We read with considerable interest the study conducted by Jiang et al. 1, which explored the longitudinal patterns of media exposure and cognitive functioning in older Chinese adults, along with the determinants shaping these developmental trajectories. The study employed a longitudinal design to determine whether older adults with greater media exposure at baseline exhibited superior initial cognitive performance, and whether temporal variations in media exposure corresponded to changes in cognitive functioning. Media exposure indices and cognitive outcomes were measured over three assessment waves (2016, 2018, and 2020) among Chinese individuals aged 60 years and older, and their concurrent developmental trajectories were evaluated. Conducted in a large sample (n = 7625), this well-designed and clearly presented study fills an important gap in the literature. We would like to offer several considerations regarding the study. First, the reliance on self-reported data may represent a significant confounding factor affecting the results, as also acknowledged by the authors in the limitations section. Additionally, the observed decline in media exposure over time may partly reflect reduced social support, which could limit access to media resources. Alternatively, the development of secondary comorbidities or new medical conditions may have led to declines in Activities of Daily Living (ADL) or Instrumental Activities of Daily Living (IADL) scores, thereby indirectly contributing to decreased media exposure. Another important consideration is that multiple clinical factors may have influenced changes in cognitive function over time. The presence of comorbid conditions—such as diabetes or vascular diseases—overall comorbidity burden, and hospitalization history may all act as confounding variables affecting the study outcomes. A meta-analysis published in 2024 2 demonstrated that the presence of diabetes, having two or more comorbidities, and physical frailty each independently increased the risk of developing mild cognitive impairment (MCI) in older adults. Moreover, a meta-analysis published in 2025 3, encompassing 41 studies with a combined sample of 138 030 participants, indicated that the overall pooled prevalence of cognitive impairment (CI) among individuals with chronic obstructive pulmonary disease (COPD) was approximately 20%–30% (95% CI, 0.17–0.28). In the same analysis, the pooled prevalence of MCI in this population was reported as 24% (95% CI, 0.17–0.32). In the study, baseline (2016) depressive symptoms were adjusted for in the model, and a minimal sensitivity analysis showed that the direction and magnitude of the main path coefficients between media exposure and cognitive growth factors remained largely unchanged. However, it should also be considered that the emergence of a new major depressive disorder (MDD) during follow-up—despite the absence of depressive symptoms at baseline—may negatively affect both cognitive performance and media exposure. Cognitive impairment has long been acknowledged as a fundamental characteristic of MDD. In a recent investigation, Ou et al. 4 further highlighted the link between cognitive deterioration and MDD, reporting that individuals with untreated MDD showed marked deficits in processing speed as well as visuospatial memory and learning. Additionally, although to a lesser degree, impairments were also observed in executive functioning and verbal memory and learning. Lastly, although the current study aimed to underscore the potentially protective influence of media exposure on cognitive performance, additional investigations are needed to elucidate the possible cognitive consequences associated with excessive or long-term media use. In conclusion, we acknowledge the significant contribution of this well-conducted study to the literature. Nevertheless, we believe that consideration of the above-mentioned factors would further strengthen the interpretation of the findings. The authors have nothing to report. The authors declare no conflicts of interest. Data sharing not applicable to this article as no datasets were generated or analysed during the current study.
Kars et al. (2026) studied this question.