Phenotype‐Dependent Effect of Bedtime Antihypertensive Dosing on Nocturnal Blood Pressure: A Meta‐Analysis Guided by Non‐Dipper Prevalence

ABSTRACT

Bedtime dosing of antihypertensive medications is associated with a reduction in nocturnal blood pressure (BP), but whether this effect is modified at the study level remains uncertain. We conducted an updated meta‐analysis of randomized controlled trials comparing bedtime versus morning antihypertensive administration, evaluating the overall treatment effect and its potential modification by baseline non‐dipper prevalence. PubMed, EMBASE, and the Cochrane Library were searched, with nighttime ambulatory systolic BP (SBP) and diastolic BP (DBP) as primary outcomes. Subgroup analyses were performed based on baseline non‐dipper proportion (< 60% vs. ≥ 60%).

Sixty‐six RCTs involving 29,654 participants were included. In the overall hypertensive population, bedtime dosing significantly reduced nighttime SBP (WMD: 4.87 mmHg; 95% CI: 3.75–6.00) and DBP (WMD: 2.80 mmHg; 95% CI: 2.23–3.37), with modest reductions in 24 h SBP (WMD: 2.06 mmHg) and daytime SBP (WMD: 0.99 mmHg). In trials restricted to non‐dippers, a more pronounced reduction in nighttime BP was observed (SBP WMD: 7.57 mmHg; DBP WMD: 3.77 mmHg). Similarly, study populations with a higher baseline proportion of non‐dippers (≥ 60%) experienced approximately two‐fold higher nighttime SBP reductions versus those with lower prevalence (< 60%) (WMD: ∼9.7 vs. 4.4 mmHg; p < 0.01), with a consistent pattern for nighttime DBP (5.1 vs. 2.6 mmHg; between‐group p < 0.05).

Bedtime antihypertensive dosing reduces nocturnal BP across the general hypertensive population; however, the magnitude of effect appears to be modified at the study level by baseline non‐dipper prevalence. These findings suggest that chronotherapy effects are heterogeneous across trial populations rather than uniform across all hypertensive individuals.

1. Introduction

Hypertension management has shifted from a focus on clinic blood pressure reduction toward sustained 24 h BP control, particularly during the nighttime [1, 2]. Ambulatory blood pressure monitoring (ABPM) studies consistently demonstrate that nighttime BP and circadian BP patterns—especially the non‐dipping phenotype—are stronger predictors of target organ damage and cardiovascular outcomes than daytime or office BP [2, 3, 4]. Individuals with impaired nocturnal BP decline remain at elevated risk despite similar mean BP levels [5, 6].

Given the circadian variation of BP, the timing of antihypertensive drug administration has been proposed as a potential strategy to improve 24 h BP control [7, 8]. A recent meta‐analysis published in Hypertension synthesized 72 randomized trials and confirmed that bedtime dosing is associated with a reduction in average nocturnal BP compared with morning administration in unselected hypertensive populations [9]. However, that analysis primarily focused on overall treatment effects in unselected hypertensive populations and did not specifically evaluate between‐study heterogeneity in treatment effects according to nocturnal BP phenotype [9, 10]. In addition, several recently published randomized trials were not included in that synthesis [11, 12].

Whether chronotherapy exerts phenotype‐dependent effects—particularly across populations with differing prevalence of non‐dipping patterns—therefore remains uncertain [2, 10, 13].

We conducted an updated systematic review and meta‐analysis of randomized controlled trials comparing bedtime versus morning antihypertensive dosing. In addition to incorporating newly published trials, we specifically evaluated the interaction between dosing time and nocturnal BP phenotype to determine whether nocturnal BP phenotype is associated with between‐study heterogeneity in the effect of bedtime versus morning dosing on nighttime BP reduction across randomized trials.

2. Methods

This meta‐analysis adhered to standard guidelines and was written in accordance with the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) statement [14]. It is registered in PROSPERO (CRD42022304270).

2.1. Data Sources and Searches

We conducted a comprehensive search across PubMed, EMBASE, and the Cochrane Library from inception to November 22, 2024, using a structured search strategy encompassing three core concepts: “chronotherapy,” “hypertension,” and “randomized controlled trials (RCTs).” Both Medical Subject Headings (MeSH) and relevant keywords were applied in the search. To further ensure completeness, reference lists from included studies and relevant reviews were manually screened, and experts in the field were consulted to identify additional publications.

