Critical appraisal of recent translational chronopharmacology and chronotherapeutic reviews, meta-analyses, and pragmatic patient trials discloses significant deficiencies of design and conduct and suspect findings

ABSTRACT

The conduct of molecular and laboratory animal circadian rhythm research has increased exponentially in the past few decades, such that today investigations are being performed by scientists of many diverse disciplines. Knowledge gained from past works is now being explored for translational applications to clinical medicine, often termed “circadian medicine,” through the implementation of patient trials. However, these trials are being led, more often than not, by investigators who have little or no formal training and in-depth expertise in the methods of human circadian rhythm research, causing them to be deficient in design and produce dubious findings that have already led to unnecessary medical controversy at the expense of advances in patient care. Evidence of the very significant shortcomings of today’s translational circadian medicine research is exemplified in two recent publications in well-read reputable medical journals concerning the chronotherapy of blood pressure (BP) medications: one a review and meta-analysis by Maqsood et al. published in the journal Hypertension in 2023 that pertains to ingestion-time differences in the extent of BP reduction exerted by hypertensive medications and the other a report by Mackenzie et al. in the journal Lancet in 2022 that details the results of the pragmatic TIME study that assessed ingestion-time differences in cardiovascular disease outcomes. Herein, we appraise the inaccurate trial selection, lack of quality assessment, and the numerous other shortcomings that culminated in suspect findings and faulty conclusions of the former, as well as the deficiencies in design and conduct of the latter using as reference the eight items identified in 2021 by a working committee of the International Society for Chronobiology and American Association for Medical Chronobiology and Chronotherapeutics as being necessary for high-quality research of circadian rhythm-dependencies of the therapeutic effects of BP-lowering medications. The TIME study when rated for its quality according to the extent to which its investigational methods satisfy all of the eight recommended items attains a very low overall score of + 1 out of a possible range of −1 to + 7. Moreover, our review of the methods of the currently ongoing pragmatic BedMed trial discloses major deficiencies of the same sort rending a poor quality score of + 0.5. Although the focus of this article is the appraisal of the quality of contemporary circadian medicine hypertension chronotherapy research, it additionally exposes the inadequacies and dubious quality of the critique of such manuscripts submitted for publication to influential journals, in that some peer reviewers might also be deficient in the knowledge required to properly rate their merit.

KEYWORDS:

• Circadian medicine
• hypertension chronotherapy
• TIME study
• BedMed trial
• Hygia Chronotherapy Trial
• research quality

Introduction

The conduct of basic and translational circadian rhythm research has increased markedly during the 21^st century, in part stimulated by the award of the Nobel Prize in Physiology and Medicine to Jeffrey C. Hall, Michael Rosbash, and Michael W. Young for their pioneering work in elucidating their molecular mechanisms. Our PubMed search (19 July 2023) using “Circadian Rhythms” retrieved a total of 26,182 journal articles published during the past decade, with 13,859 retrieved using “Circadian Medicine.” It is clear that a vast number of investigators of diverse disciplines are presently researching circadian rhythms, including their relevance to medicine. Unfortunately, many investigators lack sufficient training in and competency of the methods of circadian medicine research. Inadequate understanding of the fundamental requirements, particularly for the design and conduct of quality clinical trials, can result in erroneous and inconsistent findings, resulting in risk for unnecessary disputes and medical controversies among the authors of those investigations at the expense of meaningful advances in the care of patients (Hermida et al. 2021f). This is exemplified by the great disparity in the rigor in which published recent reviews and clinical studies of administration-time-dependent differences in the effects of blood pressure (BP)-lowering medications have been done. Such disparity motivated the International Society for Chronobiology (ISC) and American Association for Medical Chronobiology and Chronotherapeutics (AAMCC) to appoint a working committee to develop guidelines for the design and conduct of future such human trials plus a scoring system to rate the quality and representativeness of the findings of previously reported ones. The guidelines and the developed scoring system were published in this journal in 2021 (Hermida et al. 2021f). It consists of eight items (Table 1) identified to be relevant for the proper design and implementation of circadian rhythm-based ingestion-time human hypertension medication clinical trials. For each one of these eight items, a numerical score of −0.5, 0, 0.5, or 1 is assigned depending on the extent to which it complies with the prescribed guidance (Table 1). Accordingly, the total score per trial can range between −1 and + 7, the higher is the total score, the higher is the quality of the trial.

Table 1. ISC/AAMCC recommendations for design and conduct of prospective hypertension chronopharmacology and chronotherapy trials plus scoring system for rating such reports.

Protocol item	Statement of procedure	Score
Specific chronopharmacology and chronotherapy trial criteria
1. Inclusion criteria for patient recruitment	Reliance on wake-time OBPM or wake-time HBPM to certify subjects as arterial hypertensive	−0.5
	Recruitment primarily of properly controlled treated patients according to diagnostic thresholds for hypertension	−0.5
	Failure to require as inclusion criterion that participants have a confirmed similar activity/sleep routine	0
	Recruitment of hypertensive patients according to ABPM diagnostic thresholds (0.5), all with similar activity/sleep routine (0.5)	1
2. Tested treatment times	Arbitrary clock-time or spans of the 24 h (e.g. 08:00 h vs. 20:00 h, morning vs. evening, etc.)	0
	Surrogate markers of endogenous biologic-time (e.g. upon awakening vs. at bedtime of sleep/wake routine)	1
Specific BP measurement criteria
3. Primary BP measurement method	Wake-time OBPM	0
	Wake-time HBPM	0
	Around-the-clock ABPM by a validated and calibrated device	1
4. Primary ambulatory BP and other endpoints	24 h mean or daytime/awake BP mean	0
	Nighttime/asleep BP mean and/or sleep-time relative decline calculated using fixed common clock-time span across all subjects	0
	Dipping based on sleep-time relative decline derived from MAP or DBP, instead of only SBP	0
	Asleep SBP/DBP means calculated using the individualized activity/rest cycle per subject and SBP dipping (only if sleep-time relative SBP decline properly calculated from awake/asleep SBP means)	1
5. Intended ABPM duration	<48 h	0
	≥48 h	0.5
6. Calculation of ABPM-based mean values	Unadjusted (arithmetic mean) procedure	0
	Adjusted procedure	0.5
General criteria applicable to any clinical trial
7. Sample size calculation	Not done or mistakenly done	0
	Properly calculated	1
8. Study design	Non-randomized or crossover open-label without washout period	0
	Crossover (with sufficient washout period of ≥2 weeks) or randomized (PROBE or double-blind)	1

AAMCC: American Association for Medical Chronobiology and Chronotherapeutics. ABPM: Ambulatory blood pressure monitoring. BP: Blood pressure. HBPM: Home blood pressure measurement. ISC: International Society for Chronobiology. MAP: Mean arterial pressure. OBPM: Office blood pressure measurement. PROBE: Prospective, randomized, open-label, blinded endpoint trial. Sleep-time relative SBP decline, index of BP dipping, defined as percent decrease in mean SBP during the sleep span relative to mean SBP during the activity span, calculated as: ([awake SBP mean – asleep SBP mean]/awake SBP mean) x 100; participants are designated as dipper if the sleep-time relative SBP decline is ≥ 10%, and as non-dipper otherwise (Hermida et al. 2013b). The total score, utilized either to evaluate appropriateness of any given trial or to compare quality among trials, is calculated as the addition of the rating obtained for each of the eight listed protocol items (Hermida et al. 2021f).

The purpose of this article is to discuss the deficiencies of recent works that are exemplary of the faulty methods that have resulted in unrepresentative and/or discordant findings and needless medical controversy. Our intent is not to attack or embarrass well-meaning investigators but to point out the mistakes that have been made in the design and conduct of circadian clinical trials, with the goal of improving the quality and rigor of future investigations. Herein, we accomplish these goals by elucidating the shortcomings of: (i) a recently published purported comprehensive review of past publications pertaining to treatment-time differences in the effects of hypertension therapy (Maqsood et al. 2023) and (ii) the design and conduct of both the pragmatic Treatment in Morning versus Evening (TIME) (Mackenzie et al. 2022; Rorie et al. 2016, 2017) and BedMed (Garrison et al. 2022) cardiovascular disease (CVD) outcomes trials. These works, which involved investigators with seemingly little or no formal training in the methods of circadian medicine research, are assessed, respectively, in reference to the related findings of our previously (2021) published in-depth review of the literature of this topic (Hermida et al. 2021b, 2021c, 2021d, 2021e) and the ISC/AAMCC recommended guidelines for the design and conduct of treatment-time trials that assess differences in the efficacy of BP-lowering medications (Hermida et al. 2021f).

