Should Bedtime be Med-Time for Hypertension?

Author(s)
Vivian Cheng, PharmD, BCPS
Joseph J. Saseen, PharmD, BCPS, BCACP

Reviewed By
Daniel Longyhore, PharmD, BCACP
Eric MacLaughlin, PharmD, BCPS

Citation
Hermida RC, Crespo JJ, Domínguez-Sardiña M, et al. Bedtime hypertension treatment improves cardiovascular risk reduction: the Hygia Chronotherapy Trial. Eur Heart J. 2019. pii:ehz754. doi: 10.1093/eurheartj/ehz754

Please note that in April 2020 the editors of the European Heart Journal released a “Issue of Concern” statement regarding the HYGIA study. An investigation regarding the conduct and results of the HYGIA study have been initiated and readers have been advised to interpret the study with caution.

The Problem

Few medications are specifically dosed at night to optimize outcomes. However, antihypertensive medications may soon belong on our patients’ nightstands. It is common practice for patients to take all antihypertensive medications in the morning, but perhaps daytime dosing doesn’t maximize cardiovascular risk reduction? Chronotherapy involves dosing medication at a time that targets the body’s natural circadian rhythms. The typical diurnal blood pressure (BP) pattern is rising values starting in the early morning before awakening, culminating in a mid-morning peak, followed by a gradual decline over the rest of the day. BP values are typically lowest during sleep, a phenomenon called “dipping.”¹ Approximately 50% of patients with essential HTN are “non-dippers” and have elevated nighttime ambulatory blood pressures (ABP) and an increased risk for fatal and non-fatal cardiovascular (CV) events.^2,3,4 The potential benefits of chronotherapy and its impact on BP and CV outcomes have been investigated since the 1980s.⁵ However, current practice guidelines do not explicitly recommend dosing antihypertensive medications at bedtime.^6-7

What’s Known

In a small post-hoc analysis of the HOPE study with patients with 24-hour ambulatory BP monitoring (ABPM) randomized to placebo or nighttime-dosed ramipril, there was no difference in the office BP measurements (OBPM) between groups but mean 24-hour and nighttime BP reductions were larger with ramipril.⁸ The CONVINCE trial evaluated whether using a novel dosage form of verapamil at night versus morning dosing of atenolol or hydrochlorothiazide reduced CV events. Although 16,602 patients with HTN were enrolled in this prospective, randomized, double-blind trial, it was stopped early for futility.⁹ Verapamil and atenolol are unacceptable first-line HTN agents, so application of these results is problematic.¹⁰

More recently, the MAPEC (Monitorización Ambulatoria para Predicción de Eventos Cardiovasculares) trial prospectively compared dosing at least 1 antihypertensive medication at bedtime versus all antihypertensive medications taken in the morning in Spanish patients with HTN.² Significantly fewer patients in the bedtime dosing group were non-dippers (34% vs 62%, p<0.001) and more patients had controlled 24-hour ABP (62% vs 53%, p<0.001). Surprisingly, the risk of CV events was dramatically reduced in the bedtime treatment group (RR 0.39; 95% CI 0.29-0.51, p<0.001) when compared to morning dosing. With mixed results and different approaches to chronotherapy, there is a lack of clarity about whether this approach should be widely adopted.

What’s New

The Hygia Chronotherapy trial was a multicenter, controlled, prospective, randomized, open-label, blinded endpoint (PROBE) study conducted between 2008 and 2018 across 40 primary care centers in Spain.¹¹ Patients with HTN were randomized to take at least 1 antihypertensive medications at bedtime (n=9,552) or all antihypertensive medications upon awakening (n=9,532). Confirmation of HTN using 48-hour ABP criteria was required and patient were required to meet at least 1 of the following criteria: 1) mean awake systolic blood pressure (SBP) ≥135 mmHg, 2) mean awake diastolic blood pressure (DBP) ≥85 mmHg, 3) mean asleep SBP ≥120 mmHg, 4) mean asleep DBP ≥70 mmHg, or prescription(s) for antihypertensive agents. Angiotensin receptor blockers (ARBs), angiotensin converting enzyme inhibitors (ACEIs), calcium channel blockers (CCBs), beta-blockers (BB), and diuretics were used alone or in combination as first-line treatments. Fixed-dose combination treatments were prescribed when available.

