Journal Logo


Body, indoor, outdoor temperature − and arterial blood pressure

Tikhonoff, Valériea; Casiglia, Edoardob

Author Information
doi: 10.1097/HJH.0000000000002782
  • Free

In the field of hypertension, many topics are still uncertain and would need to be clarified. Temperature is usually considered a minor problem because in real world, temperature [that of the body, that outdoor, and that indoor at home or at the outpatient clinic wherein office blood pressure (BP) is measured] is considered a negligible factor. Unfortunately, between BP and the temperatures listed above, there is a complex interaction largely unknown and quite always not considered.

This editorial article comments an interesting study that has the quality of being well done and the inconvenience of being cross-sectional and of taking into consideration outdoor temperature only.

That BP is lower in hot than in cold seasons has always been sustained by patients, worried about modifying their antihypertensive therapy accordingly (and they are not wrong, as actually the BP control rate decreases in cold climate). In this case, popular opinion preceded scientific data. Today, wide confirmations exist that BP values are higher and diagnosis of hypertension is more frequent in winter than in summer [1,2]. Outdoor temperature is the most commonly used in the temperature/BP studies because it is easy to obtain, for example considering seasonal variations or consulting for free the data provided by meteorological databases that are in the public domain.

Research mainly focused on low temperatures, and controlled studies demonstrated that cold climates are associated to higher values of SBP and DBP. To this purpose, analyses on small numbers of participants [3–5] were preferred, while great numbers of individuals from communities or populations are rare [6]. In general, in the Northern hemisphere, the highest BP levels are observed in February, when BP is in average 11–12/7–8 mmHg higher than in summer [3,7]. Kimura et al. found that in elderly individuals, 1°C outdoor temperature increases by +0.23/+0.49 mmHg [3]. More recently and more precisely, Hu et al.[8] found that for a 1°C increase in temperature, SBP decreases by 0.37 mmHg in normotensive individuals, by 0.21 mmHg in newly detected hypertensive individuals and by 0.81 mmHg in known hypertensive individuals, and DBP decreases by 0.19, 0.01 and 0.81 mmHg, respectively. Furthermore, prevalence of hypertension is 14% higher for every 10°C decrease in outdoor temperature [8]. Apart from the interesting stratification into three classes of BP, Hu et al's results corroborate those summarized above.

It is noteworthy that this is referred to peripheral BP only, as − in contrast − in the experiments by Hintsala et al., 15-min whole-body exposure to −10°C increases central aortic blood pressure by +2 mmHg [4]. This topic is briefly discussed below and deserves to be better studied.

Lower attention has been paid to indoor temperature. To obtain this information, it is necessary to measure the temperature in the outpatient clinic [9] or at the patient's home [5], or to be in an experimental setting [4], all of which are challenging conditions. It is mandatory to stay in a certain room and actively measure the temperature at the exact moment in which the BP is measured, a practice that requires time, effort, now-how and trained personnel.

In the Health Survey for England 2014 (HSE) on 4659 individuals, −1°C in indoor temperature was associated to +0.5 mmHg increase in both SBP and DBP, while +1°C corresponded to −0.73/−0.46 mmHg DBP [9]. We too can confirm this trend, as in our experience that involved 4000 unselected individuals aged 50 ± 15 years (range 18–90) of the Italian general population, both SBP and DBP were inversely correlated to the indoor temperature measured directly by a doctor in clinic or at the patient's home (Fig. 1) even after adjustment for a number of plausible confounding factors. In particular, age significantly contrasted but did not nullify the inverse effect of indoor temperature on BP values. In these studies, by our research group, indoor temperature ranged from 15.3°C to 35.0°C (in average 24.3 ± 3.3°C).

Negative linear regression between indoor temperature and SBP and DBP among 4000 unselected individuals from general population, adjusted for age, blood glucose, smoking and serum uric acid (all P < 0.0001). Adjusted R 2 is 0.289 for SBP (coefficient −0.647, 95% confidence intervals −0.018 to −0.376, P < 0.0001) and 0.180 for DBP (coefficient −0.511, 95% confidence intervals −0.669 to −0.156, P < 0.0001) (E. Casiglia, V. Tikhonoff, unpublished data on authorized informed-consent data).

Little is known about the effect of body temperature on BP, although it has been well established in complex experiments that higher values of body temperature are associated with higher values of peripheral and central BP [10]. The physiological mechanisms of this are far from being clarified [4,10]. We can only add that not only central BP but also cardiac output increase with increasing body temperature, as in our experience among 1000 unselected individuals from general population body temperature correlates directly to cardiac index (R2 = 0.235, P = 0.004) despite adjustment for significant confounders such as age, blood glucose, serum uric acid and smoking. However, everyday clinical reality does not have to be as simple as it seems, as in our experience, SBP and pulse pressure are not related to body temperature, although there is a direct age-adjusted correlation for diastolic BP only (R2 = 0.196, P = 0.006). This is an ever-changing field.

