Correlates of Nocturnal Hypertension in a Real-World Cohort

Abstract

BACKGROUND

Nocturnal hypertension (NH) is associated with adverse cardiovascular outcomes beyond and even independent of daytime hypertension (DH). Although cohort studies have evaluated correlates of NH, there is comparably less data available from real-world clinical practice and for population subsets that tend to be under-represented in cohort studies.

METHODS

This retrospective cohort study included all patients who underwent ambulatory blood pressure monitor (ABPM) testing at a large US academic medical center from 1 January 2013 to 31 December 2023. We used multivariable-adjusted logistic regression to assess DH as a correlate of NH, covariates related to the co-occurrence of DH and NH, and correlates of isolated NH.

RESULTS

Of 1,566 patients, 812 (51.9%) had DH, 1,125 (71.8%) had NH, and 363 (23.2%) had isolated NH. A total of 762 (48.7%) patients had co-occurring daytime and NH. In multivariable analysis, significant correlates of NH included DH, male sex, age, Black race, and Hispanic ethnicity. By comparison, significant correlates of co-occurrent DH and NH included male sex, age, Asian race, Black race, and renal disease; coronary artery disease (CAD) was inversely associated with this co-occurrence. Among all covariates, only CAD was associated with isolated NH.

CONCLUSIONS

Our real-world study results highlight the generally under-recognized prominence of isolated NH, as well as the presence of NH among Hispanic and Asian-American populations. Further prospective investigations are needed to evaluate whether broader ABPM screening for NH is needed across all populations at risk, including but not limited to persons with more easily identified DH.

Graphical Abstract

Graphical Abstract.

Nocturnal hypertension (NH) is associated with adverse cardiovascular outcomes beyond and even independent of daytime hypertension (DH).¹ Whereas DH is more easily diagnosed via clinic- or self-measured blood pressure (BP), NH is rarely diagnosed in the absence of ambulatory blood pressure monitor (ABPM) use, an out-of-office BP monitoring technique utilized by a minority of physicians’ offices. Although cohort studies have evaluated correlates of NH, there is comparably less data available from real-world clinical practice and for population subsets that tend to be under-represented in cohort studies.²

METHODS

We conducted a retrospective cohort study of all patients who underwent ABPM testing at a large US academic medical center from 1 January 2013 to 31 December 2023. We excluded ABPMs with incomplete data. We curated from the electronic health record (EHR) data on demographic characteristics (age, sex, race, and ethnicity) and cardiovascular comorbidities (diabetes mellitus, coronary artery disease [CAD], heart failure, renal disease, history of myocardial infarction, and history of hypertension) documented via ICD-10 coding at the time of testing. Using contemporary US guidelines, we classified DH as a mean daytime systolic BP ≥ 130 mmHg or diastolic BP ≥ 80 mmHg, and NH as a mean evening systolic BP ≥ 110 mmHg or diastolic BP ≥ 65 mmHg.³ Isolated NH was defined as the presence of NH, yet with normal daytime BP. We used multivariable-adjusted logistic regression to assess DH as a correlate of NH, covariates related to the co-occurrence of DH and NH, and correlates of isolated NH. We conducted sensitivity analyses using European ABPM definitions for DH (mean daytime systolic BP ≥ 135 mmHg or diastolic BP ≥ 85 mmHg) and NH (mean evening systolic BP ≥ 120 mmHg or diastolic BP ≥ 70 mmHg).⁴ All statistical analyses were conducted using STATA v18, with a two-sided P < 0.05 considered significant. Institutional Review Board approval was provided by Cedars-Sinai Medical Center, with a waiver of informed consent. Analysis and manuscript preparation followed STROBE guidelines.

RESULTS

Of the 1,619 patients who completed ABPMs during the 11-year study period, 53 were excluded due to incomplete data. For the remaining 1,566 patients, 734 (46.9%) were male, 1061 (67.8%) were White, 125 (8.0%) were Black, 195 (12.5%) were Hispanic, 88 (5.6%) were Asian, and 1118 (71.4%) had a history of hypertension (Figure 1, Panel A). In non-mutually exclusive groups, there were 812 (51.9%) patients with DH, 1,125 (71.8%) with NH, and 363 (23.2%) with isolated NH; 104 (83.2%) Black patients and 148 (75.9%) Hispanic patients had NH. A total of 762 (48.7%) patients had co-occurrent DH and NH, and 93.8% of patients with DH also had NH. In multivariable analysis, significant correlates of any NH included DH, male sex, age, Black race, and Hispanic ethnicity (Figure 1, Panel B). By comparison, significant correlates of co-occurrent DH and NH included male sex, age, Asian race, Black race, and history of renal disease; CAD was inversely associated with this co-occurrence (P < 0.01 for all, Figure 1, Panel C). In analyses of isolated NH, the only significant correlate was CAD (odd ratio 1.60; 95% confidence interval [1.13, 2.25]). In sensitivity analyses using European definitions for BP thresholds, similar findings were observed with the following exceptions: Hispanic ethnicity was associated with co-occurrent DH and NH (1.61 [1.13, 2.29]) and CAD was not associated (P = 0.38); male sex (1.60 [1.24, 2.07]) and age (1.10 [1.03, 1.19]) were associated with isolated NH, while CAD was not associated (P = 0.42).