At follow-up visits, at least 3 consecutive OBPMs were taken and adherence was assessed using the Morisky-Green test. After the clinic visit and at least once annually, ABP measurements were recorded every 20 minutes between 07:00 and 23:00 and every 30 minutes overnight for 48 consecutive hours (48-hour ABPM). Medication adjustments were made based on the 48-hour ABPM results and were instituted if awake BP was greater than 135/85 mmHg or asleep BP was greater than 120/70 mmHg, regardless of OBPM.

The primary endpoint was the composite of MI, coronary revascularization, heart failure (HF), ischemic stroke, hemorrhagic stroke, and CV death. Secondary endpoints were stroke, coronary events (CV death, MI, coronary revascularization), and cardiac events (coronary events, HF). Baseline characteristics between groups were similar (Table 1).

Table 1: Mean Baseline Characteristics of Patients:

Characteristic	Total (n=19,084)	Awakening Dosing (n=9,552)	Bedtime Dosing (n=9,532)
Median age, yr	60.5	60.5	60.6
Male sex, %	55.6	56.2	55.0
Body-mass index, kg/m2	29.7	29.6	29.7
Obstructive sleep apnea, %	4.1	4.2	3.9
Smoking, %	15.2	15.6	14.8
Prior HTN treatment, %	57.4	57.9	56.9
Previous CV events, %	10.4	10.8	10.0
Mean Clinical Laboratory Tests
Estimated glomerular filtration rate (eGFR), mL/min/1.73m2	79.0	78.6	79.3
HDL cholesterol, mg/dL	53.0	52.8	53.1
LDL cholesterol, mg/dL	123.8	123.9	123.8
Mean BP, mmHg
Office SBP	149.4	149.4	149.5
Office DBP	86.1	86.3	86.0
Awake SBP	136.0	136.1	135.9
Asleep SBP	123.6	123.3	123.7
48-hr SBP	131.6	131.4	131.7
Non-dipper, %	49.3	49.0	49.5

The most frequent monotherapies used were ARBs (valsartan, telmisartan) or ACEIs (enalapril, ramipril) and CCBs (mostly amlodipine). The most common combinations were ACEI or ARB with a diuretic, most commonly hydrochlorothiazide. At study conclusion, there were small (but statistically significant) differences in the number of prescribed antihypertensive medications and distribution of specific antihypertensive classes (Table 2).

Table 2: Final Mean Characteristics After Study Conclusion

Characteristic	Awakening Dosing (n=9,552)	Bedtime Dosing (n=9,532)	P-value
HTN medications, no.	1.80	1.71	<0.001
HTN treatment, %
ARB	53.1	53.1	NS
ACEi	25.3	23.4	0.002
CCB	32.7	36.8	<0.001
BB	22.0	17.5	<0.001
Diuretic	46.5	39.5	<0.001
Mean Clinical Laboratory Tests
eGFR, mL/min/1.73m2	75.7	79.3	<0.001
HDL-cholesterol, mg/dL	51.8	53.0	<0.001
LDL-cholesterol, mg/dL	120.7	118.2	0.002
Mean BP, mmHg
Office SBP	143.2	140.0	<0.001
Office DBP	82.4	81.4	<0.001
Awake SBP	129.5	129.2	NS
Asleep SBP	118.0	114.7	<0.001
48-hr SBP	125.6	124.3	<0.001
Non-dipper, %	50.3	37.5	<0.001

During the median follow-up of 6.3 years, the primary endpoint occurred in 1,752 subjects. Patients in the bedtime group had a 45% relative reduction in the primary endpoint compared to the upon awakening group (see Table 3).

Table 3: Primary and Secondary Outcomes

	Adjusted HR for Bedtime Dosing	95% CI
Total CV events	0.57	0.53-0.62
Primary Endpoint	0.55	0.50-0.61
Secondary Endpoints
Stroke	0.51	0.41-0.63
Coronary events	0.56	0.49-0.64
Individual Endpoints
Total death	0.55	0.48-0.63
CV death	0.44	0.34-0.56
Ischemic stroke	0.54	0.42-0.69
Hemorrhagic stroke	0.39	0.23-0.65
Myocardial infarction	0.66	0.52-0.84
Coronary revascularization	0.60	0.47-0.75
Heart failure	0.58	0.49-0.70
Peripheral artery disease	0.52	0.41-0.79

Total CV events: CV death, MI, coronary revascularization, HF, stroke, angina pectoris, PAD, TIA; Primary endpoint: CV death, MI, coronary revascularization, HF, stroke; Coronary events: CVD, MI, coronary revascularization

Hazard ratios were adjusted for age, sex, type 2 diabetes mellitus, chronic kidney disease, smoking, HDL-cholesterol, previous CV event, mean asleep SBP, and sleep-time relative SBP decline. All subgroup analyses and other endpoints favored bedtime dosing except for TIA, which did not achieve statistical significance. There were no differences in adverse effects between bedtime (6.0%) vs upon awakening dosing (6.7%), and the incidence of poor adherence was comparable (<3.0%). Only 0.3% of patients experienced sleep-time hypotension.

