We read with great interest the article by Shabil et al. entitled “Association of electronic cigarette use and risk of COPD: a systematic review and meta-analysis”1. The authors pooled 17 observational studies and concluded that e-cigarette use (current, former, or ever) is associated with a higher risk of COPD. While we acknowledge the importance of the topic and the effort invested in this review, we have significant concerns regarding the methodology, interpretation, and framing of the findings, which risk overstating the evidence and misleading readers.

The primary limitation is inadequate adjustment for cigarette smoking, the predominant cause of COPD. The majority of e-cigarette users are either current or former smokers; without careful adjustment for detailed smoking history (e.g., pack-years, age at initiation, years since cessation), any observed association is highly likely to reflect the effects of combustible cigarette use rather than e-cigarette use per se. Indeed, prior work2 shows that pack-years and cessation timing are critical determinants of COPD risk. These variables were rarely considered in the included studies, and the meta-analysis does not stratify or adjust accordingly. Moreover, the absence of a direct comparison to cigarette smokers is a major missed opportunity. Public health authorities in the UK and US acknowledge that many smokers transition to e-cigarettes to reduce harm. Without benchmarking the risk of e-cigarette use against that of cigarette smoking, the analysis cannot inform in a meaningful or balanced way.

A second critical concern is the lack of information on the timing of COPD diagnosis relative to the initiation of e-cigarette use. As demonstrated by Rodu & Plurphanswat3, cross-sectional associations are often driven by smokers with pre-existing COPD who switch to e-cigarettes after diagnosis. This reverse causation produces a spurious association in which e-cigarette use appears linked to COPD, when in fact COPD preceded the exposure.

Shabil et al. acknowledge that the pooled odds ratio is smaller and statistically non-significant in longitudinal cohort studies (OR = 1.145; 95% CI: 0.842–1.557) compared with cross-sectional studies. This is a crucial finding, as cohort designs better address temporality and causality, whereas cross-sectional studies are intrinsically vulnerable to bias. For this reason, we believe the main conclusions should rest on longitudinal evidence. A better alternative is to focus on never-smokers. There is no evidence that e-cigarette use is directly associated with risk of COPD among never-smokers4.

The reported pooled OR for former e-cigarette use (1.84) exceeds that for current use (1.48), and the confidence intervals do not overlap indicating a statistically significant difference. This is biologically implausible if e-cigarette use were causal, since ongoing exposure would be expected to confer equal or greater risk. The pattern is, however, entirely consistent with residual confounding by cigarette smoking: many “former” e-cigarette users may have returned to smoking or never ceased smoking in the first place.

Many of the included studies have methodological weaknesses that undermine the reliability of the pooled results. Most relied on self-reported COPD diagnoses, which are prone to recall bias and misclassification. The review also omits a detailed risk-of-bias assessment (an essential component given the well-documented shortcomings in this field). Several studies draw on the same large datasets (e.g., PATH), increasing the likelihood of overlapping participants and inflating the apparent sample size. Such double-counting can distort precision estimates and misrepresent the true breadth of the evidence. The meta-analysis should have identified, quantified, and adjusted for this overlap. In addition, definitions of “current,” “former,” and “ever” e-cigarette use varied widely, and the extent of dual use with cigarettes was often unclear, making pooled estimates potentially misleading. Finally, the findings diverge from those of a recent umbrella review (the highest level of evidence) which found no significant short- or medium-term changes in respiratory function for any pattern of e-cigarette use5.

In conclusion, residual confounding, reverse causation, biologically implausible patterns, and data overlap substantially weaken the reported associations. Given these limitations, the evidence does not support a causal link between e-cigarette use and COPD. At most, it shows that many COPD patients have used e-cigarettes, a finding more plausibly explained by smoking history and switching behaviour than by e-cigarette harms. Drawing causal conclusions from such data is akin to claiming nicotine replacement therapy causes COPD because many patients have used it, ignoring confounding and temporality, and risking misinformation to clinicians, policymakers, and the public. These issues, evident in Shabil et al. and widespread in the field6, call for greater reliance on longitudinal data, explicit control for confounding, reverse causation, and data overlap, and more cautious, hypothesis-generating conclusions.