This cross-sectional study assesses whether an association exists between extreme heat exposure and blood pressure among adults receiving care at US community health centers.
Key Points
Question
Is exposure to extreme heat associated with acute changes in blood pressure among adults receiving care in US community health centers?
Findings
In this cross-sectional study of 4 221 866 ambulatory encounters, same-day extreme heat was associated with small but significant reductions in systolic and diastolic blood pressure, with larger reductions among adults aged 65 years or older.
Meaning
Findings indicate that extreme heat was associated with modest acute reductions in blood pressure, suggesting that extreme heat may have limited influence on blood pressure in routine outpatient settings and is unlikely to be meaningfully associated with individual-level cardiovascular risk assessment.
Abstract
Importance
Extreme heat has well-documented cardiovascular effects, yet its acute association with blood pressure (BP) in routine clinical settings remains incompletely characterized, particularly in large, diverse outpatient populations.
Objective
To assess the association between extreme heat exposure and BP among patients receiving care at community health centers (CHCs).
Design, Setting, and Participants
This cross-sectional study was conducted using electronic health record data from a national network of 2554 CHCs across 40 US states during the summer months (June through August) of 2019 to 2023. Participants included adult patients (≥18 years of age) with ambulatory encounters that included BP measurements. Data were analyzed October 2025 to April 2026.
Exposure
Same-day and short-term cumulative exposure to extreme heat, defined using the GridEX dataset (500 × 500 m spatial resolution) linked to clinic locations. Extreme heat was operationalized as days exceeding location-specific thresholds.
Main Outcomes and Measures
The primary outcomes were systolic and diastolic BP recorded during ambulatory visits. Associations were estimated using mixed-effects models with patient- and facility-level random intercepts, adjusting for age, sex, race and ethnicity, hypertension and medication status, comorbidities, year, and geographic region. Effect modification by age and hypertension and medication status was evaluated. Distributed lag nonlinear models assessed cumulative associations across an 8-day window for the binary extreme heat exposure indicator.
Results
Of 4 221 866 ambulatory encounters (cohort mean [SD] age, 47.6 [17.0] years; 63% female), most were for patients without a hypertension diagnosis (81%). Same-day extreme heat exposure was associated with lower systolic BP (−0.7 [95% CI, −0.8 to −0.6] mm Hg) and diastolic BP (−0.3 [95% CI, −0.3 to −0.2] mm Hg). Decreases were larger among adults aged 65 years or older (systolic: −1.1 [95% CI, −1.3 to −0.9] mm Hg vs −0.6 [95% CI, −0.7 to −0.5] mm Hg among adults aged <65 years; interaction P < .001; diastolic: −0.8 [95% CI, −1.0 to −0.6] mm Hg vs −0.3 [95% CI, −0.3 to −0.2] mm Hg among adults aged <65 years; interaction P < .001). Hypertension status modified the association between extreme heat exposure and systolic BP (omnibus interaction P = .003) but not diastolic BP (omnibus interaction P = .84). In distributed lag analyses, BP reductions peaked on the day of exposure and the following day and were attenuated by day 3, with minimal cumulative differences thereafter (cumulative difference across lags 0 through 7: −0.8 [95% CI, −1.0 to −0.6] mm Hg for systolic and −0.5 [95% CI, −0.6 to −0.4] mm Hg for diastolic).
Conclusions and Relevance
In this cross-sectional study, extreme heat exposure was associated with modest acute reductions in BP among adults receiving care at CHCs, with slightly larger reductions among older adults. These findings suggest physiologic responses to heat that may be relevant for clinical BP measurement and cardiovascular risk assessment in outpatient settings.
Introduction
The frequency and intensity of extreme heat events are increasing worldwide and are commonly considered an emerging threat to cardiovascular health.1,2,3 Major cardiovascular societies now consider environmental exposures, including heat extremes, as important and potentially modifiable contributors to the global burden of cardiovascular disease.4 This is driven by evidence that at the population level short-term increases in ambient temperature are associated with higher cardiovascular mortality and excess deaths during heatwaves, with particularly pronounced effects among older adults.5,6,7
Yet the pathways by which heat affects cardiovascular physiology outside controlled experimental settings remain incompletely understood. Specifically, the extent to which ambient heat exposure meaningfully alters blood pressure (BP)—a routinely measured indicator of cardiovascular function—is not well established in outpatient settings. Physiological responses to heat may plausibly act in opposing directions: heat-related vasodilation and redistribution of blood flow to the skin may lower BP, whereas dehydration, increased cardiovascular demand, and sympathetic activation may increase it, particularly among vulnerable populations.8,9,10,11
Prior studies reported heterogeneous BP responses to heat exposure, with uncertainty regarding whether individuals with hypertension experience greater vulnerability.12,13,14 The emerging framework of environmental hypertensionology emphasizes that BP reflects not only genetic, behavioral, and therapeutic factors but also contextual environmental exposures.15 However, most prior studies of heat and cardiovascular outcomes relied on laboratory environments or severe clinical end points, such as hospitalization and mortality, offering limited insight into early or subclinical physiologic responses.8,16
The acute physiological response to ambient heat in ambulatory populations, especially those served by safety net health care systems, remains largely uncharacterized. Community health centers (CHCs) provide a unique setting to study these effects. They serve patients from low-income households, routinely measure BP during outpatient visits, and maintain electronic health records (EHRs) suitable for exposure-outcome investigation.