Example I. Incomplete and erroneous information leading to invalid findings of a recently published review and meta-analysis

Background: current knowledge of ingestion-time differences in the pharmacodynamics of hypertension medication

In 2021, we conducted a systematic and comprehensive review and meta-analysis – registered with PROSPERO (International Prospective Register of Systematic Reviews; no. CRD42020201220) – of published human trials that explored individual hypertension medications and their combinations for ingestion-time differences in BP-lowering effects, markers of hypertension-induced target organ pathology of the kidney and heart, drug adverse effects, and patient adherence to therapy (Hermida et al. 2021b, 2021c, 2021d, 2021e). In keeping with the a priori established criteria (Hermida et al. 2021c), our systematic review yielded 155 valid short-term efficacy trials published between 1976 and 2020 that collectively represented 23,972 hypertensive persons treated with 37 different single and 14 fixed-dose dual-combination therapies. Some 127 of them assessed participants by around-the-clock ambulatory BP (ABP) monitoring (ABPM), among which only 62, representing 6,120 hypertensive persons, provided quantitative data on the ingestion-time-dependent effects of hypertension treatment on the awake and asleep systolic BP (SBP) means and the sleep-time relative SBP decline (dipping) – percent decrease in mean SBP during the sleep span relative to the mean SBP during the activity span and calculated as [(awake SBP mean − asleep SBP mean)/awake SBP mean] × 100 (Hermida et al. 2013b). Among the 155 identified trials, 25 were “neutral,” i.e., reported non-inferiority of the at-bedtime/evening vs. upon-waking/morning treatment schedule, while the remaining 130 (83.9%) reported superiority of the at-bedtime/evening treatment schedule for one or more of the designated outcomes of:

1. Enhanced asleep SBP reduction (without induction of sleep-time hypotension) by an average of 5.17 mmHg (95% confidence interval (CI) [4.04, 6.31], P < 0.01 between treatment-time groups), but not awake SBP mean (average treatment-time difference of only 0.71 mmHg, 95%CI [−0.04, 1.46], P = 0.06), which consequently further increased the sleep-time relative SBP decline by an average of 3.22% (95%CI [2.42, 4.02], P < 0.01) towards the normal dipper 24 h BP pattern (Hermida et al. 2021d). This is an extremely important finding given that the asleep SBP mean and asleep SBP dipping – independent of office BP measurements (OBPM) and wake-time and 24 h ABP means being low or elevated – are the strongest joint prognosticators of patient CVD risk (Ben-Dov et al. 2007; Boggia et al. 2007; Fagard et al. 2008; Hermida et al. 2011, 2018; Kario et al. 2020; Minutolo et al. 2011; Ohkubo et al. 2002; Roush et al. 2014; Salles et al. 2016; Staplin et al. 2023) and thus documented therapeutic targets for increasing patient event-free survival (Hermida et al. 2011, 2018).

2. Increased proportion of adequately controlled patients, i.e., attainment of BP values, mainly asleep SBP mean, below thresholds for the diagnosis of hypertension or study-specified therapeutic targets.

3. Diminished incidence of medication adverse events (10.9 ± 14.8% vs. 14.2 ± 14.9%, respectively, with the at-bedtime/evening vs. upon-waking/morning treatment schedule; P = 0.022), mainly when entailing the angiotensin-converting enzyme inhibitor (ACEI) and dihydropyridine calcium-channel blocker (CCB) classes of medication. Indeed, no single published trial reported superiority of the upon-waking/morning treatment regimen in terms of patient safety and tolerance to therapy (Hermida et al. 2021c, 2021e).

4. Improved markers (indicative of lessened severity) of hypertension-associated target organ pathology of the kidney – including reduced albuminuria and increased glomerular filtration rate – and heart – including decreased left ventricular posterior diameter and left ventricular mass.

The clinically and significantly enhanced reduction of the asleep SBP mean achieved with the at-bedtime/evening vs. upon-waking/morning treatment strategy was largest for trials of the: (i) dual-combination (average of 8.91 mmHg, [4.62, 13.21], P < 0.01) (Hermida-Ayala et al. 2022) and poly therapies (8.46 mmHg, [3.66, 13.25], P < 0.01) compared with the mono therapies of either angiotensin receptor blockers (ARBs) (4.10 mmHg, [2.03, 6.18], P < 0.01), ACEIs (4.58 mmHg, [2.54, 6.62], P < 0.01), or CCBs (3.11 mmHg, [1.58, 4.63], P < 0.01); and (ii) special high-CVD-risk patient populations – non-dippers, i.e., patients with sleep-time relative SBP decline < 10% (8.30 mmHg, [6.39, 10.21], P < 0.01), those diagnosed with diabetes or chronic kidney disease (CKD), or those with medical history of previous major CVD event (7.99 mmHg, [3.03, 12.95], P < 0.01) – relative to the differential at-bedtime/evening vs. upon-awakening/morning BP-lowering effect on hypertensive individuals of the general population (4.20 mmHg, [3.09, 5.31], P < 0.01) (Hermida et al. 2021d). Furthermore, there was no significant ingestion-time difference in the reported average patient adherence to therapy among the trials, i.e., 95.8 ± 5.9% vs. 94.8 ± 8.2%, in patients randomized, respectively, to the upon-waking/morning and at-bedtime/evening treatment schemes (P = 0.306). Finally, a most highly noteworthy finding of our comprehensive and systematic review is that no single trial reported significantly greater BP-lowering or other benefits of the most recommended upon-waking/morning hypertension treatment-time strategy (Hermida et al. 2021c, 2021d, 2021e).

Quality evaluation of ingestion-time efficacy trials of hypertension medications

Even though the vast (83.9%) majority of the 155 reviewed trials reported the ingestion of mono and fixed-dose dual-combination hypertension therapy of all classes at-bedtime/evening rather than upon-waking/morning resulted in clinically and statistically significant enhanced benefits, the methods of many of them were deficient to various extent, particularly those 16.1% that failed to substantiate enhanced benefit of the at-bedtime dosing strategy (Hermida et al. 2021b). Indeed, according to the eight-item quality ranging system of the ISC/AAMCC (Table 1), among the total of the 155 reviewed trials, 68 were of low quality, with 34 that scored ≤ 1 and 34 others that scored 1.5 or 2. Some 55 of the trials were of high quality, with a score ≥ 4, and all but three (5.5%) of these reported marked advantages of the at-bedtime/evening compared to the upon-waking/morning hypertension treatment schedule. In fact, the proportion of “neutral” studies markedly decreased with increasing quality score, ranging from 52.8% for trials that scored < 1 to 17.6% for those that scored 1 or 1.5, 12.5% for those that scored 2 or 2.5, and a very low 2.8% for those that scored ≥ 5 (Hermida et al. 2021b). Among the most frequent compromising deficiencies were:

1. Absence or miscalculation of the minimum required sample size (Item 7 of Table 1; 83.2%).

2. Inadequate ABPM duration, i.e., <48 h (Item 5; 78.1%).

3. Improper computation of the ABPM-derived mean values (Item 6; 77.4%).

4. Erroneous selection of the primary ABP endpoint and their expression as arbitrary “daytime” and “nighttime” means that are neither biologically relevant nor clinically meaningful (Item 4; 53.6%).

5. Inappropriate designation of the tested treatment times because of referencing them externally to clock hour rather than internally according to indicative surrogate markers of circadian stage, i.e., at-bedtime before sleep and upon-awakening after sleep (Item 2; 53.6%).

6. Recruitment of patients: (a) without consideration of the activity/sleep routine so that it would be similar among all of those comprising the study cohort; (b) without qualification of being hypertensive based on around-the-clock ABPM but by wake-time OBPM or home BP measurement (HBPM); or (c) under treatment for their hypertension and who at the commencement of the trial were already properly controlled according to diagnostic thresholds for hypertension (Item 1; 51.0%).

Nonetheless, regardless of the deficiencies as denoted by the differential quality score of the 155 trials, no single reviewed prospective randomized trial reported significantly enhanced BP-lowering (mainly during sleep), patient safety, patient adherence to therapy, or other benefits of the still most recommended upon-waking/morning hypertension treatment-time scheme (Hermida et al. 2021b, 2021c, 2021e).

Erroneous and incomplete trial selection

Maqsood et al. (2023) recently published review of treatment-time effects of hypertension therapy is neither systematic nor accurate. Our thoroughly comprehensive and systematic 2021 publication summarized above (Hermida et al. 2021c, 2021e) identified 155 short-term treatment-time efficacy trials that were reported between 1976 and 2020 (currently numbering 158), excluding the long-duration ones that assessed CVD outcomes. Some 127 of them assessed patients by ABPM to derive the outcome variables. Among the latter, 62 specifically reported, as required by our protocol for quantitative meta-analysis (Hermida et al. 2021d), the asleep SBP mean and sleep-time SBP decline, strongest predictors of CVD risk, and documented therapeutic targets for prevention (Hermida et al. 2011, 2018). Maqsood et al. (2023), despite knowledge of our previously published systematic review and meta-analysis (summarized in their reference #4), were able to identify just 56 studies, i.e., only 44.1% of the 127 ABPM-based published trials (Hermida et al. 2021e). Moreover, they improperly included 16 trials, among the total of 72 identified by their incomplete and mistake-laden review as listed in their supplement, that we consider invalid for efficacy evaluation on the following grounds:

• 3 (their #47, #50, #62) of them entailed patient BP assessment by HBPM or OBPM (without BP measurement during sleep), but not by ABPM as required by their own inclusion criteria; nevertheless, the authors provided information on ABPM device brand and duration of ABP monitoring that is inexistent in the original publications.

• 1 (#23) of them reported BP only for two short 4 h-intervals, not identifiable as awake/asleep or daytime/nighttime spans.

• 1 (#72) of them did not assess efficacy, but only CVD outcomes after long-term follow-up and with unknown BP-lowering treatment schemes.

• 3 (#19, #22, #24) of them were preliminary reports of larger trials and thus solely providing preliminary information (two of them just in abstract form and thus lacking information on most of the variables of interest for retrieval).

• 3 (#21, #70, #71) of them entailed morning and evening patient groups treated with different BP-lowering medications and, therefore, are invalid for assessment of treatment-time differences in the efficacy of a specific therapy.

• 3 (#41, #42, #48) of them comprised the CVD-outcome MAPEC (Monitorización Ambulatoria para Predicción de Eventos Cardiovasculares, i.e., Ambulatory Monitoring for Prediction of Cardiovascular Events) sub-studies of patients with diabetes, CKD, and resistant hypertension, respectively, that further biased the results of the review by including duplicated information.

• 2 (#38, #67) of them were long-term CVD-outcome trials (in addition to #41, #42, #48 sub-studies described above) that allowed by protocol treatment adjustment to achieve ABP control, versus all of the short-term ones that did not allow such because, by study design, therapy and dose were fixed.

