $37.2 billion in 2004. – $20.9 billion in health care expenditures.

– $7.4 billion indirect morbidity costs.

– $8.9 billion indirect mortality costs.

COPD Risk Factors

• Hereditary deficiency in α1-antitrypsin (ATT) – Less than 1% of COPD patients have ATT

deficiency, so PAR <1%.

• Cigarette smoking – PAR estimated to be 80-90%;

– Only 15-20% of smokers develop COPD and 10% of COPD deaths occur in lifetime non-smokers.

• Occupational exposures – ATS (2003) estimated an occupational PAR of 15%.

– Blanc (2009) suggested occupational PAR might be higher due to joint effects of occupational exposures and smoking.

Occupational Exposures and COPD Risk

• Non-specific exposures to ‘vapors, gases, dusts, and fumes (VGDF)

• Specific exposures – Coal dust

– Welding

– Silica

– Diesel exhausts

– Cement dust

– Spray painting with isocyanate-based paints

– Wood dust

– Cadmium

DOE Sites – Large Industrial Complex Fernald 1994

DOE Medical Surveillance Programs • Established by Congress in 1993

• 1996 and 1997 DOE funded programs at Hanford, Oak Ridge, and Savannah River.

• Expanded to form the Building Trades National Medical Screening Program (BTMED) with >20 Sites

• BTMED consortium: – The Center for Construction Research and Training

(CPWR).

– University of Cincinnati

– Zenith Administrators

– Duke University

Source: https://www.btmed.org/locations.cfm

BTMED Exam Respiratory Components

• Complete medical history.

• Personal history including smoking and hobby activities.

• Respiratory history and symptom questionnaire (ATS DLD-78).

• Posterior-anterior (P-A) chest radiograph, classified by a NIOSH B-reader according to ILO criteria.

• Spirometry by ATS standards.

BTMED Work History

• Contractors

• Crafts/Trades

– Work duration at DOE Site

– Total work duration in Trade

• Materials/Agents

– Scale of exposure frequency (0-5)

• Work Tasks Performed

– Scale of task frequency (0-5)

• Work Tasks Performed by Others Nearby (Bystander Exposures)

– Scale of task frequency (0-5)

• Buildings or Work Areas

– Work in selected beryllium buildings

– List most hazardous buildings with list of known hazards

– Work in Critical Buildings (e.g. beryllium or other hazards)

• Exposure Incidents

• Additional Questions about Certain Exposures (e.g. mercury,

radiation, noise, beryllium, etc.)

Work History Frequency Scale

Tasks, Materials, and Bystander Exposures

Scale

Description

5 Continuous

4 Daily or most days per week

3 Couple times per week

2 Few times per month

1 Rarely

0 No Reported Exposure

BTMED DOE Worker COPD Study

• Study Population Included 7579 Workers:

– Completed initial exam with spirometry

through September 2008.

– Not missing data on age, race, sex, or height

needed to calculate predicted PFT values.

– Spirometry meeting ATS criteria

• Minimum of 3 recorded expiratory efforts.

• Reproducibility by ATS criteria.

BTMED COPD Study Case Definition

• Spirometry predicted values based on

Hankinson et al. equations derived from

NHANES III data.

• COPD Case Definition:

– FEV1/FVC ratio of less than 0.70

– FEV1 less than 80 percent predicted.

BTMED COPD Study Data Analyses

• Full Cohort Analyses (N=7579 workers) – Crude COPD prevalence by age, sex, race, and

trade.

– Logistic regression: COPD risk by occupation controlling for age, sex, race, smoking, and years of DOE site work.

• Cohort Restricted to 4758 workers < 5 years of trade work outside of DOE sites – Logistic regression: COPD associated with specific

exposures from work histories controlling for age, sex, race, smoking, and years of DOE site work.

Comparison of Workers by COPD Status Demographics & Spirometry

COPD

(N=1136)

Mean Age (SE) 56.9 (0.15) 64.6 (0.35) <0.01

Male Sex (%) 5974 (92.7) 1102 (97.0) <0.01

Caucasian Race (%) 5511(85.5) 1052 (92.6) <0.01

Years at DOE, Mean (SE) 8.9(0.11) 9.5 (0.30) <0.05

Years in Trades2, Mean (SE) 23.9 (0.18) 27.8 (0.44) <0.01

% Predicted FVC 89.4 (0.21) 72.7 (0.48) <0.01

% Predicted FEV1 91.3(0.20) 57.5 (0.46) <0.01

FEV1/FVC Ratio 0.79 (0.001) 0.59 (0.003) <0.01

P-value

Spirometry Mean (SE)