In the present study, we assessed whether short-term exposure to extreme heat is associated with changes in systolic and diastolic BP among adults receiving ambulatory care at CHCs across the US. Rather than assuming a directional change, we sought to characterize the direction and magnitude of BP responses associated with extreme heat in routine clinical settings. We used high-resolution gridded heat exposure data linked to geocoded clinic locations and assessed associations across multiple temporal windows and patient subgroups. Understanding how extreme heat is associated with BP in routine clinical care may inform cardiovascular risk assessment and outpatient management during such periods.
Methods
Setting
This cross-sectional study used EHR data from OCHIN, a national network of more than 2500 CHC primary care sites across 40 US states. Member clinics predominantly serve diverse, low-income patients in medically underserved communities. OCHIN leads the Accelerating Data Value Across a National Community Health Center Network (ADVANCE) Clinical Research Network, which supports patient-centered outcomes research using EHR data from community-based primary care settings. ADVANCE maintains the most comprehensive dataset in the US on community-based health care organizations and the populations they serve.17,18 The ADVANCE Research Data Warehouse contains harmonized clinical data from participating clinics standardized to the PCORnet Common Data Model19 and includes patient demographic characteristics, diagnoses, medications, vital signs, and laboratory results from primary care encounters. The present analysis draws on data from a broader, National Institutes of Health–funded research initiative focused on extreme weather-related health risks and intervention development.20 The study was approved by the Advarra institutional review board and was granted a waiver of informed consent because it involved secondary analysis of routinely collected clinical data and posed minimal risk to participants. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cross-sectional studies.
Study Design and Population
We conducted a cross-sectional exposure-outcome study to assess the short-term association between extreme heat and BP. The cohort included patients aged 18 years or older with at least 1 visit in which BP was recorded during summer months (June through August) from 2019 to 2023. Patients with multiple summer visits contributed multiple observations. Encounters, rather than patients, were the unit of analysis.
BP measurements were obtained as part of routine outpatient clinical care across CHCs. Specific measurement protocols (eg, resting time prior to measurement, number of repeated measurements, or arm selection) were not standardized or consistently captured in structured EHR fields. When multiple BP measurements were recorded within an encounter, we used the mean of available values after restricting to physiologically plausible BP measurements (systolic, 70-269 mm Hg; diastolic, 50-149 mm Hg).21 These thresholds exclude implausible values while retaining clinically relevant hypertensive and hypotensive readings.
Restriction to Summer Months
Analyses were restricted to June, July, and August, when extreme heat events occur most frequently in the northern hemisphere. BP exhibits systematic seasonal variation, with higher levels in winter and lower levels in summer, independent of short-term heat exposure22; inclusion of cooler months would therefore introduce comparisons between extreme heat and substantially different physiological baselines rather than comparisons between extreme and typical warm-season conditions. Restriction to summer months minimized this source of confounding while preserving an appropriate comparator group reflecting usual warm-season temperatures.
Extreme Heat Exposure Data
Heat exposure was derived from GridEX, a 500 × 500–m gridded dataset covering the contiguous US from 2008 through 2023.23,24,25 GridEX interpolates station-level weather data to produce daily estimates of apparent temperature, a composite metric incorporating air temperature, humidity, and wind speed. Apparent temperature was calculated using the Steadman formula,26 which more closely reflects physiologic heat stress than dry-bulb temperature alone.
Extreme heat events were defined using the Excess Heat Factor, which quantifies temperature anomalies relative to recent acclimatization periods.27 Event intensity was quantified using the Extreme Heat Magnitude Indicator (EHMI), which scales heat events from 0 to 100 and incorporates both intensity and duration, with higher values indicating greater intensity and duration. Station-level EHMI values were interpolated using inverse distance weighting to produce continuous gridded surfaces.
Daily EHMI values were linked to patient encounters by geocoding clinic locations and extracting the corresponding grid-cell value for the encounter date. Patients were classified as exposed if the grid cell of the clinic recorded an EHMI value greater than 0 on the day of the visit.