Erroneous retrieved information on trial design/conduct and reported findings

Review of the individual trial publications discloses incorrect reporting in Figures 1–3 by Maqsood et al. (2023) of the average treatment-time differences in BP lowering per trial. For example, the authors when designating the extent of reduction of the asleep SBP mean of the reviewed trials (their Figure 2) by categorizing them according to the geographical location of their conduct, rather than the quality of their design and conduct, reported inaccurate values for 11/19 of their so-designated “Hermida-(Galicia, Spain)” trials, but only for 3/42 of their so-designated as “non-Hermida-(non-Galicia)” trials, with the extent of disparity between the actual values stated in the original publications and those listed by Maqsood et al. (2023) of their latter category of trials being lower by up to 10 mmHg. Such errors inflate the otherwise clinically non-significant and irrelevant difference in the efficacy of BP-lowering according to treatment time between the trials arbitrarily sorted by geographic location of their conduct, thereby generating biased and invalid findings. Additionally, most of the 95%CI the authors listed in Figures 1–3 are incorrect, mistakenly reducing the number of individual non-Galicia-trials that demonstrate significant treatment-time differences from the real sum of 12 to only 7. These mistakes that were made in retrieving information from the original publications severely compromise the findings and conclusions of this review. Moreover, the huge difference in the distribution of errors that arise from the arbitrary categorization of trials by location of their conduct, being eightfold larger for the Galicia than for non-Galicia trials, rather than according to their quality score, further detracts from the merit and quality of the review.

Furthermore, the publication by Maqsood et al. (2023) is filled with many other types of mistakes as well as erroneous information – therefore misleading for everyone not familiar with the original publications – concerning virtually all of the listed aspects of the design and conduct of each retrieved trial (their Supplement Tables S1, S2, and S3). Ones of particular relevance, among the many that are too numerous to fully chronicle, entail the reported:

1. Number of recruited patients and duration of follow-up per trial: For example, the number of participants in the markedly under-powered HARMONY trial (Maqsood´s reference #65 in their Supplement; Poulter et al. 2018) was 95 out of the total 103 recruited – although according to the information provided by the trial investigators, the minimum required sample size needed to be 175 (Hermida et al. 2021e) – not 192 as mistakenly reported in Maqsood´s Table S2. Furthermore, the HARMONY trial publication (one of the only three past published trials with ISC/AAMCC quality score of 0; Hermida et al. 2021b) failed to specify the class or dose of any of the prescribed BP-lowering medications; thus, Maqsood´s categorization of the unknown tested treatment as “long acting” is speculative.

2. Dose(s) of the tested BP-lowering medications: This highly relevant information was provided only for a very few trials, although it is reported in, and thus available from, the majority of the original publications. Furthermore, the names of some of the medications are not even spelled correctly, e.g., “omlesartan” (instead of olmesartan; all 7 times it is mentioned), “temlisartan” (instead of telmisartan; all 3 times it is stated), “torsedmide” (instead of torasemide or torsemide; once, for the only trial that tested this diuretic), etc., while some others are improperly identified by their commercial brand name. Collectively, such mistakes are too numerous and too repetitive to be considered simply the consequence of typographical errors, causing uncertainty and even distrust of the due diligence of those charged with the retrieval of data from the original publications.

3. Tested treatment times: They are mistakenly designed as “AM/PM” across all of the trials, regardless of the real ones stated in the original publications, whether arbitrary clock times (e.g., 08:00 h vs. 20:00 h), specific spans of the 24 h (e.g., 06:00 h-10:00 vs. 20:00 h-00:00 h), non-specific “morning” vs. “evening,” or recommended representative surrogate markers of endogenous biological time (e.g., upon-awakening vs. at-bedtime, as properly done in 44.6% of the trials).

4. Evaluated ABPM-derived outcome variables: The ABP variables are incorrectly reported as “daytime/nighttime BP” across all of the trials, omitting and thus disregarding valuable information available from simultaneously conducted wrist actigraphy assessments (25 trials) and patient dairies (15 trials) that allowed accurate determination per patient of the wake-up and bed times and thus derivation of the biologically meaningful variables of the awake and asleep BP means and actual amount of the sleep-time BP dipping, a highly significant prognostic marker of CVD risk that was not evaluated or considered in Maqsood´s review.

5. Patient inclusion criteria: They are incomplete and not indicative of the specifically targeted patient cohort, thus omitting relevant information that again significantly impacts the results of their meta-analysis (e.g., their supplement reference #49: all participants were non-dippers; reference #52: all patients were African-American; reference #53: all patients had CKD and a non-dipper 24 h ABP pattern; reference #54: all patients had type 2 diabetes; and so on). Indeed, among the 62 truly valid ABPM-based trials (Hermida et al. 2021d), 17 entailed special high-CVD-risk populations, i.e., non-dippers and patients with diabetes, CKD, or previous CVD event, and 16 additional ones trialed dual combination and polytherapies. The trials entailing these cohorts and combination therapies are characterized by statistically significantly larger average asleep BP reduction with the at-bedtime/evening than the upon-awakening/morning treatment schedule, as previously reported (Hermida et al. 2021d, 2021e) and earlier summarized herein.

6. Time intervals for the calculation of the awake/asleep BP means: They are erroneously listed as fixed clock-time spans across all of the trials, unrealistically assuming a common wake/sleep routine among all of the participants of all of the trials, once more disregarding the data of wrist actigraphy and patient diaries that are available from 64.5% of the ABPM-based trials to enable accurate determination of the bed and rise times of each participant. Furthermore, the listed fixed clock-time spans for all these actigraphy/diary-based trials are purely speculative inasmuch as, obviously, none of such fictitious spans are reported in the original publications.

7. Risk of bias assessment: The information reported in Maqsood´s et al. (2023) Table S1 is filled with many implausible entries and incongruences. For example, the authors mistakenly attributed high risk of attrition bias to short-term trials that did not evidence losses during follow-up. Additionally, they attributed differential risk of bias for their assessed variables of the MAPEC study (Hermida et al. 2010) and the three published MAPEC sub-studies involving, respectively, patients with diabetes, CKD, and resistant hypertension, respectively, all of them obviously conducted following the very same investigative protocol.

8. Publication date: It is wrongly listed for several trials by up to 7 years, a minor issue compared to all of the others, but nonetheless another deficiency that further detracts from the quality and credibility of the review and reported findings of the meta-analysis.

The enormity of the stated erroneous information arising from the incorrect and incomplete retrieval of data from the original publications substantiates the poor quality of Maqsood´s review. Most importantly, those mistakes concerning the information of the reported tested treatment times, ABPM-derived outcome variables, patient inclusion criteria, and time intervals utilized to calculate the ABP means resulted in invalid and faulty findings and conclusions. Additionally, the scale of the mistakes made by the authors, including those evident in their Figure 2 regarding the duplication of the SBP findings in both panels and absence of findings on diastolic BP (DBP) – a seemingly minor oversight in comparison to all of the other discussed major errors committed by the authors – went undetected by the peer reviewers, as were the inadequacies of the review according to criteria established for sound circadian rhythm research, thereby suggesting the limited competency in the unique methods of circadian medicine research both by the authors and reviewers.

Disregard of trial design/conduct characteristics that impact results

The limitations of Maqsood et al. (2023) review paper are further evidenced by disregarding the quality of the design and conduct of the retrieved trials according to the ISC/AAMCC guidelines and recommendations (Hermida et al. 2021f) that significantly impact their findings and conclusions. Table 2 documents the proportion of the 62 valid ABPM-based trials identified by our earlier systematic review and meta-analysis (Hermida et al. 2021d) that reported significantly greater asleep SBP reduction and/or increased dipping achieved by at-bedtime/evening therapy is, among other factors, explainable in large part by being of proper sample size and high ISC/AAMCC quality score, rendering Maqsood et al. (2023) classification of trials simply by geographic location of their conduct irrelevant and improper. Table 2 further documents the huge differences between the Galicia and non-Galicia trials in the average number of participants per trial and the proportion of trials fulfilling each of the ISC/AAMCC recommended items for proper design and conduct of such trials that significantly impact the validity of findings.

Table 2. Comparison of patient sample size, quality of trial design and conduct, and rate of significant treatment-time dependent reduction of asleep SBP mean reported in ABPM-based efficacy studies categorized by location of coordinating institution.

Item	Galicia, Spain	Other locations	P between groups
Short-term ABPM-based efficacy trials	19	43
No. total patients, No.	3067	3008
No. patients per trial, mean±SD	161.4 ± 53.7	69.9 ± 125.7	0.003
Quality evaluation: Trials designed and conducted in keeping with ISC/AAMCC recommendations,* No. (%)
Patient inclusion criteria	19 (100)	20 (46.5)	<0.001
Tested treatment times	19 (100)	10 (23.3)	<0.001
Primary endpoints	19 (100)	22 (51.2)	<0.001
ABPM duration	19 (100)	3 (7.0)	<0.001
Calculation of ABPM parameters	19 (100)	4 (9.3)	<0.001
Sample size	15 (78.9)	4 (9.3)	<0.001
Study design	19 (100)	24 (55.8)	<0.001
Quality score per trial, mean±SD (−1=lowest; 7=highest)	6.79 ± 0.42	2.89 ± 1.25	<0.001
Trial categorization by sample size and/or quality score in keeping with ISC/AAMCC recommendations,* No. (%)
Higher quality trials score ≥ 4	19 (100)	14 (32.6)	<0.001
Trials with > 50 patients	19 (100)	15 (34.9)	<0.001
Trials with > 50 patients and score ≥ 4	19 (100)	7 (16.3)	<0.001
Trials reporting significantly greater asleep SBP reduction and/or increased dipping with bedtime dosing, No. (%)
All trials	17 (89.5)	24 (57.2)	0.006
Trials with quality score ≥ 4	17 (89.5)	12 (85.1)	0.832
Trials with > 50 patients	17 (89.5)	13 (86.7)	0.801
Trials with > 50 patients and score ≥ 4	17 (89.5)	7 (100)	0.949

*ISC (International Society for Chronobiology) and AAMCC (American Association for Medical Chronobiology and Chronotherapeutics) recommendations for the design and conduct of ingestion-time trials of hypertension medication (Hermida et al. 2021f) are listed in Table 1 of this article.

