Reply

To The Editor—In their Letter to the Editor, Feng et al [1] stated that they read our article [2] with great interest and described 5 concerns they had with our report. We welcome the opportunity to respond.

As their first concern, they suggest that we should have examined viral load, which is indeed a biomarker of cervical lesion progression [3] and one that we have examined in the context of sexual transmission using the same cohort [4]. In our Journal of Infectious Diseases article [2], we explored many different approaches to reveal a possible association between serum 25-hydroxyvitamin D levels and human papillomavirus (HPV) infection, such as by grouped genotype and infection outcome definition (prevalence, incidence, and clearance). We also accounted for different ways of characterizing 25-hydroxyvitamin D levels, that is, by concentration and clinical threshold. Despite the multiplicity of comparisons, we did not find any evidence for an association. Stratifying by viral load would have amounted to data dredging with no biological rationale or empirical evidence. Feng et al [1] may have misunderstood our study as one that included HPV infections with clinical manifestations. Our cohort included only asymptomatic young women, as explained in the article [2].

With respect to their second concern, we are aware that high viral load infections tend to take longer to clear, based also on our own empirical research [5]. However, we fail to see why stratifying a null association by viral load status would have revealed anything useful. If vitamin D were associated with HPV infection, it would have initially manifested via detectability of the virus. Again, our in-depth analyses by genotype and outcome did not reveal any credible dose-response relationship, despite the multiple ways that we treated exposure and outcome variables.

Their third concern stems from another apparent misunderstanding of our study methods. Our subject sample included only asymptomatic women [2]. A history of prior cervical HPV disease would have made a participant ineligible. Our study was not conducted in a clinic for sexually transmitted diseases. None of our subjects were treated for HPV disease because they did not have one. Moreover, there is no treatment for asymptomatic HPV infection, that is, in the absence of associate lesions. For the sake of argument, however, Feng et al’s [1] contention is irrelevant to the vitamin D-HPV relationship that we examined. The expectation that adjusting for treatment would have changed our clearance analysis by vitamin D has no biological or epidemiologic underpinnings.

Their fourth concern is also unfounded, because we administered follow-up questionnaires and did not observe an association between positivity being disclosed and clearance (data not shown). Again, we fail to see why that would be related to the vitamin D hypothesis. Feng et al [1] missed our reference to the Human Papillomavirus Infection and Transmission Among Couples Through Heterosexual Activity (HITCH) Cohort Study design paper, in which many details are given about the methodology [6].

Their fifth and final concern also appears incorrect. We described in our article the commercial assay we used for serum 25-hydroxyvitamin D levels [7].

Notes

Disclaimer. The funders played no role in study design, data collection and analysis, preparation of the manuscript, or the decision to publish.

Financial support. The Human Papillomavirus Infection and Transmission Among Couples Through Heterosexual Activity (HITCH) Cohort Study was funded by the Canadian Institutes of Health Research (Grant MOP-68893 and Team Grant CRN-83320; to E. L. F.) and the US National Institutes of Health (Grant RO1AI073889; to E. L. F.). This work was also funded by the Fonds de recherche du Québec - Santé (FRQS) Réseau sida et maladies infectieuses (SIDA-MI) (support for human papillomavirus testing) and by Merck-Frosst Canada and Merck & Co.

Potential conflicts of interest. E. L. F. reports grants and personal fees from Merck outside of the submitted work. M. Z. and E. L. F. hold a patent related to the discovery “DNA methylation markers for early detection of cervical cancer”, registered at the Office of Innovation and Partnerships, McGill University, Montreal, Quebec, Canada (October 2018). All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.