Parameter

No COPD

(N=6443)

Comparison of Workers by COPD Status Symptoms, ILO Category, & Smoking

COPD

(N=1136)

Respiratory Symptom Prevalence (%)

Cough (N=6258) 1518 (28.4) 523 (56.7) <0.01

Phlegm (N=6356) 1579 (29.3) 564 (58.6) <0.01

Dyspnea (N=7145) 1785 (29.4) 679 (63.2) <0.01

B-Reader Prevalence (%)

Pleural Changes Only 828 (12.9) 247 (21.7) <0.01

Parenchymal Changes Only 101 (1.6) 34 (3.0)

Both Pleural and Parenchymal 114 (1.8) 55 (4.8)

No B-Read Results Available 209 (3.2) 50 (4.4)

Cigarette Smoking Prevalence (%)

Current 1110 (17.2) 364 (32.0) <0.01

Former 2653 (41.2) 561 (49.4)

Never Smoked 2529 (39.3) 167 (14.7)

Smoking Unknown 164 (2.4) 44 (3.9)

P-value

No COPD

(N=6443)Parameter

Comparison of Workers by COPD Status Prior Respiratory Diagnoses, BMI, & Hypertension

Parameter No COPD (N=6443)

COPD1

(N=1136) P-value

Physician Diagnosis (Ever) (%)

Asthma (N=7143) 489 (8.0) 217 (21.3) <0.01

Chronic Bronchitis (N=7053) 419 (6.9) 212 (21.2) <0.01

Emphysema (N=7196) 173 (2.8) 259 (25.2) <0.01

Mean BMI (N=7954) (SE) 29.9 (0.07) 28.4 (0.16) <0.01

Hypertension Prevalence (N=7560) (%) 2107 (32.8) 416 (36.7) 0.02

Crude COPD Prevalence by Trade

COPD

Prevalence

(%)

Administrative/Scientific/Security 254 6.7 4.0 10.5

Asbestos Workers/Insulator 219 18.7 13.8 24.5

Boilermaker 150 14.0 8.9 20.6

Carpenter 531 19.6 16.2 23.0

Cement Mason/Brick Mason/Plasterer 104 24.0 15.8 32.2

Electrician 1298 14.0 12.1 15.9

Ironworker 389 16.2 12.7 20.2

Laborer 958 13.7 11.5 15.9

Millwright 153 19.6 13.6 26.7

Operating Engineer 488 14.6 11.5 18.0

Painter 219 15.5 11.0 21.0

Plumber, Steamfitter, Pipefitter 1431 15.6 13.7 17.6

Sheet Metal Worker 413 16.2 12.8 20.1

Teamster 278 16.1 12.1 21.1

All Other Workers 694 11.8 9.5 14.5

OVERALL 7579 15.0 14.2 15.8

95% LCLTrade Group 95% ULCNumber Exams

Logistic Regression Results by Trade

Prevalence

Odds Ratio

Administrative/Scientific/Security 1 Ref Ref

Asbestos Workers/Insulator 2.66 1.42 4.97

Boilermaker 1.44 0.69 2.99

Carpenter 2.39 1.37 4.19

Cement Mason/Brick Mason/Plasterer 3.09 1.53 6.22

Electrician 1.73 1.01 2.96

Ironworker 1.99 1.11 3.57

Laborer 1.93 1.12 3.33

Millwright 2.21 1.13 4.3

Operating Engineer 1.66 0.94 2.96

Painter 1.96 1.03 3.74

Plumber, Steamfitter, Pipefitter 1.83 1.07 3.11

Sheet Metal Worker 1.97 1.11 3.52

Teamster 1.79 0.96 3.31

All Other Workers 1.57 0.9 2.76

Trade Group 95% LCL 95% ULC

Adjusted for age, sex, race, smoking, and years of DOE work

Exposure Estimates for Exploratory Analyses Summary Exposure Metric

• Combined data for multiple tasks associated

with exposures.

– Considered asbestos, silica, welding & cutting,

cement dust, solvents, paints, and dust/fumes from

paint removal.

– Task frequency scores (0-5) weighted to account for

relative frequency of daily exposure.

• Weighted frequencies then summed by

exposure to provide a relative exposure metric.

Task Exposure Summary Measure Task Summary Score Example

• Asbestos Tasks – Multiple exposed tasks scored 0-5 during the

interview.