The problem probably is that in clinical practice outdoor, indoor and body temperatures are not independent from each other, and do not act independently from each other on BP. For instance, Sinha et al.[7] found that outdoor temperature has a strong confounding effect on the relationship between indoor temperature and SBP. This is not surprising, as a patient who goes to a heated outpatient clinic during the winter or in an air-conditioned outpatient clinic in the summer can experience true thermal shock. When BP is measured at a precise moment in such conditions, the net thermal effect represents the net combination of body temperature and of outdoor plus or minus indoor temperature, acting sometimes in the same and sometimes in opposite directions. For instance, during winter, peripheral BP naturally tends to show a long-term increase [6] paralleling that in central BP [4], while in summer BP naturally tends to show a long-term fall, counteracting the BP increase observed when body temperature rises. The current opinion is that cold temperature (seasonal or indoor or experimental) increases BP via a reflex increase in sympathetic drive. However, it is clear that this cannot be the answer, as this view assumes the existence of a ‘neutral temperature’ associated with ‘neutral BP’ on which the ambient temperature would act. But it is of course impossible to measure a BP independent of the combined effects of outdoor, indoor and body measures because this ideal neutral combination does not exist. An algorithm taking into consideration all these variables would be of interest when measuring BP in clinical and epidemiological settings, but such an empirical algorithm was not found yet.

The topic is of great interest, as the variations due to temperature are relevant and the guidelines are rigid, so that even +1 or −1 mmHg wrongly attributed is sufficient to label or not label an individual as hypertensive (so becoming a patient) on the basis of a BP measured in a particular moment, that is in hardly repeatable thermal conditions. This is all the more important if the recent less conservative USA guidelines are applied. As we already showed, hypertension labelling is not an innocent act, but has a long-term negative psychological impact [11]. Furthermore, a patient could receive unnecessary chronic therapy, be denied necessary therapy or have the antihypertensive therapy dose increased or decreased based on a BP momentarily falsified by outdoor, indoor or body temperature, or by a combination of the three.

For the epidemiologists, the problem is of paramount importance. Since many years, we have been used to measure in epidemiological setting not only BP and a lot of other biological, demographic and anagraphic parameters, but also body and room temperature, and to consider the season. This is necessary to avoid incorrectly framing a lot of individuals in the category ‘hypertensive’ or ‘normotensive’ based on data biased by temperature. As suggested by Wang et al.[6], temperature variations should always be considered when screening for hypertension.

Finally, temperature can act as a cofactor of high BP in predicting cardiovascular risk. Studies from European countries [12–14] have observed an increase in death rates from acute myocardial infarction and stroke in the winter not only in cold European nations but also in the mild climate of a Mediterranean areas. Moreover, in the context of climate change, sudden temperature change as has been linked with elevated short-term mortality [15]. As reported by the Multi-Country Multi-City Collaborative Research Network, nonoptimum temperature exposures (extremely hot or cold) could account for 7.7% of total mortality and could be associated with changes in vulnerability or adaptation to temperature extremes [16]. Indeed, human's thermoregulatory system could not respond efficiently to the sudden temperature changes within a very short period. Unfortunately, Hu et al.[8] could not examine this problem in their cross-sectional study and were therefore unable to confirm this hypothesis or not.

In our personal experience on thousands of all-age unselected individuals from the general population followed for many years, in in Cox models including BP, indoor temperature was irrelevant in predicting all-cause and cardiovascular events, while body temperature had a significant role in predicting inversely the cardiovascular events with an estimate of −0.297 (95% confidence intervals −0.496 to −0.099, P = 0.003). Furthermore, we defined through the receiver operator characteristic curves method a plausible cut-off value of body temperature for models including systolic BP (36.0°C temperature, 95% confidence intervals, 35.5–36.2, area under curve 0.545, sensitivity 62.29, specificity 46.41, Z = 2.790, P = 0.016): this cut-off was accepted in a Cox analysis adjusted for confounders, where it divided individuals destined to be free from cardiovascular events with a hazard ratio of 0.736 (Fig. 2). SBP was accepted in the model (estimate 0.016, 95% confidence intervals 0.002–0.020, P < 0.0001) while DBP did not.

Hazard ratio with 95% confidence intervals of being over the prognostic cut-off of body temperature (36°) from Cox model adjusted for the same confounders listed in Fig. 1. (E. Casiglia, V. Tikhonoff, unpublished data on authorized informed-consent data).

