What the study found
The study found that categorical risk estimates from studies with different exposure groupings can be converted into per-unit continuous effect estimates and then combined in meta-analysis. The authors report that the method appeared to give an unbiased estimate of the true continuous effect in their example.
Why the authors say this matters
The authors say this matters because environmental epidemiology often has studies that use different exposure categories, which makes quantitative synthesis difficult. They conclude that a practical and replicable method like this can improve comparability and support meta-analyses, including in evidence bases that are limited.
What the researchers tested
The researchers presented a three-step approach: estimate the midpoint of each exposure category, derive category-specific per-unit continuous beta coefficients from categorical effect estimates, and then calculate a study-specific continuous effect estimate with uncertainty intervals. They illustrated the method using data from Mataloni et al. (2016) to estimate the hazard risk associated with a 1 ng/m³ increase in hydrogen sulfide (H2S) exposure, and compared robustness against the true continuous effect and previously published pooling methods.
What worked and what didn't
The method appeared to provide an unbiased estimate of the true continuous effect. The results were comparable with more complex methods in both point and interval estimates, and the method was reported to be robust to different definitions of central category exposures.
What to keep in mind
The method assumes that risk is linear within categories and that category-specific estimates are independent. The abstract does not describe limitations beyond these assumptions, and the practical example was limited to the dataset used for illustration.
- The method converts categorical exposure risk estimates into per-unit continuous effect estimates.
- The approach is meant to help meta-analyses when studies use different exposure intervals.
- In the example using H2S exposure, the method appeared unbiased relative to the true continuous effect.
- Results were comparable with more complex pooling methods for both point and interval estimates.
- The method assumes linearity within categories and independence of category-specific estimates.