Additionally, only 6 out of the 62 valid ABPM-based trials reported a barely significant very small treatment-time difference in the reduction of the awake SBP, amounting to only a clinically unimportant average difference of 0.71 mmHg (P = 0.06) between the at-bedtime/evening and upon-awakening/morning treatment schedule (Hermida et al. 2021d). Indeed, the average treatment-time-dependent disparity in the awake BP reduction was not statistically significant in any of the subgroups of trials that tested monotherapies, including those conducted in Galicia, regardless of the class of the tested hypertension medication (ARB, ACEI, CCB, or others) (P always > 0.27), and it was only of borderline significance in those that entailed dual or poly therapies (P = 0.04). Findings and conclusions regarding the 24 h SBP mean are almost identical (Hermida et al. 2021d). Furthermore, as already reported above, trial categorization according to tested medication class, medication number (mono, dual, poly therapy), and patient co-morbidity reveals the average reduction of the asleep SBP mean and increase of the sleep-time relative SBP decline are most significantly enhanced with the at-bedtime/evening treatment strategy when it involves fixed-dose dual-combinations, poly therapies, and high-CVD-risk patient cohorts having a non-dipper 24 h BP pattern, those diagnosed with diabetes or CKD, and those with medical history of a previous CVD event, in comparison with lower-CVD-risk ones of the general population treated with monotherapies (Hermida et al. 2021d). In all of these cohorts, the clinically meaningful numerical treatment-time differences in the effect on these two ABP variables were > 4-fold larger than the inconsequential non-significant ones derived from the markedly erroneous retrieved data of trials arbitrarily sorted by the Galicia vs. other-than-Galicia location of study conduct.

Erroneous trial selection for evaluating treatment-time effects on CVD risk

Maqsood et al. (2023) additionally provide an incorrect comparison of the long-term CVD-outcomes trials (their Figure 4) by improperly including, apart from three MAPEC sub-studies thus to some extent duplicating information, the invalid (i) underpowered Fosinopril versus Amlodipine Cardiovascular Events Trial (FACET) (Tatti et al. 1998); (ii) the prematurely sponsor-terminated Controlled Onset Extended-Release (COER) Verapamil Investigation of Cardiovascular Endpoints (CONVINCE) trials; most important, the design of both these investigations compared patients of morning and evening groups randomized to ingest different medications; and (iii) the “pragmatic” Treatment in Morning versus Evening (TIME) trial, whose design, conduct, and reported results deserve detailed criticism as done in the subsequent section.

The FACET trial randomized a rather small cohort of only 380 patients with diabetes, who were diagnosed as being hypertensive on the sole basis of wake-time OBPM, to an open-label morning 20 mg/day fosinopril or evening 10 mg/day amlodipine treatment regimen and then followed for up to 3.5 years (Tatti et al. 1998). If the wake-time OBPM was not reduced to the target goal – defined as wake-time office SBP ≤140 mmHg and DBP ≤90 mmHg or a decrease ≥20 mmHg in SBP or DBP if the baseline SBP was >160 mmHg or DBP >100 mmHg – by the initial study medication, the second one was added for ingestion at an unspecified time of the day. Amlodipine was the added therapy for 30.7% of the uncontrolled patients whose initial treatment was fosinopril, and fosinopril was the added therapy for 26.2% of patients whose initial treatment was amlodipine. This means that more than half, 56.9%, of the participants were simultaneously treated with both of the investigated medications that were ingested at an uncertain time of day, while the remaining ones were treated either in the morning or evening with only one of the two medications that were evidently different for each treatment-time-group, making any comparison between the two treatment-arms markedly confusing and invalid. Additional limitations of the FACET trial design are: (i) The required sample size was established on the basis of the potential differing effects of fosinopril vs. amlodipine in reducing total cholesterol, not CVD events, as misleadingly implied by the trial name. (ii) Reported CVD events, assessed only as secondary outcomes, were incomplete, e.g., no information was provided on heart failure and coronary revascularization. (iii) The rather small sample size of the trial further diminishes the impact, if any, of the findings on the management of patient CVD risk (Hermida et al. 2023). (iv) At trial closure, only 41 CVD events were documented (Tatti et al. 1998), too few to draw conclusions of clinical relevance.

The CONVINCE trial was designed to explore whether the initial bedtime treatment with 180 mg of the unique COER-verapamil medication – a specially designed delayed-onset, extended-release CCB formulation intended for bedtime ingestion to achieve highest serum drug concentration upon awakening and the initial hours of the daytime activity span to effect attenuation of the rate of rise and level of BP during this span – was equivalent to morning-treatment with either 50 mg of the ß-agonist atenolol or 12.5 mg of the diuretic hydrochlorothiazide – therapies that produce a more constant serum concentration of medication throughout the 24 h than the COER-verapamil therapy – in preventing, as the primary outcome variables, myocardial infarction, stroke, and CVD death (Black et al. 2003). The CONVINCE trial was prematurely terminated 2 years earlier than planned (and thus was underpowered) by the sponsoring pharmaceutical company purely for commercial reasons. At the conclusion of the abbreviated 3-year follow-up period, no differences were detected in the primary outcome variables between the two tested treatment-time strategies. Although bedtime ingestion of COER-verapamil significantly reduces morning BP, it exerts only limited effect on asleep BP, as documented by a randomized trial of the medication that showed a two-fold greater reduction of the awake than asleep SBP/DBP means (White et al. 1998). This results in significant reduction of the sleep-time relative BP decline, thereby diminishing BP dipping, and thus inducing a non-dipper or even riser 24 h BP pattern, each of which is associated with elevated CVD risk as consistently substantiated by numerous investigations (Ben-Dov et al. 2007; Boggia et al. 2007; Dolan et al. 2005; Fagard et al. 2008; Hermida et al. 2011, 2018, 2020c, 2020d; Kario et al. 2020; Minutolo et al. 2011; Ohkubo et al. 2002; Roush et al. 2014; Salles et al. 2016). Accordingly, the results of the CONVINCE trial cannot be considered evidence against the merit of at-bedtime hypertension chronotherapy consisting of conventionally formulated once-a-day medications that targets the elevated asleep SBP level, the BP parameter most strongly linked with target-organ pathology and heightened CVD risk. In actuality, the findings of the CONVINCE trial refute the unproven theory of the 1990s that hypertension therapy should target as its major goal control of the rate of rise and level of BP upon-waking and initial hours of daily activity to diminish the risk of CVD events (Hermida et al. 2021e, 2023).

In conclusion, FACET and CONVINCE do not qualify as valid hypertension ingestion-time or chronotherapy trials because, apart from their many limitations summarized above, neither one of them included an upon-awakening-time treatment arm with the same tested medication as reference to compare the effects both upon ABP control and CVD risk reduction. Interestingly, the authors of another recent review and meta-analysis (Abuelazm et al. 2023) on hypertension treatment-time effects on CVD outcomes made the exact same mistake of including FACET and CONVINCE among their listed five identified “valid” trials, jointly with MAPEC, the Hygia Chronotherapy Trial (HCT), and the pragmatic TIME trial, the latter two further described and discussed in detail in the following sections. This further emphasizes the insufficient competency of too many authors and reviewers of the methods of circadian medicine research.

Example II. Comparison of the design and conduct of the ABPM-based Hygia Chronotherapy Trial and pragmatic TIME and BedMed studies

Background: Hygia Chronotherapy Trial (HCT)

The multicenter, prospective, randomized, open-label, blinded endpoint (PROBE) HCT was one of the several ABPM-based studies nested within the so-called Hygia Project, a large research network of 292 physician-investigators, all of whom were trained in the methods of medical chronobiology and chronotherapeutics, including certification on the use of ABPM, that was established to incorporate ABPM as routine procedure in the primary care setting to diagnose and manage hypertension, assess response to treatment, and evaluate patient CVD and other risks associated with elevated BP (Hermida 2016). The HCT entailed 19,084 protocol-correct ABPM-diagnosed hypertensive patients randomized either to ingest the entire daily dose of ≥ 1 of the major classes of prescribed hypertension medications – ARBs, ACEIs, CCBs, ß-blockers, and/or diuretics – at bedtime (one half of the patients in this group ingested all of their medications at bedtime) or all of their prescribed medications upon awakening (Hermida et al. 2020a). The design of the long-term 6.3-year follow-up CVD-outcomes HCT incorporated most of the relevant and novel features that were integrated into the earlier concluded MAPEC Study (Hermida et al. 2010), including:

1. Baseline ABPM patient assessment to ensure inclusion of only true hypertensives – patients with elevated awake or asleep ABP, regardless of daytime OBPM (Hermida et al. 2013b) – and exclusion of low-CVD-risk individuals who displayed elevated OBPM yet proven normotensive by ABPM, commonly termed isolated-office hypertension, although the more accurate term should be “masked normotension” (Hermida et al. 2013b).

2. Periodic, at least annual, ABPM patient evaluation throughout the 6.3-year median follow-up span to achieve proper titration and safety (avoidance of sleep-time hypotension) of physician prescribed first-time or add-on BP-lowering medications.

3. 48 h, instead of the most common low-reproducible and poorly representative ≤24 h, ABPM assessment to optimize accurate determination of the ABP characteristics – including mean BP values and dipping classification – and ABPM-based CVD-risk appraisal (Hermida et al. 2021a), which depend more markedly on ABPM duration than measurement frequency, as previously documented (Hermida et al. 2013a).

4. Concomitant with the conduct of each ABPM assessment done during follow-up, patient self-recording in the provided specifically designed personal diary of nocturnal bed and morning awakening times to ascertain the beginning and end of the activity and sleep spans and to enable accurate calculation on an individual basis of the awake and asleep BP means plus the sleep-time relative BP decline, rather than assuming arbitrarily a common set of fixed clock times across all of the patients for the commencement and termination of the activity and sleep spans to generate “daytime” and “nighttime” values that are neither biologically representative nor clinically meaningful.

5. Randomization of participants to designated treatment-times that are markers of internal circadian stage, i.e., upon-awakening from sleep or at-bedtime before sleep, instead of external arbitrary chosen clock hours or non-specific “morning” vs. “evening” spans of the day.

