Li Martino Mansour 2022

To improve causal inference by reducing reporting bias and residual confounding from unobserved time-invariant confounders, FE regressions were also employed.²⁶ This modeling approach explores the relationship between exposure and outcome within people over time and removes the effect of time-invariant characteristics of residents that may impact or bias the causal effect of noise exposure on mental health.

Translation:

Fixed-effects regression model was used to aid in determining the cause-and-effect relationship between environmental noise exposure and mental health, as it minimizes the effects of potential bias from self-reporting and confounding variables.

This study aims to evaluate the inequalities in the distribution of perceived noise exposure in a residential context and causally test and quantify the negative impact of reported noise exposure on mental health and psychological distress in a population-based longitudinal cohort of Australians spanning the first 2 decades of this century.

Translation:

The study seeks to investigate the differences in environmental noise perceived by individuals living in varying residential settings and to determine and quantify the cause-and-effect relationship between reported noise exposure and mental health in a population-based cohort of Australians in the last twenty years.

The results on the impacts of changes in noise exposure show that adjusting for sociodemographic characteristics and occurrence of confounding life events, mental health declined with increased noise exposure (change from reporting no/rare to common/very common exposure) in the order of −0.27 (95% CI: −0.57, 0.02) and improved with decreased noise exposure (a change from reporting common/very common to no/rare exposure) in order of 0.48 (95% CI: 0.19, 0.77) (Appendix Table 3, available online).

Translation:

After considering sociodemographic factors and life events, the results on how changes in noise exposure affect mental health show mental health decline with increased noise exposure and incline with decreased noise exposure.  A change from no/rare to common/very common noise exposure reveal that mental health worsened by -0.27 (95% Cl: -0.57, 0.02).  On the other hand, a change from common/very common to no/rare noise exposure reveal that mental health improved by 0.48 (95% Cl: 0.19, 0.77) (Appendix Table 3, available online).

These results contribute to the understanding of the health-harming impacts of residential environmental noise with a large-scale longitudinal national survey for more than 20 years and a causally focused approach to overcoming the high bias risk inherent in previous cross-sectional studies.

Translation:

The results of this extensive and long-term study enhance our understanding of the health-damaging effects of environmental noise exposure in residential settings.  By adopting a cause-and-effect analysis approach, the high bias risk associated with measurement error, confounding variables, and reverse causation present in previous studies have been addressed.

The study relied on self-reported noise exposure which varies greatly from individual to individual, depending on their perception of noise.  Future research should focus on producing more precise and more valid results, perhaps by using an instrument, like a sound level meter, to measure the noise level.

I did not fully understand the authors’ approach in establishing a cause-and-effect relationship between noise exposure and mental health.  Personally, I believe the authors achieved showing an association between noise exposure and mental health but not a direct causal relationship.