Statistical Analysis
Systolic and diastolic BP were modeled as continuous outcomes at the encounter level. For each encounter, we constructed an 8-day exposure history extending from the encounter day (lag 0) through 7 days before the encounter (lag 7). Exposure was coded as a binary indicator (EHMI >0 vs EHMI = 0). We evaluated 3 exposure parameterizations: (1) acute exposure: EHMI higher than 0 on the encounter day; (2) distributed lag nonlinear model to estimate the distributed lag associations of the binary exposure across lags 0 through 7; and (3) cumulative exposure metrics, including total number of days exposed, cumulative EHMI, and mean EHMI across lags 0 through 7.
We examined the distribution of systolic and diastolic BPs according to exposure to an extreme heat event using empirical cumulative distribution functions. These curves enabled us to explore whether mean differences between groups were associated with a large number of patients with small differences vs a subset of individuals with large differences. Analyses were conducted overall, and effect modification was evaluated a priori by hypertension status (diagnosed hypertension with medication, hypertension without medication, none) and by age group (aged <65 vs ≥65 years). Interaction terms between extreme heat exposure and each subgroup indicator were included in the models to assess heterogeneity of associations across clinically relevant groups. All estimates were derived from these interaction models using the same exposure definitions and covariate adjustment. Statistical evidence of effect modification was assessed using the interaction term for age and an omnibus test of the exposure × hypertension status interaction.
Associations between extreme heat exposure and BP were estimated using Gaussian mixed-effects models, with patient- and facility-level random intercepts to account for clustering of patient encounters within clinics. Because the data included repeated observations from individual patients who were nested within clinics, intraclass correlation coefficients (ICCs) were estimated from unconditional multilevel models to quantify how much of the total outcome variability was attributable to differences between clinics and differences between patients. The clinic-level ICC represents the proportion of variance attributable to clinic-to-clinic differences, whereas the patient-level ICC represents the proportion of variance attributable to patient-to-patient differences and reflects the similarity of repeated observations from the same individual. Larger ICC values indicate that more of the total variability is explained by differences at that level. Analyses were restricted to encounters with complete data for the exposure, outcomes, and covariates included in the models; therefore, complete-case analysis was used, and no imputation was performed.
Covariates were selected a priori based on clinical plausibility and prior literature and included age (per 5-year increment), hypertension diagnosis and antihypertensive medication use, sex, race and ethnicity, Charlson Comorbidity Index, calendar year, and geographic region. Models were additionally adjusted for time of day of the clinical encounter using a natural spline function of encounter hour (4 degrees of freedom) to account for diurnal variation in BP. Race and ethnicity were collected as separate variables in the EHR, typically self-reported by patients at clinical registration, with categories defined by the health system, and grouped according to the PCORnet Common Data Model, version 7.0.19 Ethnicity was recorded separately as Hispanic or non-Hispanic. For this analysis, race and ethnicity were combined to create mutually exclusive categories: Hispanic Asian, Hispanic Black, Hispanic White, Hispanic other race, non-Hispanic Asian, non-Hispanic Black, non-Hispanic White, and non-Hispanic other race. These variables were included to examine potential differences in heat-related BP responses across diverse populations. The other race category included individuals identifying as multiracial or as a race not otherwise specified.
For the lag analyses, models estimated the distributed lagged associations of the binary indicator across lags 0 through 7. The cumulative association across the 8-day window was prespecified as the primary estimand. All analyses were conducted from October 2025 to April 2026, in R, version 4.5.2 (R Project for Statistical Computing). Statistical significance was defined as a 2-sided P < .05.
Results
The analytic dataset included 4 221 866 ambulatory encounters during summer months. The cohort had a mean (SD) age of 47.6 (17.0) years, was predominately female (63%; 37% male), and self-reported race and ethnicity of Hispanic Asian (<1%), Hispanic Black (1%), Hispanic White (28%), Hispanic other race (8%), non-Hispanic Asian (5%), non-Hispanic Black (16%), non-Hispanic White (38%), and non-Hispanic other race (4%). Of these ambulatory encounters, 106 428 (2%) occurred on days classified as extreme heat, and 4 115 438 (98%) occurred on days classified with no extreme heat. Additional baseline characteristics stratified by exposure status are shown in the Table.