6. As currently recommended (Hermida et al. 2021f), instruction and reminders of patients upon recruitment and at every clinical visit throughout follow-up to place their prescribed medication(s) on the bedside table and to ingest it (them), depending on their randomly assigned treatment-time schedule, either immediately upon awakening from nighttime sleep or before turning the lights off to retire to sleep at night as a means to increase their adherence to the allocated schedule of therapy (Hermida et al. 2010, 2020a).

Although the estimated completion date was December 31, 2020, the trial was closed by its Scientific Committee 3 years earlier once the required sample size (including that for the protocol-specified groups of participants complicated with either diabetes or CKD) was attained and the findings of the interim analyses at that time already showed significant CVD risk reduction with the at-bedtime therapy strategy. At the final evaluation, ABPM-based hypertensives randomized to the at-bedtime, in comparison to the upon-awakening, treatment regimen evidenced significantly lower adjusted hazard ratio (HR; 0.55, 95%CI [0.50–0.61], P < 0.001) of the primary CVD outcome variable – composite of CVD death, myocardial infarction, coronary revascularization, heart failure, and stroke – as well as each of its individual components (P < 0.001 in all cases), i.e., CVD death (0.44 [0.34–0.56]), myocardial infarction (0.66 [0.52–0.84]), coronary revascularization (0.60 [0.47–0.75]), heart failure (0.58 [0.49–0.70]), and stroke (0.51 [0.41–0.63]) (Hermida et al. 2020a). Analysis of the impact of hypertension treatment-time on the primary composite CVD-outcome variable for participants further categorized by known influential markers of CVD risk indicated significantly lower HR of CVD events when BP-lowering medications were ingested at-bedtime rather than upon-waking, regardless of participant´s sex, age, smoking habits, hypertension treatment at baseline, either normal or elevated baseline awake or asleep SBP mean, dipper or non-dipper 48-h SBP pattern, and absence or presence of either diabetes, CKD, or positive or negative history of previous CVD event (Hermida et al. 2020a). Finally, throughout the follow-up there were no treatment-time differences in the proportion of patients who reported adverse effects of BP-lowering medications (6.7% of the upon-awakening and 6.0% of the at-bedtime treatment participants; P = 0.061), or in the manifestation of sleep-time hypotension, defined by current ABPM criteria (Hermida et al. 2013a) − 39 and 26 persons of the upon-awakening and at-bedtime groups, respectively (0.3% of all of the trial participants; P = 0.114 between groups) – although the findings of such low incidence may in part reflect the adopted clinical protocol that mandated the conduct of 48 h ABPM several weeks after initiating or changing one´s hypertension therapy in order to assess attainment of the clinical goal for ABP control plus avoidance of sleep-time hypotension.

Treatment in morning versus evening (TIME) study

The TIME study was a PROBE, pragmatic, and decentralized trial conducted entirely online, without personal contact between the investigators and participants and without involvement of any kind by treating physicians. Participants were self-recruited by internet and, in accord with the TIME protocol, “treatment allocation was not made available to treating physicians unless there was a specific individual clinical need” (Rorie et al. 2016). Recruitment initially commenced in 2011 as a pilot study, with these participants and their data included in the main trial, and it continued until 2018. The required sample size was adjusted upward from 10,269 (Rorie et al. 2016) to > 20,000 participants (Rorie et al. 2017) to achieve “the necessary number of events needed,” although neither the expected event-rate nor the one actually documented in the pilot study was reported (Ho and Reid 2022). The investigators also extended, without providing justification, the end-date of the study first from 07/04/2016 to 05/31/2019, then to 04/03/2021, and finally to 06/30/2022 (https://www.isrctn.com/ISRCTN18157641). Participants – required by protocol to be already treated with at least one BP-lowering medication – were informed by e-mail according to a randomization procedure to ingest all of their BP-lowering medications either in the morning, defined by the 4-h span extending from 06:00 h to 10:00 h, or all of them in the evening, defined by the 4-h span extending from 20:00 h to 00:00 h. After a median 5.2-year follow-up, there was no significant difference in the primary endpoint (composite of CVD death, myocardial infarction, or stroke) between the two treatment-time groups. Regarding safety, fewer patients in the evening than the morning dosing group self-reported adverse events (P = 0.041), although the reported difference in the adverse event-rate between the two treatment-time groups was small, 69.2% in the evening-treatment group vs. 70.5% in morning-treatment group (Mackenzie et al. 2022). However, such an extremely large proportion (~70%) of patients claiming an adverse event is not only unusually high but implausible, although being self-reported by an online questionnaire, it is likely some relevant proportion of the claimed adverse events were not the consequence of the BP-lowering medications and its administration time. Additionally, the published safety information is based on limited reporting, i.e., only from participants “who submitted at least one complete study follow-up” report to the trial website or by e-mail (Mackenzie et al. 2022), even though by protocol all participants were expected to have submitted periodic reports commencing one month after recruitment and thereafter at quarterly intervals for the duration of the study (i.e., an average of at least 21 reports per participant during the median 5.2 years of follow-up).

Comparison of trial design and conduct between HCT and TIME

Table 3 summarizes and compares the main features of the trial design, study methods, follow-up, and data analysis of the TIME study and HCT. The TIME study, apart from the randomization of participants to treatment-time groups, did not comply with any of the other eight quality requirements and recommendations (Table 1) of the ISC/AAMCC guidelines for the design and conduct of human clinical trials on ingestion-time differences of hypertension medications (Hermida et al. 2021f). As reported in Table 3, contrary to the HCT (ISC/AAMCC quality score = 7), the TIME study, designated by a score = 1, evidences multiple shortcomings and concerns, including:

1. Treatment-times being defined as broad clock-time spans of 06:00 h–10:00 h and 20:00 h–00:00 h. They are unrepresentative of circadian time and subject to disparity in adherence according to the 4 h spans of the day to which participants were randomized to ingest their treatment (Hermida et al. 2021f). While everyone is at home upon awakening (unless when traveling), a potentially relevant proportion of persons may be frequently away from home in the evening (working late, socializing, enjoying entertainment, etc.), thus compromising consistent adherence to the evening medication ingestion-time schedule. This discrepancy is solved by defining the tested treatment-times as upon-awakening and at-bedtime, as recommended by the ISC/AAMCC guidelines (Hermida et al. 2021f).

2. Participants being self-recruited by internet. Their managing and prescribing physicians were not study investigators, and by protocol they were unaware of the treatment-time allocation of their patients. One could thus wonder whether the participants were more likely to follow, in cases of discrepancy, treatment-time instruction stipulated by e-mail or that provided through direct contact with prescribing physicians. This and many other concerns listed in Table 3 indicate the TIME study does not “reflect usual care” at all, as claimed by its investigators (Mackenzie et al. 2022), inasmuch as usual care for most essential hypertensives is in fact provided by their primary care physician, not by a decentralized web page or by e-mail communication from an unfamiliar person.

3. The participants upon recruitment self-reporting medical conditions of diabetes and CKD plus history of previous CVD event that indicates their individual prevalence being markedly lower in the TIME study, not only in comparison with the higher and more representative prevalence among participants of the HCT (Table 3) but what is currently reported in the general population of the UK where the TIME study was conducted. Although the TIME investigators claimed the prevalence of these comorbidities is “similar to patients with hypertension in a previous UK population-based study” (Mackenzie et al. 2022), in reality its reported baseline prevalence of diabetes (13.1%), CKD (3.2%), and CVD (7.0%) are all substantially lower than recently reported in the UK population (diabetes: 15.5%; CKD: 13.9%; CVD: 19.4%) (Hounkpatin et al. 2020; Tapela et al. 2021). This disparity indicates a marked and unjustified bias in the recruitment of participants with significantly lower CVD risk than would be expected from a truly representative randomly selected cohort of hypertensive persons. This bias was somehow corroborated by the TIME investigators, themselves, by being forced to double the originally calculated required sample size from 10 200 to > 20,000 participants “because of lower-than-expected CVD event rates” than other outcome trials involving UK hypertensive participants (Mackenzie et al. 2022).

4. All participants being assumingly ingesting BP-lowering medications upon recruitment, although only 37.2% of them reported information about their baseline therapy and regimen. Therefore, there was no possibility of corroborating all of the required inclusion criteria for the remaining 62.8% of the TIME trial participants.

5. All participants most likely being diagnosed as hypertensive based solely on wake-time OBPM resulting in the inclusion of those with low-risk masked normotension (elevated wake-time OBPM but normal out-of-office BP according to ABPM diagnostic criteria) and the exclusion of those with high-risk masked hypertension (normal wake-time OBPM but elevated out-of-office BP according to ABPM diagnostic criteria), both conditions collectively representing a misclassification of ~ 45% of those diagnosed by OBPM as either “hypertensives” or “normotensives” when the correct diagnosis by ABPM is just the opposite (Hermida et al. 2018).

6. Most participants already evidencing controlled wake-time OBPM (<140/90 mmHg) upon recruitment, although BP information at baseline was self-reported by only 46.7% of them. As stated in the ISC/AAMCC guidelines, beyond BP-lowering efficacy being markedly associated with pre-treatment BP level, diminishing with lower baseline BP, it is judged unethical to change the treatment regimen of any patient whose BP is already properly controlled to the clinical guidelines-recommended threshold values (Hermida et al. 2021f).

7. No follow-up BP measurements of any kind being obtained from participants. ABPM evaluation of participants was never conducted in the TIME study, and only a small proportion (36%) of them provided at least one HBPM assessment at any time during the entire study (Vickneson et al. 2022). Therefore, the “BP changes” during follow-up reported by the investigators (their Supplement Figures S8-S11) are potentially wrong and misleading, first, by assumingly presenting BP values of different cohorts of participants who provided HBPM at differing times during follow-up, as opposed to proper report of the longitudinal BP changes in the same exact cohort of participants and, second, because there is neither information on the baseline BP as reference nor repetitive BP measurements of all, or at least the same subgroup, of the participants throughout follow-up to enable a valid assessment of “BP changes.” Moreover, no other clinically relevant variables – including laboratory testing of blood and urine, adjustment of therapeutic scheme, and evaluation of target organ damage – that impact CVD risk were assessed at any time during the study.

