Seasonal variation in prostate-specific antigen levels: a large cross-sectional study of men in the UK

School of Social and Community Medicine, University of Bristol, Bristol, Oncology Centre, Addenbrooke's Hospital, Cambridge, andNuffield Department of Surgery, University of Oxford, John Radcliffe Hospital, Oxford, UK.

Study Type - Aetiology (cohort) Level of Evidence 2b What's known on the subject? and What does the study add? Evidence for seasonal or climatic effects on PSA levels is mixed and inconclusive. In this large, long-term study, no seasonal or climatic patterns in PSA levels were identified.

OBJECTIVE: To assess whether a seasonal change in prostate specific antigen (PSA) levels can be detected in men recruited to a large clinical trial.

PATIENTS AND METHODS: A total of 66 969 men aged 50-69 years were drawn from a large study conducted at general practices across the UK between 2002 and 2007. Trigonometric algorithms and regression methods were used to assess the relationship between the time of year and serum PSA and blood pressure measurements. We obtained local daily mean temperatures and hours of sunlight per day to assess whether these factors were potential mechanisms for seasonal variation in PSA levels or blood pressure. The proportion of participants who would be considered clinically at risk according to their PSA or blood pressure measurement, by month, was also assessed. The strength of associations between time of year and blood pressure were used to reinforce conclusions from the PSA models.

RESULTS: There was no relationship between time of year and PSA levels (P= 0.11) or between climate and PSA levels (P= 0.42). No difference was found in the prevalence of clinically raised PSA content by month (P= 0.50). This lack of an association with PSA content was despite our data being sufficient to provide clear evidence of an association between blood pressure and time of year (systolic P < 0.001; diastolic P < 0.001), and to show that this association was largely explained by climatic factors (temperature and sunlight).

CONCLUSIONS: There was no pattern in PSA levels by time of year, air temperature or levels of sunlight in this cohort, so there is no need to take these factors into account when reviewing PSA results