Blood pressure and temperature (outdoor, indoor and clinical) interact continuously producing effects not already completely understood and very often neglected. The article by Hu et al. [8] contributes to clarify the role of outdoor and seasonal temperature, although other studies, both epidemiological and experimental, are mandatory to define the net effect of the three temperatures on measured BP. An important contribution could come from studies involving ambulatory BP monitoring, particularly if the recorded values are compared to clinically measured BP. It Is also mandatory to check deliberately, also in a prospective view, the ideas exposed in the article by Hu et al. [8] in unselected cohorts of elderly and very old people, representing − in affluent societies − the great majority of the hypertensives at a population level [17].


Conflicts of interest

There are no conflicts of interest.


1. Brennan PJ, Greenberg G, Miall WE, Thompson SG. Seasonal variation in arterial blood pressure. Br Med J 1982; 285:919.
2. Woodhouse P, Khaw K, Plummer M. Seasonal variation of blood pressure and its relationship to ambient temperature in an elderly population. J Hypertens 1993; 11:1267–1274.
3. Kimura T, Senda S, Masugata H, Yamagami A, Okuyama H, Kohno T, et al. Seasonal blood pressure variation and its relationship to environmental temperature in healthy elderly Japanese studied by home measurements. Clin Exper Hypertens 2010; 32:8–12.
4. Hintsala H, Kandelberg A, Herzig KH, Rintamäki H, Mäntysaari M, Rantala A, et al. Central aortic blood pressure of hypertensive men during short-term cold exposure. Am J Hypertens 2014; 27:656–664.
5. Imai Y, Munakata M, Tsuji I, Ohkubo T, Satoh H, Yoshino H, et al. Seasonal variation in blood pressure in normotensive women studied by home measurements. Clin Sci 1996; 90:55–60.
6. Wang S, Li M, Hua Z, Ye C, Jiang S, Wang Z, et al. Outdoor temperature and temperature maintenance associated with blood pressure in 438,811 Chinese adults. Blood Press 2017; 26:246–254.
7. Sinha P, Kumar TD, Singh NP, Saha R. Seasonal variation of blood pressure in normotensive females aged 18 to 40 years in an urban slum of Delhi, India. Asia Pac J Public Heal 2010; 22:134–145.
8. Hu J, He G, Luo J, Xu Y, Xu X, Song X, et al. Temperature-adjusted hypertension prevalence and control rate: a series of cross-sectional studies in Guangdong Province, China. J Hypertens 2021; 39:911–918.
9. Zhao H, Jivraj S, Moody A. ‘My blood pressure is low today; do you have the heating on?’ The association between indoor temperature and blood pressure. J Hypertens 2019; 37:504–512.
10. Kenney WL, Stanhewicz AE, Bruning RS, Alexander LM. Blood pressure regulation III: what happens when one system must serve two masters: temperature and pressure regulation? Eur J Appl Physiol 2014; 114:467–479.
11. Tikhonoff V, Hardy R, Deanfield J, Friberg P, Muniz G, Kuh D, et al. The relationship between affective symptoms and hypertension-role of the labelling effect: the 1946 British birth cohort. Open Heart 2016; 3:e000341.
12. Douglas AS, Dunnigan MG, Allan TM, Rawles JM. Seasonal variation in coronary heart disease in Scotland. J Epidiol Community Health 1995; 49:575–582.
13. Dilaveris P, Synetos A, Giannopoulos G, Gialafos E, Pantazis A, Stefanadis C. CLImate impacts on myocardial infarction deaths in the athens TErritory: the CLIMATE study. Heart 2006; 92:1747–1751.
14. Ostendorf T, Bernhard M, Hartwig T, Voigt M, Keller T, Stumvoll M, et al. Association between rapid weather changes and incidence of chiefly cardiovascular complaints in the emergency department. Am J Emerg Med 2020; 38:1604–1610.
15. Cheng J, Xu Z, Zhu R, Wang X, Jin L, Song J, et al. Impact of diurnal temperature range on human health: a systematic review. Int J Biometeorol 2014; 58:2011–2024.
16. Gasparrini A, Guo Y, Hashizume M, Lavigne E, Zanobetti A, Schwartz J, et al. Mortality risk attributable to high and low ambient temperature: a multicountry observational study. Lancet 2015; 386:369–375.
17. Casiglia E, Mazza A, Tikhonoff V, Pavei G, Privato G, Schenal N, et al. Weak effect of hypertension and other classic risk factors in the elderly who have already paid their toll. J Hum Hypert 2002; 16:21–31.
Copyright © 2021 Wolters Kluwer Health, Inc. All rights reserved.