8. The treatment scheme, as stated above, being self-reported only at baseline and by just 37.2% of the trial participants. As a consequence, there is no information available to enable the determination of whether the therapeutic scheme during follow-up was or was not balanced between the two treatment-time groups.

9. The protocol not providing recommendations on how to handle adjustments of the participants´ BP-lowering therapy, i.e., either increase or decrease of doses, number, and/or class of medications during follow-up, except in the case of nocturia in those of the evening-treatment group (Table 3). Such adjustments, if they were warranted, are unknown, or at least unreported, by the study team. Assumingly, alteration of therapy was based exclusively on wake-time OBPM and instituted by the treating physician who, unless informed by the study participant, was unaware of the allocated medication treatment-time. The only treatment recommendation and allowed adjustment of treatment expressed in the TIME study protocol is participants “in the evening-treatment group ingesting a diuretic should take it at 18:00 h [i.e., 2 h earlier than the protocol-specified commencement of the arbitrarily designated ‘evening treatment’ 4 h interval of 20:00 h-00:00 h]; if nocturia, ingest it in the morning” (Table 3). This recommendation seemingly forces the ingestion of dual or triple fixed-dose single-pill combination therapies that include a diuretic, frequently prescribed in the treatment of hypertension, to be ingested in the early evening or even morning despite the original randomization, therefore, further compromising patient adherence and assessment of findings when relying on the intention-to-treat analyses of collected data.

10. Patient adherence to treatment-time (not adherence to the prescribed medications and doses, themselves, as it should have been also required to be reported), adverse events and symptoms (regardless of being suspectedly associated or not with BP-lowering treatment or the allocated ingestion-time), and occurrence of CVD events all being self-reported by the participants and being subject to bias, as recognized by the TIME investigators, themselves (Mackenzie et al. 2022). In fact, compliance to the scheduled periodic online/e-mail reporting was extremely poor and, when reporting was done, the provided information was markedly incomplete (Table 3). Therefore, the availability of the primary and secondary outcomes for all participants might be questionable. In contrast, in the HCT all of the information reported at baseline and during each of the scheduled clinic visits during follow-up was registered, utilizing a specifically designed electronic data-entry-booklet programmed so that it “does not allow incomplete data forms (i.e., with missing information) to be submitted” (Hermida et al. 2020a) by the managing and treating physician-investigators, who periodically reviewed the complete electronic medical history of their patients.

11. Drop-outs and protocol violations being documented in a very high proportion (46.3%) of the recruited TIME participants (Table 3). Nonetheless, the data of all these participants were included in the final analyses, markedly biasing the reported findings. In fact, all of the analyses were inappropriately conducted by relying on the intention-to-treat basis, despite it being known that up to 40% of the participants were non-adherent to the allocated (morning vs. evening) treatment time.

12. Event-rates in the TIME study for the composite outcome of CVD death, myocardial infarction, and stroke being reported as 3.4% for evening and 3.7% for morning treatment, but heart failure being reported separately and not included in the primary outcome for unknown reason. Compared with the TIME study, the event-rate of the HCT for the same composite outcome was a very similar 3.3% for the at-bedtime treatment scheme and a higher 6.5% for the upon-awakening treatment-scheme (Hermida et al. 2020a).

13. In TIME, an unreported number of “participants randomly assigned to morning dosing changed to evening dosing in response to high-profile media coverage of the results of the HCT in October, 2019” (Mackenzie et al. 2022), i.e., more than 2 years before the completion date of the TIME study. The authors stated “the study team was aware that many of these participants reverted to morning dosing shortly afterwards (data not shown).” Considering the very low rate of adherence by the participants to the scheduled periodic online/e-mail reporting, vague statements like this by the main investigators of the TIME trial seem unsubstantiated.

Table 3. Hygia Chronotherapy Trial vs. TIME study: Comparison of trial design, study methods, follow-up, and evaluation of results.

Methodology item	Hygia Chronotherapy Trial		TIME study
Design and timeline*	Trial design: PROBE Median follow-up: 6.3 years Recruitment: 09/01/2008–06/14/2017 Estimated completion date: 12/31/2020 Actual completion date: 06/30/2018		Trial design: PROBE Median follow-up: 5.2 years Recruitment: 12/17/2011–06/05/2018 Estimated completion date: 2021 Actual completion date: 03/31/2021
Sample size*	Calculated based on reported incidence of CVD events in the general hypertensive population (n = 10,700) and in participants complicated with either diabetes (n = 3,800) or CKD (n = 3,800) Participants:  Total required: >18,300  Referred by participating primary-care investigators: 22,654  Protocol-correct, with required > 1-year follow-up: 19,084  Used for final analyses: 19,084		Adjusted upward from 10 269 to > 20,000 participants to “achieve the necessary number of events needed.” Expected event-rate unreported Participants:  Total required: >20,000  Self-registered by internet: 24,610  Randomized to treatment-time: 21,104  Protocol-correct Participants: 11329 (see “drop-outs” below)  Used for final analyses: 21,104
Main patient characteristics and CVD risk factors upon recruitment†	Age, years:	60.3	Age, years:	65.1
	Men, %:	55.6	Men, %:	57.5
	Diabetes, %:	23.9	Diabetes, %:	13.1
	CKD, %:	29.4	CKD, %:	3.2
	Previous major CVD event, %:	10.4	Previous major CVD event, %:	7.0
	BP non-dipper/riser pattern, %:	49.3	BP non-dipper/riser pattern, %:	Unknown
	Current smoker, %	15.2	Current smoker, %	4.2
	Obesity, %	43.0	Obesity, %	Unknown
	Source of the information: Electronic medical history.		Source of the information: Self-reported by participants.
	Participants providing all information, including ABPM and therapeutic scheme, %	100	Participants providing all information, including office BP and therapeutic scheme, %	<37
Investigated treatment times	Individualized awakening vs. bedtime as surrogate markers of circadian time		Broad clock-time spans (06:00–10:00 vs. 20:00–00:00), unrepresentative of circadian time
Patient recruitment	By managing and prescribing primary care physicians, all comprehensively trained in ABPM procedures and registered investigators of the trial		Patient self-recruitment by internet. Prescribing physicians were not investigators and, by protocol, unaware of treatment-time allocation
Method of diagnosing hypertension for recruitment	48 h ABPM		Current antihypertensive treatment, based on daytime office BP, thus leading to inclusion of those with low-risk white-coat hypertension (masked normotension) and exclusion of those with high-risk masked hypertension mainly associated with sleep-time hypertension and non-dipping 24 h pattern
Hypertension treatment status upon recruitment	42.6% untreated, 57.4% treated Treatment scheme registered and thus known by the investigators for all recruited participants at baseline and during follow-up		100% treated Treatment scheme self-reported, only at baseline, by 7,856 (37.2%) of the randomized participants
BP status upon recruitment	Inclusion criteria required elevated ambulatory BP (49.3% of the participants had non-dipper/riser 48 h BP pattern); average daytime office SBP/DBP 149/86 mmHg, much higher than that of the TIME study		No protocol-specified BP measurement, ABPM never conducted. Most participants had already controlled daytime office BP (self-reported by only 46.7% of the participants; average SBP/DBP 135/79 mmHg) giving rise to high risk of misleading findings
Follow-up design	Centralized, with periodic (at least annual; more frequent in ABPM-based lack of BP control and high-risk participants) clinical visits with the patient´s primary care physician/investigator		Decentralized, with participants contacted only by email and study personnel/investigators unknown to them
Follow-up BP measurement	Periodic (at least annual) office BP plus 48 h ABPM assessment		None. Only a small proportion (36.3%) of the participants provided at least one home BP assessment
Other clinically relevant variables assessed during follow-up	Clinical laboratory testing of blood and urine, plus review of therapeutic scheme (medications, doses, and ingestion-time), clinical characteristics, and CVD risk factors and events at each clinical visit for 48 h ABPM		None
Antihypertensive therapeutic recommendations	No division of medications for ingestion as split doses. Multiple therapies prescribed, when available, as single-tablet fixed medication combinations to achieve higher adherence		Participants in the evening-treatment group ingesting a diuretic should take it at 18:00 h; if nocturia, ingest it in the morning. This compromises compliance and assessment of findings, as well as preventing use of single-tablet fixed medication combinations in the evening treatment group
Antihypertensive therapeutic adjustments during follow-up	Prescribed, if needed, by the participating primary care investigators, based entirely on findings of each of the 48 h ABPM assessments during follow-up		No described in the protocol. Assumingly, prescribed, if needed, based on daytime office BP, by the participants´ treating physicians, who were not investigators and, by protocol design, unaware of their treatment-time allocation
Antihypertensive medications and dosages during follow-up	Reported and registered at every clinic visit		Unknown, unreported
Adherence assessment	Morisky-Green test at each clinic visit for 48 h ABPM assessment to assess adherence to all prescribed medications and dosages according to the allocated treatment-time		Online questionnaire only for assigned treatment-time, not for actual intake of prescribed medications and dosages. Compliance with reporting and recall unknown. Reporting subject to bias
Safety assessment	Adverse events registered when reported by patient and/or revealed through non-direct questioning and physical examination by primary care physician		Patient reporting (compliance during follow-up unknown) by email. Extremely high adverse event rate (~70%); possible relation specifically to hypertension therapy and ingestion-time, unknown. Reporting subject to bias
CVD outcome reporting	By the primary care investigators, by scheduled periodic (at least annual) review of the complete electronic medical record of each patient		Mostly self-reported by the participants; linked to anonymized National Health Service records. Reporting subject to bias
CVD outcome adjudication	Independent events committee, masked to medical records, ABPM findings, therapeutic scheme, and treatment-time allocation		Independent endpoint committee, masked to dosing time allocation
Drop-outs and protocol violation	607 participants with invalid baseline 48 h ABPM and 84 participants without required >1 year follow-up at study closing; all excluded from analysis		A total of 46.3% of the recruited participants: 847 participants withdrew full consent, 2453 withdrew from follow-up, and 6475 reported non-adherence to dosing-time; nonetheless, the data of all these participants were included in the final analyses
Primary outcome comparison	By intention-to-treat for those with required > 1-year follow-up		By intention-to-treat, despite known up to 40% non-adherence to allocated (morning vs. evening) dosing time
Statistical analyses of primary and secondary outcome variables	Cox proportional-hazards model adjusted by significant confounders		Unadjusted Cox proportional-hazards model
Reported event-rates for composite outcome of CVD death, myocardial infarction, and stroke	3.3% for bedtime treatment 6.5% for upon-awakening treatment		3.4% for evening (20:00 h–00:00 h) treatment 3.7% for morning (06:00 h–10:00 h) treatment
Main trial conclusions	“Hypertension” (preferably, altered ambulatory BP, i.e., elevated asleep SBP and/or non-dipping) must be diagnosed only by 48 h ABPM. In patients with so-defined altered ambulatory BP, bedtime treatment improves BP control, particularly of asleep BP (strongest predictor of CVD risk), and reduces CVD risk		Further research is needed to advise on dosing-time of BP-lowering medications for patients with nocturnal hypertension or non-dipper 24 h BP pattern. No data and conclusions on effect of evening, compared to morning, treatment on asleep BP control and dipping pattern (ABPM was not incorporated into trial)
Other conclusions	No treatment-time difference in adverse effects, including hypotension assessed by 48 h ABPM. Excellent adherence and compliance to prescribed circadian dosing time (utilizing awakening and bedtime as surrogate markers)		Fewer participants in the evening dosing group reported adverse events. As expected, lower compliance with dosing at evening (20:00–00:00) vs. morning (06:00 h–10:00 h)