Table. Baseline Characteristics of Ambulatory Visits During Summer Months, by Extreme Heat Exposure Status.
| Characteristic | Unexposed (n = 4 115 438) | Exposed (n = 106 428) |
|---|---|---|
| Blood pressure, mean (SD), mm Hg | ||
| Systolic | 124.8 (17.7) | 124.0 (17.3) |
| Diastolic | 76.5 (10.5) | 76.2 (10.4) |
| Encounter month, No. (%) | ||
| June | 1 325 879 (32) | 37 513 (35) |
| July | 1 315 675 (32) | 35 098 (33) |
| August | 1 473 884 (36) | 33 817 (32) |
| Encounter year, No. (%) | ||
| 2019 | 703 419 (17) | 9646 (9) |
| 2020 | 393 851 (10) | 9679 (9) |
| 2021 | 721 423 (18) | 29 141 (27) |
| 2022 | 968 695 (24) | 25 275 (24) |
| 2023 | 1 328 050 (32) | 32 687 (31) |
| Encounter time of day, No. (%) | ||
| Night (12:00-5:59 am) | 57 438 (1) | 1484 (1) |
| Morning (6:00-11:59 am) | 2 211 738 (50) | 54 943 (50) |
| Afternoon (12:00-5:59 pm) | 2 095 240 (47) | 52 365 (47) |
| Evening (6:00-11:59 pm) | 79 637 (2) | 2121 (2) |
| Age, mean (SD), y | 47.6 (17) | 47.1 (17) |
| Hypertension status, No. (%) | ||
| No hypertension | 3 344 688 (81) | 87 568 (82) |
| With medication | 725 418 (18) | 17 689 (17) |
| Without medication | 45 332 (1) | 1171 (1) |
| Sex, No. (%) | ||
| Female | 2 574 696 (63) | 65 572 (62) |
| Male | 1 540 742 (37) | 40 856 (38) |
| Race and ethnicity, No. (%) | ||
| Hispanic Asian | 9188 (<1) | 226 (<1) |
| Hispanic Black | 54 766 (1) | 1515 (1) |
| Hispanic White | 1 117 239 (27) | 30 537 (29) |
| Hispanic other race | 341 041 (8) | 7679 (7) |
| Non-Hispanic Asian | 244 222 (6) | 4463 (4) |
| Non-Hispanic Black | 699 543 (17) | 16 371 (15) |
| Non-Hispanic White | 1 490 541 (36) | 41 618 (39) |
| Non-Hispanic other racea | 158 898 (4) | 4019 (4) |
| Charlson Comorbidity Index, mean (SD)b | 2.1 (2.6) | 2.1 (2.6) |
| US region of facility, No. (%) | ||
| Midwest | 593 250 (14) | 15 063 (14) |
| West | 2 444 046 (59) | 63 628 (60) |
| South | 512 050 (12) | 12 051 (11) |
| Northeast | 566 092 (14) | 15 686 (15) |
Included individuals identifying as multiracial or as a race not otherwise specified.
Charleson Comorbidity Index ranges from 0 to 33, with higher values indicating greater comorbidity.
Mean (SD) systolic BP was 124.8 (17.7) mm Hg among unexposed encounters and 124.0 (17.3) mm Hg among exposed encounters. Mean (SD) diastolic BP was 76.5 (10.5) mm Hg in unexposed encounters compared with 76.2 (10.4) mm Hg in exposed encounters. Most encounters (81%) involved patients without a hypertension diagnosis; 18% involved patients treated with antihypertensive medication, and 1% involved patients with unmedicated hypertension. Mean BP was lower during encounters with symptom-related diagnoses on days of both extreme heat and no extreme heat (eTable 1 in Supplement 1). Blood pressure exhibited expected diurnal variation, with higher values in early morning hours, a midday decline, and a rise in the evening, as shown in Figure 1.
Figure 1. Line Graphs Illustrating Diurnal Patterns of Systolic and Diastolic Blood Pressure (BP) by Time of Day.
Smoothed curves show mean systolic and diastolic BP using generalized additive models. Shading indicates 95% CIs.
In adjusted mixed-effects models, same-day extreme heat exposure was associated with a small but significant reduction in systolic BP (−0.7 [95% CI, −0.8 to −0.6] mm Hg) and diastolic BP (−0.3 [95% CI, −0.3 to −0.2] mm Hg). Facility-level ICCs were 0.05 for systolic BP and 0.08 for diastolic BP, indicating that approximately 5% and 8% of the total variance, respectively, were attributable to differences between clinics. Patient-level ICCs were 0.40 for systolic BP and 0.41 for diastolic BP, indicating that approximately 40% and 41% of the total variance, respectively, were attributable to differences between patients, suggesting substantial similarity among repeated BP measurements from the same individual.
Differences associated with extreme heat exposure were consistent across percentiles for systolic BP, with values approximately −1 mm Hg lower at the 10th and 50th percentiles and −2 mm Hg lower at the 90th percentile in exposed compared with unexposed encounters (eTable 2 in Supplement 1). In contrast, diastolic BP showed no meaningful differences across percentiles. Cumulative distribution curves for systolic BP demonstrated a slight leftward shift for exposed encounters across the range of observed values, whereas they were nearly overlapping between exposure groups for diastolic BP (eFigure in Supplement 1). The absence of divergence in the tails indicates that, for systolic BP, differences reflected a small, population-wide shift rather than large changes concentrated in a subset of individuals.
