We thank Haoxuan Yan for the thoughtful comments regarding our study investigating dietary vitamin B intake and incident open-angle glaucoma (iOAG).
Indeed, dietary habits may change over time, and a single baseline food-frequency questionnaire cannot fully capture long-term variation in nutrient intake. This limitation was explicitly addressed in our original study. All reported associations remained consistent across cumulative follow-up periods, arguing against reverse causality. Excluding participants with less than 5 years of follow-up yielded similar associations (niacin: odds ratio 95% confidence interval [or CI] = 0.94 [0.90-0.98] per mg/d; cobalamin: odds ratio [95% CI] = 0.88 [0.81-0.96] per µg/d).
Second, we agree that the biological mechanisms underlying the observed association between cobalamin intake and iOAG remain incompletely understood. Serum homocysteine was not associated with iOAG ( P =.59), and we therefore found no evidence of mediation ( P =.73). The observed association between cobalamin and iOAG is therefore unlikely to be explained through homocysteine-related pathways and may be explained by alternative mechanisms, including direct neuroprotective effects, maintenance of myelin integrity, or preservation of retinal ganglion cell health. Consistent with the latter hypothesis, higher cobalamin intake was associated with greater ganglion cell layer thickness ( P trend =.002).
Third, regarding the interaction between vitamin B intake and genetic risk, we used a validated glaucoma genetic risk score derived from established OAG-associated variants that reflect multiple biological pathways and are not exclusively related to intraocular pressure (IOP). We agree that formal interaction testing complements stratified analyses and therefore evaluated multiplicative interaction terms between genetic risk and both niacin ( P interaction =.20) and cobalamin ( P interaction =.46) intake. Although our data do not provide strong statistical evidence for effect modification by genetic risk, nonsignificant interaction terms do not preclude interpretation of consistent subgroup-specific patterns or the underlying main effects. Accordingly, although we refrain from drawing firm conclusions regarding effect modification, the observed patterns across genetic risk strata may warrant further investigation in larger studies.
Fourth, we acknowledge that estimating untreated IOP in participants receiving IOP-lowering treatment requires assumptions. We therefore applied a commonly used correction factor to approximate pretreatment IOP levels. Restricting analyses to untreated individuals would preferentially exclude participants with elevated IOP who required treatment and are therefore particularly relevant to address the study questions. Moreover, adjustment for IOP was included as an exploratory model to assess the extent to which the observed associations may operate independently of IOP, rather than as a primary analysis.
Finally, our analyses were hypothesis driven and focused on predefined vitamin B and biologically related glaucoma outcomes. These outcomes represent clinical, pressure-related, and structural features of the glaucomatous disease process rather than independent tests. Therefore, a simple Bonferroni correction across all outcomes and models would likely be overly conservative. Importantly, the primary associations for niacin and cobalamin with iOAG were consistent across models and sensitivity analyses. Nonetheless, even under a conservative Bonferroni correction for the 6 exposures evaluated (significance threshold P <.008), the principal associations of niacin and cobalamin intake with iOAG would remain statistically significant.
We appreciate the opportunity to clarify these points and agree that further studies are needed to elucidate the biological mechanisms linking vitamin B intake and glaucoma risk.
See the original article for any disclosures of the authors.
CRediT authorship contribution statement
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