• Task Summary Score = Sum of all weighted asbestos task frequency scores.

Task Description Ordinal Task

Frequency

Exposure

Days

Per Month

Drill, grind, cut, or apply asbestos

containing insulation or Transite 3 8

Sand or refinish asbestos floor tiles 2 2

Finish or sand drywall 1 1

OVERALL ASBESTOS EXPOSURE INDEX 11

Exposure Estimates for Exploratory Analyses Exposure Modeling Procedures

• Exposure modeling used to form more homogeneous exposure groups. – Reduce attenuation of exposure-response relationship typically

found with individual exposures.

• Distributions of the exposure metrics were approximately log-normal; therefore, individual summary metrics were log-transformed in models.

• Independent variables as exposure predictors in SAS Proc GENMOD models: – Trade nested within DOE site

– Time period of first DOE work

– Sex and Race

• Exposures assigned to workers based on predicted values from log-linear models.

Exposure Estimates for Exploratory Analyses Exposure Metric and Model Validation

• Task exposure scores and model predicted exposure scores for asbestos used in logistic models of chest x-ray changes by ILO criteria.

• Cases: Workers with ILO profusion score ≥ 1/0 or having bilateral asbestos-related pleural changes. – 4758 workers with chest radiograph and with five or

fewer years of trade work outside DOE.

– 695 cases included in the analyses.

• Logistic models controlled for age, sex, race, smoking, and years of DOE work.

Exposure Metric & Model Validation Chest Radiograph Changes by ILO Criteria

1Odd-ratios controlling for age, sex, race, smoking, and years of DOE work.

The odds-ratios represent changes in risk for an increase of one standard

deviation in the given log of the exposure parameter.

Asbestos Exposure Variable

Odds-Ratio1

95% LCL

95% ULC Model AIC

Sum of Asbestos Task Scores 1.166 1.062 1.282 3363

Linear Model Predicted Sum of

Task Scores 1.348 1.218 1.492 3339

Exposure Model Validation

Odds-Ratios for Asbestos-Related CXR Changes(Adjusted for Age, Sex, Race, Smoking, and Years of DOE Work)

0.5

1

1.5

2

2.5

3

3.5

1 2 3 4 5 6 7 8 9 10

Mixed Model Predicted Log Exposure Score Category

Od

ds

-Ra

tio

Model Pred Total Score Raw Total Score

COPD Risk by Predicted Total Task Exposures Logistic Models Adjusted for Age, Sex, Race, Smoking and Years of DOE Work

Analyses based on sub-cohort of 4935 workers with five or fewer years of trade work

outside of DOE sites and data available for all other model covariates

The odds-ratios represent changes in risk for an increase of one standard deviation in the

given log of the predicted exposure score.

Statistical Model

Predicted Exposure Scores

Asbestos Task Total 1.115 1.01 1.23 3289

Welding Task Total 1.055 1.001 1.114 3290

Silica Task Total 1.123 1.017 1.241 3288

Cement Task Total 1.132 1.025 1.25 3288

95% LCLOdds-Ratio 95% ULCModel

AIC

Sheet Metal Worker COPD Study

• Sheet Metal Occupational Health Institute Trust (SMOHIT) in 1985 to study respiratory health hazards among sheet metal workers.

• Sheet metal workers employed for at least 20 years in the construction sector are eligible. – Brief occupational history.

– Chest x-rays using the ILO classification.

– Pulmonary function by ATS standards.

– A respiratory history and symptom questionnaire adapted from the ATS) DLD-78 questionnaire.

• >17,000 workers participated through 2004

Enhanced Sheet Metal Exposure Assessment

• Jobs held for at least three months.

• A qualitative assessment of exposure frequency (none to daily) for an established list of respiratory hazards (e.g. silica, asbestos, wood dust, cement dusts, paints, glues, acids, etc.) found in construction.

• A qualitative assessment of frequency of doing or working near selected construction tasks known to generate respiratory exposures.

• A qualitative assessment of home and hobby respiratory exposures.

Sheet Metal Worker Study Population

Using Enhanced Exposure Assessments (N=1967)

• Completed an examination with spirometry

through December 31, 2009

• Not missing demographic data (age, race,

sex, and height).

• Spirometry that included a minimum of three

recorded expiratory efforts and reproducibility

of FVC and FEV1 of 0.2 liters or less.