In interaction analyses, extreme heat events were associated with BP, with observed differences greater in older compared with younger patients. Among adults 65 years of age or older, extreme heat exposure was associated with a −1.1 (95% CI, −1.3 to −0.9) mm Hg difference in systolic BP and −0.8 (95% CI, −1.0 to −0.6) mm Hg difference in diastolic BP compared with −0.6 (95% CI, −0.7 to −0.5) mm Hg difference in systolic BP and −0.3 (95% CI, −0.3 to −0.2) mm Hg difference in diastolic BP in adults younger than 65 years (interaction values for both BP measures were P < .001) (Figure 2). In analyses stratified by hypertension status, the overall interaction between extreme heat exposure and hypertension status was significant for systolic BP (omnibus interaction P = .003) but not for diastolic BP (omnibus interaction P = .84) (Figure 2). In the lag analyses of the overall population, extreme heat exposure was associated with lower BP in the short-term. For systolic BP, the largest differences occurred on the encounter day (lag 0) and the following day (lag 1) with differences attenuating toward the null by lag 3. The cumulative difference across lags 0 through 7 was −0.8 (95% CI, −1.0 to −0.6) mm Hg. Diastolic BP exhibited a similar temporal pattern, with maximal decreases at lags 0 to 1 and attenuating by lag 3. The cumulative difference across lags 0 through 7 was −0.5 (95% CI, −0.6 to −0.4) mm Hg. Lag response associations are shown in Figure 3.
Figure 2. Dot Plots of the Association Between Extreme Heat Exposure and Systolic and Diastolic Blood Pressure (BP) Across Age and Hypertension Subgroups.
Adjusted mean differences in systolic BP and diastolic BP associated with extreme heat exposure across age and hypertension subgroups are shown. Adjusted mean differences estimated from mixed-effects models that included exposure by subgroup interaction terms. The vertical dashed line at 0 mm Hg indicates no difference in BP associated with extreme heat exposure.
Figure 3. Line Graphs Showing the Lag Response and Cumulative Differences in Blood Pressure (BP) Associated With Extreme Heat.
Lag-specific and cumulative differences in binary extreme heat exposure (Extreme Heat Magnitude Indicator greater than 0 vs equal to 0) across lags 0 (visit day) through 7 for BP in the overall population. Models estimated the distributed lag associations of a binary exposure indicator using a natural cubic spline for the lag structure, adjusted for prespecified covariates with a facility-level random intercept. Shaded areas indicate 95% CIs. Δ DBP indicates change in diastolic BP; Δ SBP, change in systolic BP.
Discussion
In this large, nationally distributed cohort of ambulatory visits at CHCs, acute exposure to extreme heat was associated with small decreases in both systolic and diastolic BP. Notably, these differences were augmented in older patients (≥65 years), which may reflect age-related changes in autonomic regulation, vascular compliance, and thermoregulatory capacity.5,6,28,29 In contrast, effect modification by hypertension status was observed for systolic BP but not for diastolic BP. While experimental studies have demonstrated similar compensatory heat tolerance among normotensive and mildly hypertensive adults,14,30 our findings suggest that acute systolic BP responses to extreme heat may differ across hypertension and treatment groups in community-dwelling clinical populations. By leveraging routine clinical measurements across diverse community settings, our study complements experimental and mechanistic work and provides population-level insight into early physiological responses to extreme heat in everyday clinical practice. These findings suggest that in a large community population outside controlled laboratory settings, acute BP responses to heat are not amplified.
Experimental studies of heat exposure conducted under controlled laboratory conditions have documented pronounced cardiovascular strain, including increased heart rate, elevated cardiac output, and sympathetic activation.12,14 These findings have contributed to the expectation that heat exposure would increase BP, particularly among older adults and individuals with limited cardiovascular reserve. In contrast, the present observational findings suggest that vasodilatory responses to heat may play a more prominent role in ambulatory community settings. Consistent with this interpretation, the temporal pattern observed in distributed lag models showed the greatest BP reductions on the day of exposure and the subsequent day, with attenuation over several days. This pattern is compatible with an acute, reversible response rather than sustained hemodynamic stress. The Glasgow Blood Pressure Clinic study, which examined seasonal and monthly transitions in ambient temperature, found increases in BP associated with cold exposure and minimal changes during warm periods.31 Together, these findings suggest that the BP response to temperature depends on the temporal scale of exposure, with acute changes differing from responses observed under prolonged conditions.