2.2. Eligibility Criteria

This meta‐analysis aimed to evaluate the comparative effects of evening/bedtime versus morning dosing of antihypertensive medications on ambulatory BP profiles, with a focus on nocturnal BP phenotypes as a study‐level characteristic, including variation in non‐dipper prevalence across trials. The inclusion criteria were restricted to randomized controlled trials that directly compared morning versus evening administration of antihypertensive drugs. Non‐randomized studies, investigations without a clear morning‐evening comparison, as well as reviews, editorials, and conference abstracts with insufficient data were excluded.

Study selection followed the PICO (Population, Intervention, Comparison, Outcome) framework:

Population: Adult patients diagnosed with hypertension.

Intervention: Evening or bedtime administration of one or more antihypertensive medications.

Comparison: Morning administration of the same antihypertensive medication(s).

Outcomes: The primary outcome was the mean change in nighttime ambulatory systolic blood pressure (SBP) and diastolic blood pressure (DBP) from baseline. Secondary outcomes included changes in 24 h and daytime ambulatory SBP and DBP. Furthermore, we specifically predefined the baseline proportion of non‐dippers (≥ 60% vs. < 60%) as a key exploratory variable to evaluate its potential effect modification of dosing time on nighttime BP reduction across randomized trials. The cutoff of 60% was selected based on both clinical and statistical considerations, including the expected prevalence of non‐dipper phenotype in general hypertensive populations (approximately 30%–40%) and the need to ensure a clearly enriched subgroup with predominant nocturnal BP dysregulation while maintaining a relatively balanced distribution of included trials across subgroups.

2.3. Data Extraction and Critical Appraisal

The data were extracted by two authors working independently using a predefined, standardized protocol and data collection instrument. Information was recorded on study design, demographic characteristics, BP values, and antihypertensive therapies reported in the trials. Any discrepancies were resolved by consensus among the authors.

2.4. Risk of Bias Assessment

We evaluated risk of bias using the revised Cochrane Risk of Bias Tool [15]. This tool has five domains (randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported results) and the overall score. Each domain can be judged as follows: low risk of bias, some concerns, and high risk of bias (Table S1).

2.5. Statistical Analysis

All analyses were performed using Stata (version 17.0) and RevMan (version 5.4). For continuous outcomes (BP parameters), the pooled weighted mean difference (WMD) with 95% confidence intervals (CI) was calculated. For dichotomous outcomes, the pooled odds ratio (OR) with 95% CIs was computed. All meta‐analyses employed a random‐effects model using the DerSimonian and Laird method to account for anticipated clinical and methodological heterogeneity. Heterogeneity was quantified using the I ² statistic, interpreted as follows: 25%–49% low, 50%–74% moderate, and ≥ 75% high heterogeneity [16]. Pre‐specified subgroup analyses were conducted to explore potential sources of between‐study heterogeneity and variability in pooled treatment effect estimates.

The robustness of the findings was evaluated through leave‐one‐out sensitivity analyses and influence diagnostics. Influence diagnostics included assessment of studies with disproportionate impact on pooled estimates. The sensitivity analyses involving exclusion of influential studies were not prespecified in the original protocol and were conducted post hoc as an exploratory assessment of robustness. The pooled estimates were interpreted as average treatment effects across heterogeneous trial settings rather than uniform effects applicable to all hypertensive patients. Publication bias was assessed visually using funnel plots for the outcomes and statistically using Egger's regression asymmetry test. A two‐tailed p‐value < 0.05 was considered statistically significant.

3. Results

3.1. Study Selection and Characteristics

The literature search process and its results are shown in Figure S1. The initial database search identified 7237 records. After removing duplicates and screening titles and abstracts, 118 full‐text articles were assessed for eligibility. Ultimately, 66 randomized controlled trials were included in the quantitative synthesis. This total comprised 3 newly identified studies added in the present update [12, 17, 18], while 3 studies that had been included in previous analyses were excluded because they did not meet the eligibility criteria [11, 19, 20]. The included studies investigated the effects of morning versus evening dosing on ambulatory BP and clinical outcomes.

3.2. Baseline Characteristics

66 studies evaluated the impact of dosing time on ambulatory BP, involving a total of 29 654 participants. Among these trials, 33.3% (22 out of 66) utilized a cross‐over design. The mean age of the participants was 55.1 years, and 53.1% of the individuals were male. Only 9.1% (6 out of the 66 trials) involved short‐acting antihypertensive agents, while 21.2% (14 studies) used a combination of antihypertensive medications. The median follow‐up duration was 3 months. Of the 66 included trials, 12 were restricted to non‐dipper patients, involving 686 participants.

