UNGD & Depression

Pennsylvania is one of the country’s leading producers of natural gas. Natural gas extraction is an industrial process, and as such may be linked to environmental and social impacts. Geisinger Health, an eastern-Pennsylvania based healthcare provider, decided to study ‘suspected’ influences natural gas development may have on mental health and sleep via reduced life satisfaction, feelings of disempowerment, social stress, negative psychological states, air pollution, environmental hazard, nighttime noise, and light pollution, etc. They conducted a study based on validated statistical analyses to evaluate the association of UNGD with mental health and sleep as measured by the Patient Health Questionnaire-8 (PHQ-8) and electronic health record (EHR) data among Geisinger adult primary care patients in Pennsylvania. While EHR data was obtained between 2009 and 2015, the PHQ-8 was conducted between 2014 and 2015. A UNGD metric was developed to categorize the UNGD activity (very low, low, medium and high) of each participant based on their residential proximity to UNGD wells, the well sizes, and the well activities.

There were 4762 participants in the PHQ-8 survey. The answer to each question in PHQ-8 carried a score of 0 to 3. Depression was determined based on the total score of the PHQ-8 for each participant. Depression symptoms were categorized into four levels based on the total PHQ-8 score (no significant, mild, moderate, and moderately severe to severe). Multinomial logistic models and negative binomial models were used to estimate the association of UNGD with each level of depression symptoms (mild, moderate, moderately severe/severe) compared to no depression symptoms. In order to account for the effect of participants taking antidepressants, modifications were made to include cross-products of the UNGD variables and antidepressant medication use in the final depression symptom models.

Disordered sleep diagnoses (case-events) were identified from EHR using disordered sleep ICD-9 medical code and medications used. Out of 8578 identified disordered sleep diagnoses, 3868 diagnoses were selected for analysis. A survey-weighted generalized estimating equation model was used to account for multiple diagnoses within participants.

Based on the results from the analytical models that were weighted to account for sampling design and participation, the researchers concluded that there are associations between living closer to more and bigger wells and depression symptoms, but not disordered sleep diagnoses. It is suggested further research is required to disentangle the multi-factorial pathways through which UNGD may influence mental health. In the meantime, potential mental health consequences should be included in risk-benefit calculations for future and current UNGDs.

Contributed by: Tim Yeung