Background Age related changes in blood pressure (BP) show increases around midlife. Such increases may be an early indication of arterial aging or a disease process. We investigated the presence of subpopulations with different underlying trajectories (latent classes) of midlife systolic (SBP) and diastolic blood pressure (DBP) in men and women.

Design Prospective UK birth cohort study.

Participants 1840 males and 1819 females.

Outcome BP at age 36, 43 and 53 years.

Analysis Unconditional linear growth mixture models were fitted to SBP and DBP to obtain latent classes. To examine the relevance of the extracted classes, we compared the distribution of early life body size, midlife body mass index (BMI) and weight change, lifetime socio-economic position (SEP), antihypertensive medication (HypRx) use and self reported angina in each class.

Results In men, 2 class models for SBP and DBP had the best fit in terms of the Bayesian information criterion. For SBP, the majority were in the “Normal” class characterised by a lower BP at age 36 and a gentle midlife increase (+0.9 mmHg per year). The other class (“Increaser”) had a higher increase (+3.1 mmHg per year). Similar classes were found for DBP. A “Normal” and “Increaser” class were also observed among women, together with an additional class (“High”) with high BP at age 36 (SBP=170; DBP=100 mmHg) and no evidence of a change with age. In both sexes, a smaller proportion were in the “Normal” SBP class compared to the “Normal” DBP class—for example, 94.3% of men were in the “Normal” SBP class vs 97.6% in the “Normal” DBP class. Individuals in the “Normal” classes were heavier at birth, taller at age 7, had a lower midlife BMI and midlife weight change, and were less likely to be on HypRx compared with those in other classes. Manual childhood SEP was associated with the “Increaser” SBP class in men. Undiagnosed angina (Rose Questionnaire) was more prevalent in the male “Increaser” SBP class. GP diagnosed angina (self-reported) and HypRx use were most prevalent in the “High” female SBP and DBP classes. Women in this group also had the lowest midlife weight change suggesting this was an extant morbid group.

Conclusion There was heterogeneity in the progression of midlife BP. This analytical approach may be useful for exploring determinants of BP and for identifying individuals at a high risk of future hypertension/CVD.