Figure 1.

Cohort characteristics, predictors of nocturnal hypertension, and predictors of co-occurrence of nocturnal hypertension and daytime hypertension. Cohort characteristics (Panel A), correlates of nocturnal hypertension (Panel B), and correlates of co-existing daytime and nocturnal hypertension (Panel C). Abbreviations: BP, blood pressure; SD, standard deviation; mmHg, millimeters of mercury; DH, daytime hypertension; NH, nocturnal hypertension; CAD, coronary artery disease; mi, myocardial infarction. *Percentages in the total column represent the percent of the total population. Percentages in all remaining rows are calculated across rows. ^†Daytime hypertension, nocturnal hypertension, and co-occurrent daytime and nocturnal hypertension columns are not mutually exclusive. ^‡Age is per 10 years.

DISCUSSION

In a real-world cohort of patients who underwent ABPM, over 90% of patients with DH had NH, and DH was the strongest predictor of NH. However, nearly a quarter of patients also had isolated NH, a diagnosis typically masked by normal daytime BPs. While CAD was associated with isolated NH, this relationship was no longer significant in the sensitivity analysis. Further, CAD was found to be inversely associated with co-occurrent DH and NH, though again it was no longer significant in sensitivity analysis. Our analyses suggest that CAD was not a reliable correlate of NH and underscore the importance of further studies into predictors of isolated NH.

The high prevalence of NH among Black patients, as well as the association between Black race with NH, was similar to findings from large epidemiologic cohorts.⁵ Our findings of the high prevalence of NH among Hispanic patients and the association of Hispanic ethnicity with NH are relatively under-reported, as Hispanic patients have been under-represented in large hypertension studies. In both groups, proposed mechanisms for hypertension include treatment differences, comorbidities, and socioeconomic factors.^6,7 Notably, approximately 80% of Asian patients had NH, and Asian race was the strongest predictor of co-occurrent DH and NH. Although prior studies conducted in Asia have found NH to be highly prevalent, relatively little data on NH have been available for Asian Americans and in comparison to other racial or ethnic groups in the United States.⁸ Underlying mechanisms of NH may be related to elevated salt intake and sensitivity among certain Asian populations, predisposing to increased nighttime circulating blood volume and subsequent NH.⁹

Strengths of this study included its large sample size and reliable BP data obtained using validated ABPM devices. Limitations include the single-center nature of the study, and referral sample of patients comprising a higher-risk cohort, albeit likely similar to other real-world clinical practice cohorts. Additionally, reliance on EHR-derived data implies some variation in the definitions used to document comorbidities. Other important factors which may impact cardiovascular risk, such as dipping status, were not able to be included.¹⁰ Finally, reliable data regarding anti-hypertensive medications were not available and may result in bias; however, the association between DH and NH remains.

In summary, our real-world study results highlight the generally under-recognized prominence of NH among Hispanic and Asian-American populations. This latter finding may be particularly relevant for understanding the high prevalence of stroke in Asians, given that NH appears even more strongly associated with stroke risk than DH.¹¹ Further prospective investigations are needed to evaluate whether broader ABPM screening for NH is needed across all populations at risk, including but not limited to persons with more easily identified DH.

FUNDING

This work was supported in part by National Institutes of Health grants K23-HL153888 and T32-HL116273. No funders had a role in the design/conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

CONFLICT OF INTEREST

None declared.

AUTHOR CONTRIBUTIONS

Neil Zhang (Conceptualization, Methodology, Software, Formal analysis, Writing—original draft, Writing—review & editing, Visualization), Tzu Yu Huang (Data curation, Software), Sanket Dhruva (Methodology, Writing—review & editing), Susan Cheng (Visualization, Writing—review & editing), Joseph Ebinger (Conceptualization, Methodology, Writing—review & editing, Visualization, Supervision)

DATA AVAILABILITY

The data underlying this article will be shared on reasonable request to the corresponding author. N.Z. and J.E. had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.