– COPD Case: FEV1/FVC ratio below the lower limit

of normal (LLN) using the prediction equations of

Hankinson et al. [1999].

Sheet Metal Exposure Assessment

• Specific list of tasks, working near other individuals performing tasks, or having exposures to defined list of agents.

0 - No reported agent exposure.

1 - 1-2 times per month or less.

2 - Weekly or most weeks.

3 - Daily or almost every day.

• Exposure indices: Estimated days of exposure per year multiplied by years of work to generate a cumulative estimate.

• Principal component analysis (PCA) used to account for mutual correlation of the individual task exposure scores. – PCA weights or ‘loadings’ multiplied by each worker’s task

scores to derive a summary score for each exposure of interest .

Sheet Metal Example Exposure Index

Task Description

Ordinal

Task

Frequency

Exposure

Days

Per Month

Years of

Exposure

Cumulative

Index

Asbestos Cement (Transite) Work

(drill, cut, grind) 3 20 5 100

Scrape or Remove Asbestos

Fireproofing from Beams/Decks 2 4 20 80

Sand/Refinish Asbestos Floor Tiles 1 1 10 10

OVERALL ASBESTOS EXPOSURE INDEX 190

Sheet Metal Worker

Crude COPD Prevalence Variable Number of

Workers

COPD Cases

and

Prevalence

(%) Age Category (Years)

40-49 301 30 (10.0%)

50-59 757 72 (9.5%)

60-69 587 51 (8.7%)

70+ 322 40 (12.4%) Gender

Male 1954 193 (9.9%) Female 9 0 (0.0%)

Race/Ethnicity

Caucasian 1807 177 (9.8%) All Other 160 16 (10.0%)

Crude Overall Prevalence

1967 193 (9.8%)

Logistic Regression Odd-Ratios Dichotomous Exposures with Adjustment for Age, Race, Smoking

Task or Material Crude Odds-Ratio

and (95% CI)

Adjusted Odds-

Ratio and (95% CI)

Laser Beam Welding 1.64 (1.08-2.52) 1.72 (1.11-2.68)

Use Acids or Caustics for Cleaning or Degreasing 1.38 (1.01-1.89) 1.32 (0.95-1.83)

Work Near Scraping or Removing Paints 1.36 (1.01-1.84) 1.40 (1.03-1.91)

Exposure to Coal Dust 1.42 (1.04-1.95) 1.34 (0.97-1.85)

Exposure to Lime Dust 1.48 (1.05-2.10) 1.50 (1.05-2.14)

Exposure to Talc Dust 1.55 (1.02-2.36) 1.60 (1.04-2.45)

Exposure to Wood Sealers 1.41 (1.01-1.96) 1.54 (1.10-2.18)

Exposure Principal Component Odds-Ratios Adjusted for Age, Race, Smoking

Exposure Index

Principal Components Accounting for

>10% of Total Factor Loading,

Ranked from Highest to Lowest

Odds

Ratio

95%

LCL

95%

ULC

Model

AIC

All Tasks and Materials

Principal Component

Electron Beam Welding

Laser Beam Welding Submerged Arc Welding

Carbon Arc Welding Plasma Arc Welding

Oxy-fuel Welding Brazing

Resistance Welding

Renovate/ Demolish Ducts Mold Contamination

1.25 1.08 1.43 1209

All Tasks Principal

Component

Laser Beam Welding

Submerged Arc Welding Electron Beam Welding

Oxy-fuel Welding Resistance Welding

Plasma Arc Welding Carbon Arc Welding

Brazing

Renovate/ Demolish Ducts Mold Contamination

1.21 1.05 1.39 1212

Molds/Spores Tasks

Principal Component Renovate/ Demolish Ducts Mold

Contamination 1.16 1.02 1.32 1214

Welding/Cutting Tasks

Principal Component Submerged Arc Welding 1.15 1.01 1.32 1215

Conclusions

• Construction workers experience a wide variety of exposures with prior associations with COPD.

• Construction and craft workers are at increased risk of COPD after controlling for major confounders including smoking.

• Exploratory analyses demonstrated increased COPD risk to be associated with exposures to asbestos, silica, welding, cement dust, molds/spores and possibly other exposures.

• Group-level exposure estimates from exposure models were more strongly associated with risk.

• Principal component methods proved useful in analyses of correlated exposure measures.

Future Research

• Case-control study of DOE workers to

further explore associations suggested in

the exploratory analyses

• Further prospective study of sheet metal

workers to increase the number of

participants with increasd statistical

power.