The research, published in BMC Public Health, is a prospective cohort study led by Madleen Orumaa of the Cancer Registry of Norway together with collaborators from the Danish Cancer Institute, Karolinska Institutet in Sweden, and the Icelandic Cancer Registry. The team analyzed data from 117,511 women aged 18 to 45 in Norway, Sweden, Denmark and Iceland who completed detailed questionnaires about cervical cancer risk factors, including sexual history, contraceptive use, smoking, and prior sexually transmitted infections. The women were then followed for a mean of 9.5 years through national health registries, allowing the researchers to link self-reported lifestyle information collected before diagnosis with subsequent outcomes.

During follow-up, the cohort accumulated a meaningful set of endpoints: 2,926 women were diagnosed with cervical intraepithelial neoplasia grade 3, or CIN3, the principal squamous precursor lesion; 174 developed adenocarcinoma in situ, the glandular equivalent; 66 were diagnosed with invasive squamous cell carcinoma; and 23 with invasive adenocarcinoma. Because glandular lesions are comparatively rare, the absolute numbers are small, and the authors themselves frame the findings as an important first step rather than a definitive answer. Even so, the prospective design — recording exposures before any diagnosis — gives the associations considerably more credibility than retrospective case-control studies, which are vulnerable to recall bias after a cancer diagnosis.

The statistical analysis used multivariate Cox regression, a technique that estimates hazard ratios for each exposure while adjusting for all the others, thereby isolating the independent contribution of each lifestyle factor. The results split cleanly along histological lines. For the squamous pathway — CIN3 and squamous cell carcinoma combined — four factors emerged as independently associated with risk. Each additional lifetime sexual partner raised the hazard by roughly 47 percent, with an adjusted hazard ratio of 1.47 and a 95 percent confidence interval of 1.34 to 1.56. A reported history of chlamydia infection carried a hazard ratio of 1.43 (95 percent CI: 1.31 to 1.56). Use of hormonal contraceptives was associated with a hazard ratio of 1.23 (95 percent CI: 1.02 to 1.47), and smoking added a small but measurable increment of 2 percent per pack-year (hazard ratio 1.02, 95 percent CI: 1.02 to 1.03).

The glandular pathway told a different story. Women who went on to develop adenocarcinoma in situ or invasive adenocarcinoma did share some baseline characteristics with the squamous group — both groups reported an earlier age at first intercourse and a greater number of lifetime sexual partners compared with women who remained unaffected. But in the fully adjusted models, one factor stood out as specific to the glandular lesions: a history of genital warts, which are caused by low-risk HPV types 6 and 11, was associated with a hazard ratio of 1.60 (95 percent CI: 1.15 to 2.24) for AIS/AC. Notably, this association did not hold for the squamous lesions, suggesting that a visible history of warts may serve as a marker of a distinct pattern of HPV exposure or of individual susceptibility relevant to glandular transformation.

From a biological standpoint, the divergence is plausible. Squamous cells on the ectocervix and glandular cells in the endocervical canal are different cell types with different vulnerabilities to HPV-mediated transformation, and squamous and adenocarcinomas differ in their genomic profiles and their responses to treatment. The transformation zone, where squamous and glandular epithelia meet, is the epicenter of HPV-driven carcinogenesis, but the mechanisms by which co-factors such as smoking, hormonal exposure, and co-infections accelerate progression may not act identically in both cell lineages. Smoking, for example, is thought to act partly through local immunosuppression and the delivery of carcinogenic metabolites in cervical mucus, while hormonal contraceptives have been hypothesized to influence gene expression in HPV-infected cells. Chlamydia, meanwhile, may contribute through chronic inflammation that facilitates HPV persistence.

The public health implications are considerable. The World Health Organization has set an ambitious global strategy to eliminate cervical cancer as a public health problem, built on three pillars: HPV vaccination, screening, and treatment of precancerous lesions. Most current screening programs were designed and validated in an era when squamous lesions dominated, and they rely primarily on cytology of the transformation zone. Glandular lesions, which sit higher in the endocervical canal and shed cells less readily, are notoriously under-detected by conventional Pap smears. The authors argue that their findings underscore the need for risk-based screening models — stratifying women by their individual risk profiles rather than applying a uniform interval to everyone — and for improving adherence to screening guidelines across both lesion types.

Such risk stratification could, in principle, incorporate the factors identified in this study. A woman with a history of chlamydia, extensive smoking exposure, and multiple sexual partners might warrant closer surveillance for squamous lesions, while a history of genital warts might flag elevated glandular risk. HPV vaccination, which protects against the high-risk types responsible for most cervical cancers and, in the quadrivalent and nonavalent vaccines, against the wart-causing types 6 and 11 as well, remains the most powerful preventive tool for both pathways. The study’s finding that genital warts mark glandular risk adds an intriguing dimension: vaccination against types 6 and 11 may confer benefits beyond preventing the warts themselves, potentially reducing a marker-associated risk of glandular neoplasia.

The authors are careful to note the limitations inherent in their design. The number of invasive adenocarcinoma cases — 23 — is small, which widens confidence intervals and limits the granularity of subgroup analyses. Lifestyle data were self-reported and therefore subject to misclassification, and the cohort comprised Nordic women aged 18 to 45 at enrollment, a relatively homogeneous, well-screened population whose findings may not generalize directly to settings with different screening coverage or HPV type distributions. The study was funded by Merck & Co. for questionnaire data collection and by the Nordic Cancer Union for manuscript preparation, with the sponsors having no role in data analysis or interpretation, according to the published disclosures.

Even with those caveats, the study represents one of the largest prospective efforts to disentangle the lifestyle epidemiology of the two cervical cancer pathways, and it arrives at a moment when adenocarcinoma’s rising share of cervical cancer diagnoses is drawing increasing attention worldwide. As vaccination programs mature and screening technologies shift toward HPV testing, understanding which co-factors drive which lesion type will be essential for designing the next generation of prevention strategies. The Nordic registries, with their near-complete follow-up, offer an unusually clear window onto that question — and this study suggests the window reveals two distinct diseases, not one.

Subject of Research: Lifestyle risk factors for glandular and squamous cervical neoplasia in a Nordic prospective cohort

Article Title: Lifestyle risk factors for cervical glandular and squamous cell neoplasia: insights from a prospective cohort study of over 117,000 women

Article References: Lifestyle risk factors for cervical glandular and squamous cell neoplasia: insights from a prospective cohort study of over 117,000 women. (n.d.). https://doi.org/10.1186/s12889-026-29877-7

Image Credits: AI Generated

DOI: 10.1186/s12889-026-29877-7

Keywords: cervical cancer, HPV, adenocarcinoma, squamous cell carcinoma, CIN3, adenocarcinoma in situ, smoking, hormonal contraceptives, chlamydia, genital warts, prospective cohort study, cancer screening