The baseline characteristics of the study populations are detailed in Table S2. Standards for measuring parameters of daytime, nighttime, 24 h, or 48 h mean ambulatory BP, as well as the methodologies for 24/48 h BP monitoring used in the included trials, are summarized in Table S3.

3.3. Effect of Dosing Time in all Hypertensive Patient

3.3.1. 24 H BP

Evening dosing was associated with a reduction in both 24/48 h ambulatory systolic BP (WMD = 2.06 mmHg; 95% CI, 1.33 to 2.79; I ² = 90.6%; 61 trials) and diastolic BP (WMD = 0.69 mmHg; 95% CI, 0.22 to 1.17; I ² = 89.3%; 60 trials) compared to morning dosing (Figure 1).

FIGURE 1.

The 24/48 h ambulatory blood pressure in all hypertensive patients (A), Systolic blood pressure (SBP) and (B) diastolic blood pressure (DBP). Kohno^# et al. (11) is a single trial with 2 different groups (dippers vs. nondippers) analyzed separately. Hermida C^* (14) is a single trial with 2 different groups (monotherapy vs. polytherapy groups) analyzed separately. Gorbunov ⁺ et al. (35) is a single trial with 2 different antihypertensives used (ramipril and verapamil) analyzed separately.

3.3.2. Daytime BP

Significant reductions in daytime systolic (WMD = 0.99 mmHg; 95% CI, 0.29 to 1.70; I ² = 88.9%; 68 trials) and diastolic BP (WMD = 0.68 mmHg; 95% CI, 0.21 to 1.15; I² = 87.1%; 68 trials) were observed with evening dosing (Figure S2).

3.3.3. Nighttime BP

Evening dosing led to a significantly greater reduction in both systolic (WMD = 4.87 mmHg; 95% CI, 3.75 to 6.00; I ² = 95.8%; 67 trials) and diastolic BP (WMD = 2.80 mmHg; 95% CI, 2.23 to 3.37; I ² = 91.7%; 67 trials) compared to morning dosing (Figure S3).

3.4. Effect of Dosing Time in Non‐Dipper Hypertensive Patients

3.4.1. 24 H BP

In trials enrolled non‐dipper patients only, there were no statistically significant differences in the reduction of 24 h ambulatory systolic (WMD = 3.30 mmHg; 95% CI, −1.59 to 8.20; I ² = 96.3%; 10 trials) or diastolic BP (WMD = 2.37 mmHg; 95% CI, −0.32 to 5.05; I ² = 92.6%; 9 trials) (Figure 2).

FIGURE 2.

The 24/48 h ambulatory blood pressure in non‐dipper hypertensive patients (A), Systolic blood pressure (SBP) and (B) diastolic blood pressure (DBP). Kohno^# et al. (11) is a single trial with 2 different groups (dippers vs nondippers) analyzed separately.

Substantial heterogeneity was observed in the overall analysis. A sensitivity analysis excluding two outlier studies [18, 21] yielded consistent results, with the effect on 24 h systolic BP remaining non‐significant (WMD = 2.23 mmHg; 95% CI, −0.57 to 5.03; I ² = 76.3%; 8 trials) (Figure S4A), and similarly for 24 h diastolic BP (WMD = 1.28 mmHg; 95% CI, −0.60 to 3.17; I ² = 69.4%; 7 trials) (Figure S4B).

3.4.2. Daytime BP

For trials restricted to non‐dipper participants, daytime systolic (WMD = 3.52 mmHg; 95% CI, −1.16 to 8.20; I ² = 96%; 12 trials) and diastolic BP (WMD = 1.70 mmHg; 95% CI, −0.62 to 4.03; I ² = 90.4%; 12 trials) were not significantly reduced with evening dosing compared to morning dosing (Figures 3).

FIGURE 3.

Day‐time ambulatory blood pressure in non‐dipper hypertensive patients (A), Systolic blood pressure (SBP) and (B) diastolic blood pressure (DBP). Kohno^# et al. (11) is a single trial with 2 different groups (dippers vs. nondippers) analyzed separately.

Substantial heterogeneity was observed across the studies. A sensitivity analysis, excluding two studies with high heterogeneity [18, 22], yielded consistent results: the reduction in daytime systolic BP remained non‐significant (WMD = 0.64 mmHg; 95% CI, −1.38 to 2.67; I ² = 47.8%; 10 trials) (Figure S5A), as did the reduction in daytime diastolic BP (WMD = −0.02 mmHg; 95% CI, −1.15 to 1.12; I ² = 26.9%; 10 trials) (Figure S5B).

