Impact of environmental temperature on blood pressure phenotypes: a nationwide home blood pressure monitoring study

See the editorial comment for this article ‘Is there a seasonal variation in the office vs. out-of-office blood pressure difference?’, by A. Kollias et al., https://doi.org/10.1093/eurjpc/zwae014.

The diagnosis and management of hypertension usually relies on office blood pressure (OBP) measurements, but current guidelines have encouraged the evaluation of out-of-office blood pressure (BP) measurements using home BP monitoring (HBPM) or ambulatory BP monitoring as an approach to more accurately assess the true BP burden.¹ Differences between OBP and out-of-office BP (office–home difference, OHD) may impact hypertension diagnosis and management, resulting in BP phenotypes with prognostic value, such as white-coat hypertension (WCH) and masked hypertension (MH).^1,2

There is a consistent inverse association between BP and environmental temperature (ET).^3,4 Meta-analysis results show that each 1°C ET decrease is related to increases in systolic and diastolic OBP of 0.26 [95% confidence interval (CI): 0.18–0.33] and 0.13 (95% CI: 0.11–0.16) mmHg.⁴ This aspect holds clinical significance because the impact of ET on BP behaviour is likely to increase due to ongoing global warming and climate changes.⁵ Conversely, the influence of ET on OHD is far from established. Notably, while WCH is described to increase in warmer seasons, controversial data regarding the relationship between MH and ET are reported.^6–8 To tackle these issues, this study explored the link between outdoor mean temperature (OMT), OHD, and BP phenotypes in a large nationwide sample using HBPM.

This retrospective cross-sectional study evaluated 70 949 unique individuals from 460 centres in 22 cities representing all Brazilian geographical regions [North cities: Belém (n = 1565), Manaus (n = 1513), Porto Velho (n = 1570); Northeast cities: Aracaju (n = 1916), Recife (n = 21 560), Caruaru (n = 1202), Maceió (n = 887), Petrolina (n = 2508), Salvador (n = 3285), Feira de Santana (n = 1375), São Luiz (n = 1344); Southeast cities: Belo Horizonte (n = 709), Campinas (n = 9239), Campos do Jordão (n = 2748), Rio de Janeiro (n = 2681), São Paulo (n = 4827); Central-West cities: Brasília (n = 1947), Goiânia (n = 4998); South cities: Caxias do Sul (n = 508), Curitiba (n = 997), Porto Alegre (n = 1911), Santa Maria (n = 1660)] who underwent single OBP and HBPM measurements, and collected data on environmental variables from July 2018 to July 2022.

Office blood pressure was determined by averaging two office readings, while HBPM was calculated by averaging three home BP measurements taken in the morning and evening over four consecutive days, using validated devices.^9,10 Office–home difference was defined as the OBP–HBPM difference. Blood pressure phenotypes were defined as normotension (OBP < 140/90 mmHg and HBPM < 135/85 mmHg), WCH (OBP ≥ 140/90 mmHg and HBPM < 135/85 mmHg), MH (OBP < 140/90 mmHg and HBPM ≥ 135/85 mmHg) and sustained hypertension (SH; OBP ≥ 140/90 mmHg and HBPM ≥ 135/85 mmHg).¹ Data on daily OMT, outdoor wind velocity, atmospheric pressure, and humidity were obtained from meteorological stations in the cities. Outdoor mean temperature, wind velocity, atmospheric pressure, and humidity represented the average values during the days when OBP and HBPM measurements were obtained for each individual. The protocol was approved by the Oswaldo Cruz University Hospital/PROCAPE Complex Ethics Committee. An expanded Methods section is available in Supplementary Material online.

The sample contained 39% men, with 51% using anti-hypertensive medications (AHmed), age = 57.2 ± 15.6 years, body mass index = 28.7 ± 5.3 kg/m², systolic OBP = 131.4 ± 19.6 mmHg, diastolic OBP = 83.9 ± 11.7 mmHg, systolic HBPM = 124.9 ± 15.7 mmHg, and diastolic HBPM = 79.3 ± 9.6 mmHg. The rates of normotension, WCH, MH, and SH were 47, 15, 11, and 27%, respectively. Average OMT, wind velocity, atmospheric pressure, and humidity values were 24.0 ± 4.0°C, 13.4 ± 4.2 km/h, 975.3 ± 47.0 hPa, and 73.3 ± 10.1%, respectively. Characteristics of the sample according to AHmed use and seasons are shown in Supplementary material online, Tables S1 and S2, respectively.

A multivariable linear regression analysis adjusted for potential confounding factors showed that OMT exhibited a stronger inverse association with systolic HBPM (β = −0.44 ± 0.02; P < 0.001) and diastolic HBPM (β = −0.24 ± 0.01; P < 0.001) than with systolic OBP (β = −0.20 ± 0.03; P < 0.001) and diastolic OBP (β = −0.16 ± 0.02; P < 0.001; see Supplementary material online, Figure S1). Consequently, there was a direct relationship between OMT and systolic (β = 0.25 ± 0.02; P < 0.001) and diastolic (β = 0.08 ± 0.01; P < 0.001) OHD (Figure 1A).

