Background Previous studies have shown some differences in uptake of cervical screening by sociodemographic factors. However, the available evidence on other lifestyle, reproductive and hormonal factors is limited and inconsistent. We examined screening uptake in relation to sociodemographic, lifestyle, reproductive and hormonal factors by linking data from a large prospective cohort, the Million Women Study (MWS), with the nationwide NHS Cervical Screening Programme records.

Methods Using linked NHS Cervical Screening Programme records for women before they were recruited into the MWS study, participants were classified as non-attenders or attenders for routine cervical screening. Logistic regression models were used to calculate odds ratios (OR) and 95% confidence intervals (CI) of non-attendance versus attendance by deprivation status, smoking history, body mass index (BMI), parity, age at first birth, oral contraceptive (OC) use and menopausal hormone therapy (MHT) use. All analyses were stratified by year of birth and year of recruitment into the MWS and adjusted for other factors, where appropriate.

Results Of 871,732 study participants who were eligible to have been invited for cervical cancer screening, 25,261 were non-attenders and 846,471 were attenders. The odds of being a non-attender were increased with deprivation (OR [95%CI] 1.46 [1.40–1.53] for most versus least deprived fifth), obesity (1.38 [1.33–1.43] for BMI ≥30 versus 20–25/kgm²) and smoking (1.25 [1.19–1.30] for heavy current smokers versus never smokers). Nulliparous women were much more likely to be non-attenders (5.80 [5.28–6.38]) and women who were younger at their first birth were also less likely to attend (1.24 [1.12–1.37] for <17 versus ≥25 years at first birth). By contrast, women who had used OCs or MHT were much less likely to be non-attenders (0.38 [0.37–0.40] for ≥10 years versus never OC use and 0.33 [0.32–0.35] for current versus never MHT use).

Conclusion In this large cohort of women in England, attendance for cervical screening varied considerably, not only by deprivation, as had been reported previously, but also by lifestyle, reproductive and hormonal factors. Non-attendance was associated with obesity, heavy current smoking, nulliparity, giving birth at younger ages, and non-use of OCs or MHT.