3.4.3. Nighttime BP

In trials restricted to non‐dipper patients, evening dosing was associated with greater reductions in nighttime systolic BP (WMD = 7.57 mmHg; 95% CI, 1.08 to 14.06; I ² = 99.1%; 12 trials) and diastolic BP (WMD = 3.77 mmHg; 95% CI, 0.48 to 7.05; I ² = 95.9%; 12 trials) than morning dosing. These estimates, however, were based on a limited number of trials and should be interpreted with caution due to reduced statistical precision (Figures 4).

FIGURE 4.

Night‐time ambulatory blood pressure in non‐dipper hypertensive patients (A), Systolic blood pressure (SBP) and (B) diastolic blood pressure (DBP). Kohno^# et al. (11) is a single trial with 2 different groups (dippers vs. nondippers) analyzed separately.

Substantial heterogeneity was observed in the initial analysis. A sensitivity analysis excluding three outlier studies [18, 22, 23] yielded consistent conclusions: evening dosing continued to show a significant reduction in both nighttime systolic BP (WMD = 7.22 mmHg; 95% CI, 5.31 to 9.13; I ² = 0%; 9 trials) (Figure S6A) and nighttime diastolic BP (WMD = 3.01 mmHg; 95% CI, 0.94 to 5.08; I ² = 51.1%; 9 trials) (Figure S6B).

3.5. Subgroup Analyses and Publication Bias

The results of our comprehensive subgroup analyses are detailed in Tables S4–S6 and Figures S7–S9. We evaluated several potential modifying factors, including the total number of participants, follow‐up duration, baseline proportion of non‐dippers (≥ 60% or < 60%, prespecified cutoff based on clinical and statistical considerations), type of antihypertensive medication, drug half‐life (long‐ vs. short‐acting), and the strictness of the definition of daytime/nighttime dosing (strict: single‐pill regimen; non‐strict: combination pills or multiple medications).

The main observation from subgroup analyses was that the association between dosing time and nighttime BP reduction varied according to the baseline proportion of non‐dippers across trials. Specifically, greater nighttime BP reductions were observed in studies with a higher proportion of non‐dipper participants (≥ 60%) compared with those with lower proportions (< 60%). In the subgroup of studies with ≥ 60% non‐dippers, the WMD in nighttime SBP was approximately 9.7 mmHg, whereas the WMD was 4.4 mmHg in the < 60% non‐dipper subgroup (p = 0.001 for daytime dosing subgroup; p = 0.002 for evening dosing subgroup). In addition to nighttime SBP, a statistically significant association was observed for nighttime DBP. The reduction in nighttime DBP was notably greater in subgroups with a higher proportion of baseline non‐dippers (5.1 mmHg vs. 2.6 mmHg, p < 0.001 for both daytime and evening dosing subgroups, with p < 0.05 between groups). Furthermore, consistent, though not statistically significant, trends were noted for 24 h SBP/DBP and daytime SBP/DBP, where a higher proportion of non‐dippers was generally associated with enhanced overall BP lowering particularly in the evening‐dosing group.

3.6. Assessment of Publication Bias

Egger's regression tests revealed statistically significant publication bias across all BP parameters in the overall hypertensive population (all p < 0.05). Specifically, significant bias was observed for 24/48 h systolic BP (bias coefficient = 1.71; 95% CI: 0.80 to 2.63; p < 0.001), 24/48 h diastolic BP (bias coefficient = 1.07; 95% CI: 0.16 to 1.98; p = 0.02), daytime systolic BP (bias coefficient = 1.46; 95% CI: 0.66 to 2.27; p < 0.001), daytime diastolic BP (bias coefficient = 1.22; 95% CI: 0.47 to 1.97; p < 0.001), nighttime systolic BP (bias coefficient = 1.87; 95% CI: 0.48 to 3.26; p = 0.01), and nighttime diastolic BP (bias coefficient = 2.02; 95% CI: 1.14 to 2.90; p < 0.001). In contrast, analyses restricted to non‐dipper hypertensive patients showed no significant publication bias for any BP outcome (all p > 0.05), indicating greater robustness of findings in this subgroup (Table S7). Funnel plot inspection supported the Egger's test results, demonstrating asymmetry suggestive of possible underrepresentation of small studies with null or smaller effect sizes (Figures S10 and S11).

