Longitudinal analysis of respiratory symptoms in population studies with a focus on dyspnea in marine transportation workers

Our plain-English summary of Longitudinal analysis of respiratory symptoms in population studies with a focus on dyspnea in marine transportation workers, by V.H. Arrandale, M. Koehoorn, Y. MacNab, et al (International Archives of Occupational and Environmental Health, 2009).

The study

• Authors: V.H. Arrandale, M. Koehoorn, Y. MacNab, S.M. Kennedy
• Journal: International Archives of Occupational and Environmental Health
• Published: 2009;82(9):1097-105
• Cited by: 16 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

Purpose: Longitudinal respiratory symptoms are rarely studied in occupational epidemiology. We investigated dyspnea change over time and predictors of change over time using two longitudinal modeling techniques, a semi-parametric group-based approach (SAS® Proc Traj) and a generalized linear mixed model (SAS® Proc Glimmix), and compared the two techniques for use in longitudinal studies of respiratory symptoms.

Methods: Data were previously collected from a lung health surveillance study of marine transportation workers. Subjects were seen two to four times over 12 years (1987-1999). At each visit the American Thoracic Society questionnaire was administered and lung function was tested. The semi-parametric group-based model and the generalized linear mixed model were applied to the data.

Results: The group-based trajectory model supported two groups of dyspnea change over time. Group 1 (73%) had a steady low-level probability of reporting dyspnea over follow-up, while Group 2 (27%) had an increasing probability of reporting dyspnea over follow-up. The generalized linear mixed model (random intercept) estimated that the probability of reporting dyspnea was increasing over time in the population. Current smoking, female sex, lower lung function and older age were associated with increased probability of reporting dyspnea in both models.

Conclusions: Results from both models indicate that the probability of reporting dyspnea was increasing over time in this occupational cohort. The group-based model is capable of identifying multiple patterns of linear and non-linear change while the generalized linear mixed model is preferable when the population mean change (linear) is of interest. Both approaches were able to identify similar characteristics associated with longitudinal dyspnea symptoms.

In plain English

This study reanalyzed surveillance data on marine transportation workers on ferries in British Columbia. The surveillance arose from concern about possible low-level asbestos exposure. Of 1,701 participants, the 925 with at least two visits between 1987 and 1999 were analyzed. The authors tracked reported shortness of breath (dyspnea) and compared two statistical methods. Dyspnea prevalence rose from 15.6% of men and 27.3% of women at baseline to 28.8% and 43.8% at final follow-up. One method found two groups. In 73% of workers the probability of dyspnea stayed steady and low. In the other 27% it rose over time (odds ratio 1.15 per year, 95% confidence interval 1.09–1.21). Older age and lower lung function were linked to more dyspnea. Current smoking was also linked, but no longer statistically significantly after accounting for lung function. Historical asbestos exposure, assigned from job history, was not statistically significantly associated with dyspnea. Current exposure to respiratory irritants was linked to less dyspnea, possibly a healthy worker effect, the authors suggested. The authors concluded that dyspnea reporting increased over time and both methods were suitable.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Longitudinal analysis of previously collected surveillance data (repeated visits), comparing two statistical modeling methods (a group-based trajectory model and a generalized linear mixed model)
• Who was studied: Marine transportation workers (vehicle and passenger ferries) in British Columbia, Canada. Of 1,701 participants, 925 who attended at least 2 visits were analyzed (824 men, 101 women). Enrollment was in 1987 or 1989, with follow-up visits in 1991, 1994 and 1999.
• Exposure: Historical asbestos exposure was assigned from job history. Workers who had worked at the maintenance dock, in the engine room or on live-aboard vessels were classed as exposed. No fiber type is named. Current exposure to respiratory irritants and smoking were also studied. The surveillance arose from concern about low-level asbestos exposure.
• Main results:
  • At baseline (1988), dyspnea prevalence was 15.6% among men and 27.3% among women.
  • At final follow-up, dyspnea prevalence was 28.8% in men and 43.8% in women.
  • The trajectory model found two groups. Group 1 (73%) had a steady low probability of reporting dyspnea. Group 2 (27%) had a steadily increasing probability.
  • Among Group 2 members, the authors reported that the probability of reporting dyspnea increased by 15% per year of follow-up (OR 1.15, 95% CI 1.09–1.21).
  • In the mixed model covering the whole population, visit year was positively associated with reporting dyspnea (OR 1.03, 1.00–1.07).
  • In the mixed model, age (OR 1.05, 1.04–1.07) and being female (OR 1.61, 1.03–2.51) were associated with reporting dyspnea over time.
  • In the mixed model, current smoking (compared with never smoking) was positively associated with dyspnea (OR 1.61, 1.12–2.31). Former smoking was not.
  • After adjustment for lung function, current smoking was no longer statistically significantly associated with dyspnea. In the mixed model, the adjusted odds ratio was 1.26 (0.87–1.83).
  • In the mixed model, higher lung function (FEV1 percent predicted) was associated with less dyspnea reporting (OR 0.96, 0.95–0.97).
  • In the trajectory model, historical asbestos exposure had an odds ratio of 1.54 (0.94–2.51) for membership in Group 2, the group with increasing dyspnea.
  • The authors reported that historical asbestos exposure did not reach statistical significance as a predictor of Group 2 membership (P > 0.05). It did not remain in the mixed model.
  • In the mixed model, current exposure to respiratory irritants was associated with less dyspnea reporting (OR 0.66, 0.50–0.88).
  • The authors suggested that the link between irritant exposure and less dyspnea may reflect a healthy worker effect, in which workers with symptoms move to less exposed jobs or leave work.
• The authors' conclusion: The authors concluded that the probability of reporting dyspnea increased over time in this population. They also concluded that both statistical methods are suitable for analyzing unbalanced repeated yes/no measures.
• Limitations the authors state:
  • There were few women compared with men.
  • Predictors other than age and sex did not reach significance in women, which the authors attributed likely to small sample size.
  • The estimation method used in the mixed models (penalized quasi-likelihood) may produce biased estimates. The authors re-ran the analyses with another method, got similar results, and judged the biases likely negligible.

Topics

Medical terms in this study, explained in our Lexicon: Dyspnea, Epidemiology.

Related studies

• Longitudinal pattern of reported respiratory symptoms and accelerated ventilatory loss in asbestos-exposed workers — C.A. Brodkin, S. Barnhart, et al. (1996)
• Correlation between respiratory symptoms and pulmonary function in asbestos-exposed workers — C.A. Brodkin, S. Barnhart, et al. (1993)
• Radiographic abnormalities in asbestos insulators: effects of duration from onset of exposure and smoking. Relationships of dyspnea with parenchymal and pleural fibrosis — R. Lilis, A. Miller, et al. (1991)
• Malignant peritoneal mesothelioma presenting with respiratory symptoms — M. Daskalogiannaki, P. Prassopoulos, et al. (2000)
• Abnormal wasted ventilation fraction and normal pulmonary hemodynamics during exercise in patients with exertional dyspnea — Z. Mohsenifar, H.V. Brown, et al. (1982)
• Exposure to mineral oil mist and respiratory symptoms in marine engineers — K. Svendsen, B. Hilt (1997)

Browse more research in International Archives of Occupational and Environmental Health, or return to medical research.