Mental health indicators and lung function following a large oil spill
- Kaitlyn G. Lawrence1,
- Emily J. Werder1,
- Richard K. Kwok1,2,
- Lawrence S. Engel1,3 and
- Dale P. Sandler1⇑
- 1Epidemiology Branch, National Institute of Environmental Health Sciences, Research Triangle Park, NC, USA
- 2Office of the Director, National Institute of Environmental Health Sciences, Bethesda, MD, USA
- 3Dept of Epidemiology, Gillings School of Global Public Health, University of North Carolina, Chapel Hill, NC, USA
- Dale P. Sandler (sandler{at}niehs.nih.gov)
Abstract
Among GuLF Study participants, findings indicate better lung function associated with severe depression and generalised anxiety, and no association between lung function and PTSD symptoms or perceived stress https://bit.ly/3xpLNRB
To the Editor:
Natural and anthropogenic disaster-related experiences can exacerbate background stressors, which may lead to negative mental health outcomes [1]. However, few studies have assessed mental health indicators related to measured lung function in the aftermath of a disaster. In 2010 the Deepwater Horizon disaster resulted in the largest maritime oil spill in US history and negatively affected proximal communities, which experienced increased distress and post-traumatic stress disorder (PTSD)-like symptoms attributed to the disaster [2, 3]. We investigated the relationship between mental health indicators and lung function following this disaster.
We used cross-sectional data from the Gulf Long-Term Follow-up Study (GuLF Study), a prospective cohort of adults (n=32 608) who participated in response and cleanup activities following the 2010 Deepwater Horizon oil spill, and others who received worker safety training but were not hired [4]. Enrolled participants completed telephone interviews between March 2011 and March 2013. Participants who were English- or Spanish-speaking residents of Alabama, Florida, Louisiana, Mississippi or Texas (n=25 304) were eligible for a home visit exam. 11 193 eligible participants completed a home visit exam (May 2011 to May 2013) and, of these, 10 040 completed spirometry testing. We analysed spirometry data from participants with spirometry tests meeting American Thoracic Society/European Respiratory Society (ATS/ERS) quality standards [5]. All home visit participants provided written informed consent. This study was approved by the National Institute of Environmental Health Sciences institutional review board.
Participants completed psychological assessments on multiple mental health endpoints using validated questionnaires. Recommended standardised cut-off points were used to classify participants. Perceived stress was measured at enrolment using Cohen's perceived stress scale and being positive for stress was defined as having a score ≥9 (range 0–16) [6]. The remaining tests were measured at the home visit. Depression was measured using the eight-item Patient Health Questionnaire depression scale (PHQ-8) tool and defined as having a score ≥10 (range 0–27) [7]. Generalised anxiety disorder was measured using the 7-item Generalized Anxiety Disorder scale (GAD-7) and defined as having a score ≥10 (range 0–21) [8]. Finally, PTSD was measured using the Primary Care PTSD (PC-PTSD) screener and defined as having a score ≥3 (range 0–4) [9].
Home visit participants performed spirometry using an ultrasonic spirometer (EasyOne; ndd Technologies, Andover, MA, USA) administered by trained examiners. Lung function outcomes of interest included forced expiratory volume in 1 s (FEV1; mL), forced vital capacity (FVC; mL) and their ratio (FEV1/FVC; %). Tests for 10 019 participants (workers and non-workers) were overread by a spirometry expert for quality control and 7487 met the ATS/ERS within- and between-manoeuvre acceptability criteria, defined as at least three acceptable manoeuvres, with variability of either FEV1 or FVC ≤150 mL or equivalent quality deemed by a spirometry over-reader expert (defined as having a score of either A, B or C) [5]. After excluding participants with missing information on any mental health outcome (n=833) and excluding participants with missing covariate information (n=858; 49% of which were missing income data) the final analysis sample was composed of 5796 participants.
Information on demographic, socioeconomic, lifestyle and health measures was collected at enrolment using a structured telephone interview. Anthropometric measures, including height and weight, were obtained by examiners at the home visit. Potential confounders were selected using a directed acyclic graph and included age at home visit (years), sex (male, female), height at home visit (cm), weight (lbs), race (white, black, other), ethnicity (Hispanic, non-Hispanic), smoking status (current heavy, current light, former, never), and secondhand smoke exposure (yes, no), education level (less than high school/equivalent, high school diploma/gED, some college/2 year degree, 4 year college graduate or more), and income level (<USD 20 000, USD 20 000–50 000, >USD 50 000). Height squared (cm2) was included in models based on its reported quadratic relationship to lung function measures [10]. We also controlled for pre-existing lung disease (doctor's diagnosis of bronchitis, emphysema or asthma; yes, no) and oil spill response and cleanup work hierarchical job class (response work, operations, decontamination, cleanup on water, cleanup on land, non-worker). FEV1 and FVC were statistically significantly associated with height, height squared, age and comorbid lung disease (p<0.05) in our data, providing further support for inclusion in the adjustment set and support for the validity of our spirometry results. Finally, we adjusted for area deprivation using a published index linked to participant home visit addresses [11].
