The interplay of temperature with clinic and ambulatory blood pressure and mortality: a prospective observational cohort study

M. Tokcan (Homburg/Saar)¹, I. Schwantke (Homburg/Saar)¹, A. De la Sierra (Barcelona)², T. Keßler (Homburg/Saar)¹, B. Haring (Wien)³, E. Vinyoles (Barcelona)², M. Gorostidi (Oviedo)⁴, J. Segura (Madrid)⁵, B. Williams (London)⁶, N. Staplin (Oxford)⁷, F. Mahfoud (Basel)⁸, L. Lauder (Basel)⁸, C. V. Schneider (Aachen)⁹, L. M. Ruilope (Madrid)⁵, M. Böhm (Homburg/Saar)^1
1Universitätsklinikum des Saarlandes Innere Medizin III - Kardiologie, Angiologie und internistische Intensivmedizin Homburg/Saar, Deutschland; ²Barcelona, Spanien; ³Klinik Hietzing 4. Medizinische Abteilung - Innere Medizin mit Kardiologie Wien, Österreich; ⁴Oviedo, Spanien; ⁵Madrid, Spanien; ⁶University College London London, Großbritannien; ⁷Oxford, Großbritannien; ⁸Universitätsspital Basel Abt. für Kardiologie Basel, Schweiz; ⁹Uniklinik RWTH Aachen Med. Klinik III - Gastroenterologie, Stoffwechselerkrankungen und Internistische Intensivmedizin Aachen, Deutschland

Background:
Rising ambient temperatures driven by climate change bear a significant healthcare challenge. The association between temperature and 24h-mean, daytime, and nighttime blood pressure (BP), as well as all-cause and cardiovascular mortality, was assessed in a large national European cohort recruited over a decade.

Methods:
Data of 58,222 patients of the prospective Spanish  ambulatory BP monitoring  (ABPM Registry) were linked with high spatiotemporal precision to temperature data from  Spanish State Meteorological Agency (AEMET). Inclusion criteria were guideline-based indications for ABPM, including white-coat, masked, borderline, or high-risk hypertension. Data collection took place at 223 primary care centers, recruited across all 17 regions of Spain, between March 1, 2004, and December 31, 2014, and follow-up for vital status extended to December 31, 2019. Temperature–BP associations were modelled using linear mixed-effects models, and temperature–mortality associations using Cox regression, adjusting for seasonality and individual risk factors.

Results:
The median (interquartile range, IQR) distance to the nearest weather station was 5.7 (3.0-9.6) km, and the median (IQR) time difference to the closest ABPM reading was 7.0 (3.5–16.3) minutes at daytime and 8.0 (4.0–13.0) minutes at nighttime. The overall temperature range was -11.9 to 43.9°C. The strongest association between temperature and systolic BP was observed during daytime, with a 4.2 mmHg (95% CI 4.0–4.5) increase per 10°C decrease. At nighttime, higher temperatures were associated with higher BP in patients with sustained and masked hypertension. Median (IQR) follow-up time was 9.7 (7.7–11.3) years, with 7,062 all-cause deaths and 2,327 cardiovascular deaths. 24h-mean and daytime temperature showed a significant U-shaped association with all-cause (p=0.0007 for 24h-mean and p<0.0001 for daytime) and cardiovascular death (p=0.0454 for 24h-mean and p=0.0232 for daytime), whereas nighttime temperature showed a J-shaped association with all-cause death (p=0.0204) but not cardiovascular death (p=0.2982).

Conclusion:
This large-scale, high-precision dataset highlights the importance of temperature for BP and mortality. Clinical guidelines should incorporate the role of temperature in hypertension management.