These findings highlight the complexity of cardiovascular responses to extreme heat exposure in everyday community settings. Although extreme heat exposure was associated with statistically detectable changes in BP, the magnitude of these differences was minimal and unlikely to be clinically meaningful at the individual level. However, the consistency of small reductions across multiple analytic approaches and across the distribution of BP values supports a modest, population-wide physiological response, likely reflecting thermoregulatory vasodilation. These findings are somewhat at odds with expectations derived from controlled laboratory studies,12,14 which often emphasize cardiovascular strain under heat stress. In contrast, our results suggest that, in everyday community settings, compensatory mechanisms may largely preserve BP, resulting in minimal net change. This may help explain why epidemiologic associations between extreme heat and adverse cardiovascular outcomes are unlikely to be mediated primarily by acute changes in BP.
Several features of this population suggest that these estimates represent a lower bound on the physiologic response to extreme heat. CHCs capture routine ambulatory care, and patients who experience the largest heat-related reductions in BP, including those with symptomatic hypotension, dehydration, or heat-related illness, are more likely to present to emergency or urgent care and therefore to fall outside the encounters analyzed here. Consistent with the presence of such a subgroup, encounters carrying diagnoses of dizziness, dehydration, or heat-related illness showed lower BP than asymptomatic encounters among older adults (eTable 1 in Supplement 1), although the population-level association persisted after their exclusion. As such, even though the mean change at a population level was small, extreme heat days could produce a detectable hemodynamic response and adverse clinical consequence in many patients. Extreme heat days may therefore warrant proactive outreach to older patients, particularly those who are socioeconomically vulnerable and less likely to have reliable home cooling, including review of antihypertensive and diuretic regimens and attention to hydration and orthostatic symptoms.
Limitations
Important limitations warrant consideration. First, exposure assignment was based on clinic geocodes rather than residential locations or individual activity patterns. CHCs disproportionately serve socially vulnerable populations, and prior work suggests that these patients often reside in disadvantaged environments, which may partially mitigate exposure misclassification when using clinic locations.32 Second, BP measurements were obtained during routine clinical care, without control for recent activity or measurement protocol. Variability in clinical measurement practices across sites, including differences in resting time, measurement technique, and number of readings, may reduce measurement precision compared with standardized research protocols. Third, we cannot rule out selection bias related to health care utilization patterns. Extreme heat may influence whether patients attend scheduled visits, cancel appointments, or present for different types of care (eg, urgent vs routine visits). If such patterns vary by exposure status, the observed associations may partially reflect differences in the composition of encounters rather than true physiologic changes in BP. Fourth, although we did not observe effect modification of extreme heat based on hypertension with or without medications, we were unable to examine the impact of specific types of antihypertensive medications. Future studies should focus on a more granular assessment of antihypertensive medication classes, particularly diuretics and vasodilators, to clarify whether pharmacologic mechanisms contribute to heterogeneity in heat-related BP responses. Finally, although repeated encounters were included, not all patients contributed visits under both exposed and unexposed conditions. Because extreme heat exposure was relatively uncommon (approximately 2.5% of encounters), analyses restricted to within-patient comparisons would therefore substantially reduce the analytic sample and limit generalizability. To account for within-patient correlation and time-invariant individual characteristics, we incorporated patient-level random intercepts in the final models, although fully within-patient (fixed-effects) designs could yield additional insights.
Conclusions
In this cross-sectional study of a large ambulatory population, acute extreme heat exposure was associated with small reductions in BP, with slightly larger differences observed among older adults, and no meaningful modification by hypertension status. Although the magnitude of these differences was modest and unlikely to be clinically meaningful at the individual level, these findings indicate that cardiovascular responses to environmental heat in community settings may differ from those observed under controlled laboratory conditions. These findings underscore the importance of considering temporal scale, age, and everyday clinical and community contexts when evaluating heat-related cardiovascular risk, suggest that acute changes in BP are unlikely to be the primary pathway linking extreme heat to cardiovascular risk, and highlight the importance of exploring alternative physiological pathways of heat-related health effects.