ABPM: ambulatory blood pressure monitoring. BP: blood pressure. CKD: Chronic kidney disease. CVD: cardiovascular disease. DBP: diastolic blood pressure. SBP: systolic blood pressure. TIME: Treatment in Morning versus Evening.

*Reported timeline and sample size information for the TIME study is markedly inconsistent in different published reports by the same main investigators. For example, Rorie et al. (2017) – in a paper submitted on June 16, 2017 – reported 31,695 self-registered individuals, 21,116 of whom were randomized, and recruitment ending on March, 2017. On the contrary, Mackenzie et al. (2022) reported 24,610 registered and 21,104 randomized individuals and dated end of recruitment on June 5, 2018.

^†(i) In the Hygia Chronotherapy Trial (HCT), all information extracted from the electronic medical record of each recruited participant; CKD defined as: either estimated glomerular filtration rate <60 ml/min/1.73 m², albuminuria (albumin/creatinine [Cr] ratio ≥30 mg/g_Cr), or both, on at least two occasions ≥3 months apart. (ii) In the TIME study, all information self-reported by the participants through internet and/or e-mail; CKD reported as unspecific “impaired kidney function”. (iii) Previous major CVD events of the HCT include: myocardial infarction, ischemic stroke, hemorrhagic stroke, coronary revascularization, and heart failure, the latter two not reported in TIME.

Collectively, these deficiencies, especially entailing the above last two items, suggest the proportion of participants in the TIME study who truly ingested BP-lowering medications in the evening might have been similar in both treatment-time groups defined by the initial randomization. This, therefore, invalidates the reported results and conclusions that are based on the investigators´ intention-to-treat method of data analysis, despite their knowledge of the high rate of non-compliance of the participants to their assigned treatment-time. Additionally, collective findings of our systematic review and meta-analysis of published ingestion-time hypertension trials document the clinically relevant superiority of the at-bedtime/evening treatment schedule by: (i) enhancing asleep SBP reduction (without induction of sleep-time hypotension) and in the absence of significant treatment-time differing effect on awake SBP mean, resulting in increase of the sleep-time relative SBP decline towards the lower-CVD-risk normal dipper 24 h BP pattern; and (ii) diminishing the severity of the hypertension-associated target organ pathology of the kidney (by reducing albuminuria and/or increasing glomerular filtration rate) and heart (by decreasing left ventricular posterior diameter and left ventricular mass) (Hermida et al. 2021d). Thus, it does not seem biologically plausible that such benefits do not have the advantageous effect of decreasing CVD risk. Finally, the TIME study team concluded that “further research is needed to advise on dosing time of BP-lowering medications for patients with nocturnal hypertension or other disorders of diurnal BP variation” [non-dipper/riser 24 h BP pattern, we assume]. In reality, as discussed above, the asleep SBP mean and sleep-time relative SBP decline (non-dipper/riser 24 h BP pattern) are in combination the strongest predictors of CVD risk and when abnormal the documented therapeutic targets for CVD risk prevention (Hermida et al. 2011, 2018), regardless of the wake-time OBPM and awake or 24 h ABP means, leading to the recommendation that the true diagnosis of hypertension and prescription of pharmacologic treatment to manage it be restricted only to individuals who exhibit an altered ABP characterized by an elevated asleep SBP mean and/or non-dipper BP pattern (Crespo et al. 2020).

Bedtime versus morning use of antihypertensives for CVD risk reduction (BedMed) trial

The still ongoing Canadian BedMed trial (Garrison et al. 2022) is a PROBE, pragmatic, and decentralized study that shares some of the limitations in its design as the above-discussed TIME study, including:

• Participants being followed only by telephone. The degree of interaction during follow-up between the trial investigators and the primary care providers, if any, being uncertain.

• All participants being already treated for hypertension upon recruitment and registration of baseline treatment scheme and changes in therapy during follow-up being unknown.

• Participants being likely diagnosed as hypertensive solely on the basis of the wake-time OBPM, thus including individuals with low-risk masked normotension and excluding those with high-risk masked hypertension.

• Limited number of follow-up BP measurements being obtained from participants. A single 24 h ABPM evaluation was scheduled for a small cohort of 302 participants only at 6 months following recruitment, but in the absence of baseline data due to the failure to conduct an ABPM assessment upon recruitment of these patients, it is impossible to evaluate differences in the achieved therapeutic BP-lowering efficacy between the two treatment-time groups. Moreover, no other clinically relevant variables – including laboratory testing of blood and urine and evaluation of target organ damage – are scheduled to be assessed at any time during the study.

• Study protocol lacking specific recommendations on how to handle adjustments of the participants´ BP-lowering therapy if required. The BedMed study protocol only states “if bedtime use is problematic, participants will be asked to ingest their medication with dinner; if morning use is problematic, participants will be asked to ingest their medication with lunch,” thus compromising adherence to the assigned treatment-time at recruitment and consequently the integrity of study findings. Furthermore, this recommendation implying equivalence between the ingestion of treatment in the “morning” and “lunch” (or between “bedtime” and “dinner”) disregards the basic concepts and implications of the field of chronopharmacology relating to the circadian rhythm-dependent effects of BP-lowering medications (Hermida et al. 2020b) and, as demonstrated by Umeda et al. (1994), differences on BP regulation when the same medication is ingested by the same individuals upon-awakening on an empty stomach vs. with a meal at breakfast in the morning.

• The BedMed trial, apart from the randomization of participants to the defined “morning” and “bedtime” treatment-time groups, not complying with any of the other eight quality requirements and recommendations (Table 1) of the ISC/AAMCC guidelines for the design and conduct of human clinical trials on ingestion-time differences of hypertension medications (Hermida et al. 2021f). As a consequence, it is rated with an ISC/AAMMCC quality score of only + 0.5.

The reported design of the BedMed trial evidences additional concerns, apparently not shared by TIME, such as:

1. Defined as being an event-driven study, the required sample size calculation was initially based on a prespecified but unjustified number of 406 primary outcomes (composite of all-cause death, myocardial infarction, stroke, and heart failure), rather than on the actual or expected event-rates (Garrison et al. 2022). Due to the lower-than-expected recruitment of only 6% of the contacted candidates, the BedMed investigators decided to reduce the event target to 254 with the only justification given for selecting this number being to match the number of events reported in the MAPEC study (Hermida et al. 2010). However, the actual 255 reported events in MAPEC included, apart from those composing the stated BedMed primary outcome, ones of angina pectoris, coronary revascularization, aortoiliac occlusive disease, and thrombotic occlusion of the retinal artery (Hermida et al. 2010), which are not listed as being recorded for evaluation in BedMed (Garrison et al. 2022). Additionally, participants in the MAPEC Study were at recruitment either untreated hypertensives or resistant hypertensive patients, all being diagnosed as such according to ABPM criteria, not OBPM, and who were receiving medical care at a tertiary hospital. Most importantly, a large proportion of participants of the MAPEC Study were at relatively high CVD risk, i.e., those with conditions of diabetes (21%) and CKD (31%) and those having non-dipper/riser 48 h BP pattern (54%), therefore, representing a cohort markedly different than the comparatively lower CVD risk one of the BedMed trial.

2. The number of reported major CVD events (CVD death, myocardial infarction, heart failure, ischemic stroke, and hemorrhagic stroke), number of recruited participants, and median duration of follow-up in the HCT (1,450 events, 19,084 participants, 6.3 years, respectively), TIME study (954 events, 21,104 participants, 5.2 years), and BedMed trial (254 targeted events, 3,227 participants up to 2022, targeted median duration of follow-up unreported), differ markedly, indicating extensive disparity in CVD risk of the participants of the TIME and BedMed cohorts relative to those of the HCT, eventually limiting the capability of BedMed to document significant treatment-time differences in CVD outcome.

3. The allowance of twice-daily medications to be converted to once-daily alternatives, although “not actively promoted” (Garrison et al. 2022). Treatment in this manner with split-doses of medications at both tested treatment-times causes uncertainty to which “randomized” group these so-treated participants might eventually be assigned, further compromising the assessment and interpretation of findings.