4. Discussion

In this updated systematic review and meta‐analysis of randomized trials comparing bedtime versus morning administration of antihypertensive medications, we found that dosing at bedtime was associated with a reduction in ambulatory BP, with the most pronounced relative reduction observed during the nighttime period, whereas differences in daytime and 24 h BP were comparatively modest. This finding is consistent with the hypothesis that chronotherapy primarily modifies circadian BP patterns rather than overall BP burden [24, 25, 26]. This finding is consistent with prior evidence that bedtime dosing preferentially lowers nocturnal BP [24, 26, 27, 28].

A key contribution of the present study is the dedicated evaluation of nocturnal phenotype as an effect modifier. We found that the magnitude of nighttime BP lowering was not uniform across study populations. Compared with the overall hypertensive population, study populations restricted to non‐dippers or enriched for individuals with a non‐dipping pattern showed more pronounced reductions in nighttime BP with bedtime dosing. This profile suggests that bedtime dosing may function less as a universal intensification strategy and more as a potential population‐level strategy in cohorts with a higher prevalence of nocturnal BP abnormalities [2, 10, 29, 30]. Clinically, this aligns with the concept that non‐dipping represents a distinct risk phenotype and a potential risk marker [2, 31, 32, 33], and it supports prioritizing ABPM‐defined nocturnal phenotyping when considering dosing‐time interventions [2, 34]. From a clinical workflow standpoint, ABPM (or validated home nighttime BP strategies where applicable) becomes central —not merely to confirm control, but to identify populations in which dosing‐time effects are most pronounced [1, 35].

A recent meta‐analysis published in Hypertension confirmed an overall nocturnal BP advantage with bedtime dosing but largely summarized average effects in unselected hypertensive populations [9]. By updating the evidence base with several recently published randomized trials and by explicitly testing phenotype‐dependent effects, our work extends the prior literature in a clinically actionable direction. These findings support a study‐level interpretation of effect modification in chronotherapy and help reconcile why prior studies and syntheses have reached inconsistent conclusions when averaging treatment effects across heterogeneous populations [2, 9, 31, 34, 36, 37, 38, 39].

Contemporary hypertension management largely relies on long‐acting antihypertensive agents designed to provide sustained 24 h BP control, and current expert recommendations emphasize optimization of such regimens [8, 13]. Despite the widespread use of long‐acting therapies across included trials, bedtime dosing conferred an additional nocturnal BP reduction, particularly in populations with a higher prevalence of non‐dipping patterns. These observations suggest that, within the context of currently available long‐acting antihypertensive therapies, pharmacologic duration alone may not fully normalize nocturnal BP in all individuals, and that adjustment of dosing time may represent a potential strategy to optimize nocturnal BP control in such populations.

Several limitations merit consideration. First, heterogeneity across trials remained substantial for some endpoints, reflecting differences in study design, populations, and ABPM definitions of daytime/nighttime periods. This variability should be considered when interpreting the pooled estimates as average effects across heterogeneous trial settings. Second, many trials were relatively short, and ABPM‐based phenotype classification may vary over time, which may introduce misclassification and further contribute to between‐study heterogeneity. Third, background antihypertensive therapy patterns and drug classes differed across studies, which may have contributed to observed heterogeneity in treatment effects. In addition, sensitivity analyses involving exclusion of influential studies were conducted post hoc, although they support the robustness of the main findings, these studies may reflect genuine clinical heterogeneity rather than statistical noise. The analysis was not designed to determine the optimal drug class–specific timing strategy.

5. Conclusion

In this updated synthesis of randomized trials, bedtime dosing of antihypertensive medications was associated with greater nocturnal ambulatory BP reduction compared with morning dosing. This effect appeared to be modified at the study level by nocturnal BP phenotype, with bedtime dosing showing more pronounced nocturnal BP reductions in trials restricted to non‐dippers or in those with a higher prevalence of non‐dipping patterns. These findings suggest that the effects of chronotherapy are heterogeneous across study populations rather than uniform across all patients with hypertension, although conclusions regarding non‐dipper subgroups should be interpreted with caution given the limited sample size.

Author Contributions

Min Li did the literature search, accessed and verified the data, analyzed and interpreted the data, and wrote the original draft. Zuoyi Zhou conceptualized the study, did the literature search, accessed and verified the data, analyzed and interpreted the data, and wrote the original draft. Fangfang Fan wrote the original draft, conceptualized, and interpreted the data. Wei Ma, Jianping Li, and Yan Zhang conceptualized and interpreted the data

Funding

This study was funded by Beijing Natural Science Foundation, L246054.

Conflicts of Interest

There are no conflicts of interest.

Supporting information

Supporting File: jch70331‐sup‐0001‐SuppMat.docx.