Our Critical Appraisal

The Hygia Chronotherapy trial was the largest ABPM-based study to compare bedtime versus upon awakening dosing of HTN medications and demonstrated significant CV benefits. The prospective, multicenter, randomized study design was ideal.  Both genders were well-represented. The duration of follow-up was long. The use of 48-hour ABPM provided more data than a typical 24-hour ABPM. The antihypertensive medications used in the study were reasonable and reflect current evidence-based guideline recommendations, except the use of beta-blockers.

Several factors limit this trial’s generalizability. This study enrolled exclusively Caucasian Spanish patients. Additionally, adherence rates were suspiciously high and do not reflect real-world practice in the United States. The Morisky-Green adherence test is a four-item patient questionnaire using closed-ended yes-no questions. Thus, medication adherence was likely overestimated. Objective adherence measures (e.g., proportion of days covered) were not assessed.

To replicate the benefits of Hygia in clinical practice, providers would need to monitor and adjust BP medications using 48-hour ABPM and disregard OBPM. ABPM is not commonly used in primary care and 24-hour ABPM is more commonly performed in the United States.  While some payors are now reimbursing for ABPM for specific indications (white coat or masked HTN), it is not widely available.^12,13 Other limitations included an open-label design and lack of algorithms for medication selection and dose adjustments. Importantly, there was no guidance for determining which antihypertensive medication(s) were selected to be dosed at bedtime.  Furthermore, it is unclear why unadjusted hazard ratios were not provided or whether the adjusted hazard ratios included dipper status.  

Finally, the surprisingly large CV benefits relative to the small differences in BP is notable. Mean SBPs and DBPs in the bedtime group were a mere 0.3–3.3 mmHg lower when compared to the morning dosing group but this translated into a 45% difference in the primary endpoint. The CV benefits with bedtime dosing in Hygia are difficult to explain based on BP reductions alone. Although previous landmark HTN trials did not evaluate bedtime dosing, the magnitude of benefit in Hygia is clearly an outlier.¹⁴ Thus, the dipping effect from bedtime dosing may have disproportionately contributed to the observed benefits. A 2014 meta-analysis of five HTN chronotherapy trials found that the relative risk for CV events with bedtime dosing was 0.63 (95% CI 0.43-0.92, p=0.016), which is similar to the Hygia findings.¹⁵ Improvements in renal function, LDL-cholesterol, and HDL-cholesterol were also seen with bedtime dosing, but changes were small and unlikely to explain the differences in CV event rates between groups.

The Bottom Line

The Hygia Chronotherapy trial found that dosing at least 1 antihypertensive medication at bedtime significantly reduced CV events compared to dosing all antihypertensive medications upon awakening.  However, the striking reductions in CV event rates are hard to explain and several questions remain. Would other populations benefit as dramatically? Does it matter which antihypertensive medications are dosed at bedtime? Would shift workers with alternating sleep-wake schedules derive similar benefits? The Hygia Chronotherapy trial is well-designed with results supporting bedtime BP medication dosing, however additional studies in a more diverse population are needed to validate this potential paradigm change in clinical practice.

 

The Key Points

• In Caucasian Spanish patients with hypertension, defined and monitored using ambulatory blood pressure monitoring, taking at least 1 antihypertensive medication at bedtime significantly reduced the risk for CV events when compared to taking all antihypertensive medications upon awakening, without increasing the rate of side effects.
• The study has low external validity due to the population studied, unusually high adherence rates, and use of 48-hour ABPM.
• It is reasonable to recommend patients to take at least 1 antihypertensive medication at bedtime, but it is unclear which and how many medications should be dosed this way.

FINAL NOTE:  This program will be available for recertification credit through the American Pharmacists Association (APhA) Ambulatory Care Review and Recertification Program.  To learn more, visit https://www.pharmacist.com/ambulatory-care-review-and-recertification-activities.