Discussion

In the middle of the twentieth century, Biological Abstracts, Index Medicus, and other indexing periodicals listed only around 200 to 300 publications per year that specifically involved circadian rhythm studies of plants, aquatic creatures, laboratory animals, and human beings. Since then, the number of publications devoted to circadian rhythms has grown tremendously. In those early years, such investigations were primarily conducted by a relatively limited number of academic biologists trained in the science and methods of chronobiology, the field of biological rhythm study. Today, scientists of many different disciplines, who have had no or little formal schooling of the methods of quality circadian rhythm research, in particular, the design and implementation of translational circadian medicine projects, are leading large patient outcomes trials. In the absence of the necessary expertise, certain recent endeavors are fundamentally deficient. Nonetheless, the non-representative findings of these flawed investigations have been accepted for publication in reputable medical journals and are even cited in clinical practice guidelines, because reviewers solicited to critique circadian medicine reviews, meta-analyses, and clinical trials often lack sufficient knowledge of medical chronobiology and chronotherapeutics to adequately assess their merit.

We authors were trained in the concepts and methods of circadian rhythm research, and we have spent our academic career investigating 24 h rhythms and their translational applications to the clinical practice of medicine, especially the accurate diagnosis of hypertension by around-the-clock ABPM and optimization of the efficacy and safety of hypertension treatment. Over the years, we have carried out investigations that respect the requirements for high-quality circadian rhythm research projects, particularly ones concerning ingestion-time differences in the therapeutic effects of various classes of hypertension medications and through two large outcomes trials the differential reduction of the pathological consequences of elevated BP, especially risk for CVD events and development of diabetes and CKD. The motivation for this manuscript arose from a series of recent events: (i) letters to the editor of the European Heart Journal, which published the findings of the HCT, questioning the rationale of certain key features of its design and conduct, merit of the selected ABPM parameters, and validity of reported CVD outcomes; (ii) deficiencies of the methods of the just published review by Maqsood et al. (2023) and from our perspective its incomplete and flawed findings; (iii) shortcomings of the recently communicated pragmatic TIME study; and (iv) similar faulty attributes of the ongoing BedMed pragmatic trial. The methods of the recent review and TIME projects, undoubtedly done in good faith, that were utilized to explore administration-time differences, respectively, in the efficiency of hypertension medications to reduce BP and diminish CVD risk, are markedly deficient, particularly according to the criteria recommended by the working committee of the ISC/AAMCC (Hermida et al. 2021f). Thus, it is not surprising their findings and conclusions differ greatly from those of the more comprehensive systematic review and meta-analysis published in 2021 (Hermida et al. 2021c, 2021d) and also the high-quality HCT reported in 2020 (Hermida et al. 2020a). The discrepancy of findings between the high-quality HCT and the low-quality TIME trial has led to unnecessary confusion and controversy among the medical community and lay public about the merit of bedtime chronotherapy as a cost-effective means of potentiating the efficacy of hypertension medications to better control elevated BP and reduce the risk for CVD events, diabetes, and CKD and its progression, among other hypertension-associated pathologies.

Maqsood et al. (2023) erroneous conclusions on the merits of bedtime hypertension chronotherapy are, unfortunately, shared by authors of other poorly conceived “systematic reviews.” For example, apart from the faulty review and meta-analysis by Abuelazm et al. (2023) mentioned above, a recent consensus statement by the International Society of Hypertension (ISH) (Stergiou et al. 2022), based on their own markedly wrong and incomplete “systematic review,” states the treatment-time-dependent effects of BP-lowering treatment on BP reduction have been investigated by around “50 trials reporting ABP data” [in actuality, the correct number is 127 trials (Hermida et al. 2021c)] that provided a “modest reduction in 24 h BP means (our italics) of about 2/1 mmHg,” and it once again cites the underpowered and very low quality “neutral” HARMONY trial of Poulter et al. (2018) as proof of the lack of treatment-time differences in efficacy. Given the overlap of the authorship of the HARMONY, TIME, and ISH publications, it is of no surprise that they share the same deficiencies in their investigative methods. From our perspective, the ISH statement is invalid because its authors rely exclusively on the irrelevant 24 h BP mean as the key outcome variable to base their opinion, without recognition that individuals with markedly different 24 h BP dipping phenotype and either low, normal, or elevated asleep BP mean, who are at markedly disparate CVD risk, might have the exact same 24 h BP mean (Hermida et al. 2013b, 2021e). They also state “treatment-time is unlikely to play an important role on BP-lowering when using long-acting medications,” disregarding not only the markedly opposite conclusions of the numerous randomized studies of long-acting ARBs, ACEIs, CCBs, and their combination therapies (Hermida et al. 2021c, 2021d) but the basic concepts and evidence of publications concerning the chronopharmacokinetics and chronopharmacodynamics of hypertension medications (Hermida et al. 2020b). Moreover, they apparently are unaware of possible circadian rhythm-dependent differences in the serum concentration-medication effect relationship, which has been recognized by circadian medicine researchers for more than four decades (See Reinberg 1983), plus the substantiated clinical advantage of targeting hypertension therapy to optimize the control of sleep-time BP, as opposed to wake-time BP, to successful reduce hypertension-associated pathology in the most meaningful manner (Hermida et al. 2011, 2018). Presumably, the limited recognition of these perspectives led Stergiou et al. (2022) to make the recommendation that the “treatment of hypertension should be based on long-acting agents in monotherapy or in combinations administered in a single morning dose, aiming at optimal and consistent 24 h BP control,” thereby disregarding the impossibility that such a therapeutic approach is appropriate to manage individuals with elevated sleep-time BP and/or non-dipper/riser ABP pattern, in actuality the only ones who should be deemed as truly hypertensive and who should ingest their BP-lowering medications at-bedtime rather than upon-awakening from sleep or later in the morning (Hermida et al. 2011, 2013b, 2018, 2020b, 2021a, 2021e, 2023).

Authors of reviews and meta-analyses pertaining to ingestion-time differences in the efficacy of hypertension medications commonly include quality assessment of the cited individual investigations that is solely restricted to the traditional concerns, among others, of sample size, trial duration, measured endpoint, measurement quality, treatment adherence, statistical methods, heterogeneity, and bias. To our knowledge, only one review (Hermida et al. 2021b) reports, as an integral feature, an in-depth evaluation of the quality of the design and conduct of past published hypertension medication chronotherapy trials, using as reference the extent to which there is compliance to the eight items identified by the ISC/AAMCC as being indispensable for the conduct of quality research of this type. Apart from the numerous errors committed by Maqsood et al. (2023) in retrieving important information from the reviewed published trials, as extensively documented herein, they failed to evaluate their quality according to circadian medicine perspective (Table 2) relative to reported findings. The vast majority of the studies reviewed by those authors that report disparate findings that the at-bedtime/evening treatment strategy is not superior to the upon-awakening/morning one are, according to the ISC/AAMCC eight-item rating system, of very low-quality, scoring < 2.0 out of a possible scoring of 7. The extensive errors committed by Maqsood et al. (2023) plus the failure to weigh the quality of each retrieved trial in relation to its findings invalidate the faulty results and conclusions of their review. Evaluation of the TIME trial by the same eight ISC/AAMCC criteria also discloses its very low-quality rating, a score of 1, also calling into question the validity of the findings and conclusion of this study that an evening-time hypertension chronotherapy does not better reduce CVD outcomes than a morning-time one. The ongoing pragmatic BedMed trial unfortunately not only shares most of the same shortcomings of the TIME study, but it also evidences additional concerns regarding, among others, its limited sample size, not specifically calculated in terms of the actual event-rate of the targeted population, and allowance of twice-daily medications that as a consequence is rated with a very low ISC/AAMMCC quality score of only + 0.5.

The purpose of this article has been exclusively appraisal of the quality of recent publications entailing the differential beneficial and adverse effects of hypertension medications when ingested upon-awakening/morning vs. at-bedtime/evening, respectively, on BP control and/or CVD outcomes, as appropriate, in reference to the ISC/AAMCC criteria. In the future, the conduct of other types of translational circadian medicine investigations is anticipated to become more numerous, since their implementation will be facilitated by medical-grade implantable and wearable monitoring devices that can collect data of clinical relevance continuously or at investigator-designated discrete times around the clock for many days. We advocate the methods of future translational pursuits, specially entailing such innovative monitoring technology, ensure adequacy of patient inclusion criteria (Item 1 of Table 1), proper designation of tested treatment times, preferably in terms of actual or representative surrogate markers of endogenous biologic-time (Item 2), optimal method of assessing the circadian oscillating outcome variable (Item 3), optimal choice of primary indicator endpoint(s) of assessment, preferably not the 24 h mean of collected values (Item 4), appropriate duration of patient monitoring to accurately diagnose them to ensure valid qualification for study inclusion and evaluate administration-time differences in effects of the trialed intervention (Item 5), correct “adjusted calculation” procedure to derive accurate wake-time, sleep-time, and other clinically relevant mean values of response variables (Item 6), right calculation of sample size required for valid data analyses, findings, and conclusions (Item 7), and acceptable protocol design, preferably a crossover one of sufficient washout period, or a randomized PROBE or double-blind one (Item 8). Satisfaction of these eight criteria hopefully will help ensure high-quality future circadian medicine trials and findings that will lead to meaningful advances in patient care and outcomes. Furthermore, we recommend these same elements of the ISC/AAMCC rating scale be utilized to guide not only the peer review of manuscripts devoted to circadian rhythm research to determine if they are of sufficiently high quality to be worthy of publication but to guide the assessment of the methods and goals of grant proposals submitted to governmental agencies, foundations, and other entities to rank their merit for financial support.

Disclosure statement

All authors have shares of Circadian Ambulatory Technology & Diagnostics (CAT&D), a technology-based company developed by and in partnership with the Universidade de Vigo.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.