eTable 1. Mean Blood Pressure by Symptom Status and Extreme Heat Exposure Among Adults Aged ≥65 Years During Summer Months
eTable 2. Distribution of Systolic and Diastolic Blood Pressure by Percentile and Extreme Heat Exposure Status
eFigure. Empirical Cumulative Distribution Functions of Systolic and Diastolic Blood Pressure by Extreme Heat Exposure Status
Data Sharing Statement
References
- 1.Liu J, Varghese BM, Hansen A, et al. Heat exposure and cardiovascular health outcomes: a systematic review and meta-analysis. Lancet Planet Health. 2022;6(6):e484-e495. doi: 10.1016/S2542-5196(22)00117-6 [DOI] [PubMed] [Google Scholar]
- 2.Alahmad B, Khraishah H, Royé D, et al. Associations between extreme temperatures and cardiovascular cause-specific mortality: results from 27 countries. Circulation. 2023;147(1):35-46. doi: 10.1161/CIRCULATIONAHA.122.061832 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Périard JD, Eijsvogels TMH, Daanen HAM. Exercise under heat stress: thermoregulation, hydration, performance implications, and mitigation strategies. Physiol Rev. 2021;101(4):1873-1979. doi: 10.1152/physrev.00038.2020 [DOI] [PubMed] [Google Scholar]
- 4.Münzel T, Lüscher T, Kramer CM, Churchwell K, Mbakwem A, Rajagopalan S. Environmental stressors and cardiovascular health: acting locally for global impact in a changing world: a statement of the European Society of Cardiology, the American College of Cardiology, the American Heart Association, and the World Heart Federation. Glob Heart. 2026;21(1):3. doi: 10.5334/gh.1514 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Millyard A, Layden JD, Pyne DB, Edwards AM, Bloxham SR. Impairments to thermoregulation in the elderly during heat exposure events. Gerontol Geriatr Med. 2020;6:2333721420932432. doi: 10.1177/2333721420932432 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Meade RD, Akerman AP, Notley SR, et al. Physiological factors characterizing heat-vulnerable older adults: a narrative review. Environ Int. 2020;144:105909. doi: 10.1016/j.envint.2020.105909 [DOI] [PubMed] [Google Scholar]
- 7.Kenny GP, Yardley J, Brown C, Sigal RJ, Jay O. Heat stress in older individuals and patients with common chronic diseases. CMAJ. 2010;182(10):1053-1060. doi: 10.1503/cmaj.081050 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Arsad FS, Hod R, Ahmad N, et al. The impact of heatwaves on mortality and morbidity and the associated vulnerability factors: a systematic review. Int J Environ Res Public Health. 2022;19(23):23. doi: 10.3390/ijerph192316356 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Charkoudian N. Skin blood flow in adult human thermoregulation: how it works, when it does not, and why. Mayo Clin Proc. 2003;78(5):603-612. doi: 10.4065/78.5.603 [DOI] [PubMed] [Google Scholar]
- 10.Kellogg DL Jr, Pérgola PE, Piest KL, et al. Cutaneous active vasodilation in humans is mediated by cholinergic nerve cotransmission. Circ Res. 1995;77(6):1222-1228. doi: 10.1161/01.RES.77.6.1222 [DOI] [PubMed] [Google Scholar]
- 11.Chaseling GK, Uchmanowicz I, Bäck M, et al. Heat extremes and cardiovascular diseases: a scientific statement of the Association of Cardiovascular Nursing & Allied Professions, Association for Acute Cardiovascular Care, European Association of Preventive Cardiology, Heart Failure Association, European Heart Rhythm Association of the ESC, the ESC Council on Hypertension, the ESC Council on Stroke, and the ESC Working Group on Cardiovascular Pharmacotherapy. Eur Heart J. 2025;46(30):2950-2958. doi: 10.1093/eurheartj/ehaf326 [DOI] [PubMed] [Google Scholar]
- 12.Sahu S, Sett M, Kjellstrom T. Heat exposure, cardiovascular stress and work productivity in rice harvesters in India: implications for a climate change future. Ind Health. 2013;51(4):424-431. doi: 10.2486/indhealth.2013-0006 [DOI] [PubMed] [Google Scholar]
- 13.Stotz A, Rapp K, Oksa J, et al. Effect of a brief heat exposure on blood pressure and physical performance of older women living in the community—a pilot-study. Int J Environ Res Public Health. 2014;11(12):12623-12631. doi: 10.3390/ijerph111212623 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Leach OK, Fisher KG, Cottle RM, Kenney WL. No effect of stage 1 hypertension or hypertensive medication on critical environmental limits (PSU HEAT Project). Am J Physiol Heart Circ Physiol. 2025;329(2):H395-H400. doi: 10.1152/ajpheart.00038.2025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Brook RD, Weder AB, Rajagopalan S. “Environmental hypertensionology” the effects of environmental factors on blood pressure in clinical practice and research. J Clin Hypertens (Greenwich). 2011;13(11):836-842. doi: 10.1111/j.1751-7176.2011.00543.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bunker A, Wildenhain J, Vandenbergh A, et al. Effects of air temperature on climate-sensitive mortality and morbidity outcomes in the elderly; a systematic review and meta-analysis of epidemiological evidence. EBioMedicine. 2016;6:258-268. doi: 10.1016/j.ebiom.2016.02.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bensken WP, Cottrell E, Templeton A, Gioia S, Lowe S, Stowe S, et al. The ADVANCE Clinical Research Network: past, present, and future: accelerating partnerships for patient-centered research in community-based primary care settings. Med Care. 2026;64(2S suppl 3):S205-S212. doi: 10.1097/MLR.0000000000002232 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.PCORnet . Network spotlight: ADVANCE. Accessed July 30, 2026. https://pcornet.org/network/advance/