We evaluated the relationship of each mental health outcome (PTSD, depression, stress and anxiety) to the lung function measures. We modelled mental health outcomes separately, treating each as an independent variable and adjusting for age, sex, race/ethnicity, height (cm), height squared (cm2), weight, smoking, secondhand smoke, income, education, neighbourhood deprivation, pre-existing lung disease, and oil spill response and cleanup work jobs in each model. All analyses were conducted using SAS v 9.4 (SAS Institute, Cary, NC, USA).
Perceived stress and PTSD symptoms were not associated with any lung function measure (table 1). Compared to those without depression, those with severe depression scores had a higher FEV1 (β 61 mL, 95% CI 21, 101) and FVC (β 62 mL, 95% CI 16, 109), but not FEV1/FVC ratio. Participants with severe generalised anxiety disorder scores versus those without also had higher FEV1 (β 78 mL, 95% CI 44, 113), FVC (β 71 mL, 95% CI 31, 112), and FEV1/FVC ratio (β 0.56%, 95% CI 0.15, 0.97).
We examined the relationship between adverse mental health measures and lung function in a large population of adults living in US Gulf states enrolled in a cohort following the 2010 Deepwater Horizon disaster. Despite findings of adverse mental health outcomes associated with response and cleanup work following the Deepwater Horizon disaster [2], results here indicate better lung function associated with severe depression and generalised anxiety.
Our results are discordant with the existing literature on depression and anxiety, which has reported links between mental health and/or stress and poor respiratory health. Psychological stress can lead to hyperresponsiveness of the hypothalamic pituitary axis, increasing secretion of inflammatory cytokines and the release of cortisol that may trigger hormones (e.g. corticotrophin) [12]. This can potentiate the immune system, or autonomic control, resulting in bronchoconstriction [13]. Higher depression scores have been associated with lower FEV1 [14] while war-related stressors have been associated with increased risk of asthma [15]. It is unclear why we observed better lung function for those with depression and generalised anxiety. Results may be attributable to limited generalisability, uncontrolled confounding, such as lack of data on clinical management of stress, or limitations of cross-sectional analyses. It is possible that selection bias could have affected study results. However, we assessed the prevalence of adverse mental health outcomes among excluded home visit participants and found that the prevalence of each outcome was very similar to that in the analytic sample: depression (15% included versus 15% excluded); anxiety (22% included versus 24% excluded); stress (26% among included versus 29% excluded); PTSD (5% included versus 6% excluded) so we do not expect large differences in results due to these exclusions. Findings for PTSD symptoms agree with a prior study that found no association between PTSD and lung function measures, although the authors did find an association with asthma [16]. In that study, traumatic stress was associated with modestly lower FEV1 and FEV1/FVC ratio measures. Given the dearth of information on specific mental health disorders and adverse respiratory health and inconsistent results across studies, prospective studies are needed to clarify this relationship.
Shareable PDF
Supplementary Material
This one-page PDF can be shared freely online.
Shareable PDF ERJ-00712-2021.Shareable
Acknowledgements
Thank you to Symielle Gaston and Thanh Hoang for reviewing an earlier draft of this manuscript.
Footnotes
Code: The data are not available for replication.
Author contributions: D.P. Sandler, L.S. Engel, and R.K. Kwok collected the data for this study. K.G. Lawrence, D.P. Sandler, L.S. Engel, R.K. Kwok conceptualised and designed the study. K.G. Lawrence conducted the statistical analysis and led the interpretation of the data. K.G. Lawrence drafted the manuscript. D.P. Sandler, E.J. Werder, L.S. Engel, and R.K. Kwok provided scientific review and contributed to the interpretation of findings and final manuscript content.
Conflict of interest: K.G. Lawrence has nothing to disclose.
Conflict of interest: E.J. Werder has nothing to disclose.
Conflict of interest: R.K. Kwok has nothing to disclose.
Conflict of interest: L.S. Engel has nothing to disclose.
Conflict of interest: D.P. Sandler has nothing to disclose.
Support statement: This research was funded by the Intramural Research Program of the NIH, the National Institute of Environmental Health Sciences (Z01 ES 102945). Funding information for this article has been deposited with the Crossref Funder Registry.
- Received March 10, 2021.
- Accepted July 3, 2021.
- The content of this work is not subject to copyright. Design and branding are copyright ©ERS 2021. For commercial reproduction rights and permissions contact permissions{at}ersnet.org