I read with great interest the methodological systematic review by Duranté et al, which offers a comprehensive overview of biological research on non-physical pain.^{Reference Duranté, Ribeiro, Gaspard-Boulinc, Boutron, Henry and Petit1} The authors highlight several important gaps in the literature, and I would like to expand on three methodological issues that we believe require urgent attention in order to strengthen the field’s translational potential.

First, the heavy reliance on healthy volunteers (71.7% of all included studies) presents a substantial limitation. Although healthy samples offer practical advantages, this approach restricts the generalisability of biological findings to clinical populations who experience mental, social or emotional pain with greater severity and complexity. Biomarkers derived from individuals without psychiatric, neurological or psychosocial burden may not adequately capture the neurobiological signatures of distress found in patients with depression,^{Reference Fillingim, Tanguay-Sabourin, Parisien, Zare, Guglietti and Norman2,Reference Misiak, Samochowiec, Kowalski, Gaebel, Bassetti and Chan3} trauma histories, suicidality or chronic disease. Future studies should prioritise clinically relevant samples or adopt stratified recruitment strategies that reflect diverse neuropsychological profiles.

Second, the review’s finding that 32.4% of studies did not adjust for key confounders – such as baseline mood, chronic stress, social sensitivity or past adversity – raises concern regarding potential spurious associations.^{Reference Michels, McDade, Krause, Miller and Giletta4,Reference Qin, Xu, Qin, Shao, Ma and Ou5} Because non-physical pain is inherently multidimensional, failing to account for these variables risks overestimating or misattributing biological effects. I support the authors’ call for improved methodological rigour and propose the development of a preregistered confounder framework to standardise reporting and analytic procedures across future studies.

Third, the limited number of rodent studies (n = 4) underscores a significant translational gap. Animal models have historically been crucial in elucidating mechanisms underlying physical pain,^{Reference Ferrari and Taylor6} yet analogous progress for non-physical pain remains elusive. The scarcity of preclinical research limits the field’s ability to determine whether observed neural patterns represent species-general mechanisms or context-dependent responses specific to experimental paradigms in humans. A coordinated effort to refine and validate animal models of social or affective distress is needed to advance mechanistic understanding.

In conclusion, the review by Duranté et al provides an essential foundation for future improvements in the biological study of non-physical pain. Addressing the challenges of sample selection, confounder control and cross-species modelling will be critical for developing robust, clinically meaningful biomarkers in this evolving area of research.