- 19.PCORnet Common Data Model (CDM) Specification, version 7.0. Accessed July 15, 2026. https://pcornet.org/wp-content/uploads/2025/05/PCORnet_Common_Data_Model_v70_2025_05_01.pdf
- 20.Gold R, Estiri H, Bunce A, et al. Developing a community-prioritized research agenda on primary care strategies for minimizing the health impacts of heat and poor air quality events: protocol. Front Clim. 2025;7:1703757. doi: 10.3389/fclim.2025.1703757 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tapela N, Collister J, Clifton L, Turnbull I, Rahimi K, Hunter DJ. Prevalence and determinants of hypertension control among almost 100 000 treated adults in the UK. Open Heart. 2021;8(1):e001461. doi: 10.1136/openhrt-2020-001461 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Asayama K, Yatabe J, Metoki H, Kikuya M, Ohkubo T. Current perspective on seasonal variation in blood pressure. Kardiol Pol. 2026. Published online May 6, 2026. doi: 10.33963/v.phj.112495 [DOI] [PubMed] [Google Scholar]
- 23.Fard P, Chung MKJ, Estiri H, Patel CJ. Spatio-temporal interpolation and delineation of extreme heat events in California between 2017 and 2021. Environ Res. 2023;237(Pt 2):116984. doi: 10.1016/j.envres.2023.116984 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Fard P, Patel CJ, Estiri H. GridEX: spatiotemporal repository of extreme heat and cold exposure in the United States (2008-2023). Published online 2025. Accessed July 26, 2026. doi: 10.7910/DVN/LEK2RQ [DOI] [PMC free article] [PubMed]
- 25.Fard P, Hügel J, Keivabu RC, et al. Developing a spatiotemporal repository of extreme heat and cold exposure in the United States for precision public health research. Environ Sci Technol. 2025;59(37):19872-19884. doi: 10.1021/acs.est.5c03658 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Steadman RG. A universal scale of apparent temperature. J Appl Meterorol Climatol. 1984;23(12):1674-1687. doi: 10.1175/1520-0450(1984)023<1674:AUSOAT>2.0.CO;2 [DOI] [Google Scholar]
- 27.Sheridan SC, Lee CC, Allen MJ. The mortality response to absolute and relative temperature extremes. Int J Environ Res Public Health. 2019;16(9):9. doi: 10.3390/ijerph16091493 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Alrosan AZ, Heilat GB, Alrosan K, et al. Autonomic brain functioning and age-related health concerns. Curr Res Physiol. 2024;7:100123. doi: 10.1016/j.crphys.2024.100123 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Balmain BN, Sabapathy S, Louis M, Morris NR. Aging and thermoregulatory control: the clinical implications of exercising under heat stress in older individuals. Biomed Res Int. 2018;2018:8306154. doi: 10.1155/2018/8306154 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kenney WL, Kamon E. Comparative physiological responses of normotensive and essentially hypertensive men to exercise in the heat. Eur J Appl Physiol Occup Physiol. 1984;52(2):196-201. doi: 10.1007/BF00433392 [DOI] [PubMed] [Google Scholar]
- 31.Aubinière-Robb L, Jeemon P, Hastie CE, et al. Blood pressure response to patterns of weather fluctuations and effect on mortality. Hypertension. 2013;62(1):190-196. doi: 10.1161/HYPERTENSIONAHA.111.00686 [DOI] [PubMed] [Google Scholar]
- 32.Bensken WP, Navale SM, McGrath BM, et al. Variation in multimorbidity by sociodemographics and social drivers of health among patients seen at community-based health centers. J Multimorb Comorb. 2024;14:26335565241236410. doi: 10.1177/26335565241236410 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Citations
- Fard P, Patel CJ, Estiri H. GridEX: spatiotemporal repository of extreme heat and cold exposure in the United States (2008-2023). Published online 2025. Accessed July 26, 2026. doi: 10.7910/DVN/LEK2RQ [DOI] [PMC free article] [PubMed]
Supplementary Materials
eTable 1. Mean Blood Pressure by Symptom Status and Extreme Heat Exposure Among Adults Aged ≥65 Years During Summer Months
eTable 2. Distribution of Systolic and Diastolic Blood Pressure by Percentile and Extreme Heat Exposure Status
eFigure. Empirical Cumulative Distribution Functions of Systolic and Diastolic Blood Pressure by Extreme Heat Exposure Status
Data Sharing Statement