Figure 1

Relationship of environmental temperature with office-home blood pressure difference (OHD) and blood pressure phenotypes. (A) Linear regression analysis between systolic and diastolic office–home difference and outdoor mean temperature (OMT) adjusted for age, sex, body mass index, anti-hypertensive medications use, calendar time, centre, seasons, city altitude, and daily outdoor wind velocity, atmospheric pressure, and humidity in all studied individuals (n = 70 949). The dashed lines indicate the 95% confidence intervals. Bubble plots for the association between outdoor mean temperature and rates of white-coat hypertension (B), masked hypertension (C), normotension (D), and sustained hypertension (E) in the studied cities. Rates of blood pressure phenotypes for each city (y-axis) were estimated by using logistic regression analysis adjusted for age, sex, body mass index, anti-hypertensive medication use, calendar time, centre, seasons, and daily outdoor wind velocity and humidity. City altitude and daily outdoor atmospheric pressure were not included in the model due to collinearity. Outdoor mean temperature values in (B–E) corresponded to the average outdoor mean temperature values related to office blood pressure and home blood pressure monitoring for all individuals in each city (x-axis). Weighted regression lines are shown in black. The sizes of the bubbles correspond to the weight of the city data. The regions of the studied cities are distinguished by different colours.

Open in new tabDownload slide

A multivariable logistic regression analysis revealed that each 1°C OMT increase was associated with 1.3% (95% CI: 0.4–2.1%; P = 0.003) higher risk of WCH, 4.3% (95% CI: 3.6–4.9%; P < 0.001) higher risk of normotension, 3.9% (95% CI: 3.1–4.8%; P < 0.001) lower risk of MH, and 3.5% (95% CI: 2.9–4.1%; P < 0.001) lower risk of SH (see Supplementary material online, Table S3). Furthermore, there was a stronger relationship of OMT with normotension in participants not using AHmed and with SH in participants using AHmed, which disappeared after further adjusting for OBP (see Supplementary material online, Table S3).

Cities located in northern regions, characterized by higher-average OMT, exhibited higher rates of WCH and normotension. Conversely, cities in southern regions, characterized by lower-average OMT, demonstrated higher rates of MH and SH (Figure 1B–E).

This analysis, conducted on a large nationwide sample from a continental country, revealed that warmer cities and higher ET were linked to increased rates of WCH and normotension, while lower ET and colder cities were associated with higher MH and SH rates. These findings may have potential clinical implications. First, they suggest that greater WCH and MH suspicion should be considered among individuals with elevated OBP in warmer regions and individuals with normal OBP in colder regions, respectively. Given the limited availability of out-of-office BP measurements, incorporating ET data could help improve BP phenotypes diagnosis. Second, the higher rates of MH and SH in colder cities raise the idea that ET should be considered in the formulation of public policies on the prevention and control of hypertension, given the elevated cardiovascular risk of these phenotypes.¹ Third, in multi-centre randomized clinical trials, ET may emerge as a potential cause of OHD, serving as a confounding factor that may affect the impact of intervention strategies on BP, especially out-of-office BP, across different geographical regions. Conversely, a potential explanation by which HBPM decreases more than OBP in warm temperatures could be that Brazilian houses usually do not have widespread installation of air conditioning systems and therefore may experience higher temperatures than at the office.

This study has limitations. Data on other risk factors for BP phenotypes, including diabetes, salt intake, smoking, and indoor temperature, which could be affected by air conditioning and heating,³ were unavailable. Home blood pressure monitoring was measured using triplicate measurements and not duplicate ones as stipulated by the recommendations of guidelines.¹ Furthermore, due to the retrospective and observational nature of the study, the potential influence of selection bias and unmeasured factors on our results cannot be disregarded.

In conclusion, our data indicate that ET is directly associated with WCH and inversely related to MH, suggesting that there is a need to consider ET when evaluating and treating hypertension based on combined OBP and out-of-office BP measurements.

Supplementary material

Supplementary material is available at European Journal of Preventive Cardiology.

Author contribution

E.C.D.B., A.D.M.F., and W.N. contributed to the conception or design of the work. E.C.D.B., A.D.M.F., M.V.R.S., M.A.M.-G., W.S.B., R.D.M., A.A.B., G.F., J.L.L.-F., J.A., M.L.S.N., I.C.B.G.P., B.C.B., M.C.V.F., H.P.S., A.C.S., M.C., A.C., and W.N. contributed to the acquisition, analysis, or interpretation of data for the work. E.C.D.B. and W.N. drafted the manuscript. A.D.M.F., M.A.M.-G., M.V.R.S., W.S.B., R.D.M., A.A.B., G.F., J.L.L.-F., J.A., M.L.S.N., I.C.B.G.P., B.C.B., M.C.V.F., H.P.S., A.C.S., M.C., and A.C. critically revised the manuscript. All gave final approval and agreed to be accountable for all aspects of work, ensuring integrity and accuracy.

Funding

The study was supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq; grant 310869/2021-8 for W.N.).

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

References

Author notes