Association between Time-of-Day for Eating, Exercise, and Sleep with Blood Pressure in Adults with Elevated Blood Pressure or Hypertension: A Systematic Review

Abstract

The purpose of this review is to synthesize results from studies examining the association between time-of-day for eating, exercise, and sleep with blood pressure (BP) in adults with elevated BP or hypertension. Six databases were searched for relevant publications from which 789 were identified. Ten studies met inclusion criteria. Four studies examined time-of-day for eating, five examined time-of-day for exercise, and one examined time-of-day for sleep and their associations with BP. Results suggested that later time-of-day for eating (n=2/4) and later sleep mid-point (n=1/1) were significantly related to higher BP in multivariable models, whereas morning (n=3/5) and evening (n=4/5) exercise were associated with significantly lower BP. Although this small body of work is limited by a lack of prospective, randomized controlled study designs and underutilization of 24-hour ambulatory BP assessment, these results provide preliminary, hypothesis-generating support for the independent role of time-of-day for eating, exercise, and sleep with lower BP.

Introduction

Hypertension is the leading preventable cause of overt cardiovascular diseases (CVD) and all-cause mortality both globally and in the United States (US) [1]. Consistent with this level of burden, almost half of all adults in the US are hypertensive (46.7%) [2], and in approximately 50% of cases, their hypertension is uncontrolled [3]. Each 5-mmHg reduction in systolic, and 2-mmHg reduction in diastolic blood pressure (BP) can reduce the risk of subsequent CVD by 6–10% [4–6]. However, rates of BP control are declining [7], such that the number of hypertension-attributable deaths have risen by 65% over the last 10 years [8]. Moreover, antihypertensive medications have been shown to be insufficient in curtailing the progression of hypertension to overt CVD [9,10]. These converging lines of evidence suggest that novel, auxiliary first-line lifestyle strategies are urgently needed to achieve national goals of reducing high BP [11] and lowering CVD death rates [12–14].

Achieving adequate sleep duration, eating a healthy diet, and being physically active are key health behaviors for lowering BP and preventing CVD [15]. Public health guidelines for these behaviors are largely quantity based. For example adults should eat ≥3 servings of vegetables, ≥2 servings of fruits [16], achieve ≥150 minutes of moderate to vigorous aerobic physical activity each week [17], and get 7–9 hours of sleep per night [18]. The cardioprotective effects of achieving these quantity-based guidelines are well demonstrated, however, they fail to embrace recent scientific advances in chronobiology (i.e. the study of biological rhythms [19]) [20]. The field of chronomedicine seeks to prevent or treat illnesses according to biological rhythms, or more specifically, align the time-of-day for the administration of a therapeutic agent with the body’s endogenous rhythms to optimize therapeutic efficacy or reduce side effects [21,22]. Hypertension may be particularly amenable to chronotherapeutic approaches since BP follows a predictable circadian rhythm that typically peaks after waking and lowers during sleep. The absence of appropriate BP decline during sleep, or ‘nondipping’ (or reverse dipping if there is an antithetical increase in BP during sleep) and severe spikes in the morning increase in BP are key determinants of adverse cardiovascular outcomes [23]. Evening versus morning antihypertensive dosing has been hypothesized to lower BP and promote dipping during the critical nighttime period as well as protect patients with hypertension during the morning BP surge [24]. Indeed, a Cochrane systematic review of 21 randomized controlled trials indicated that evening versus morning dosing of antihypertensives was associated with significantly lower 24-hour ambulatory BP [25]. However, some studies have shown no benefit of evening versus morning antihypertensive dosing [26,27]. This body of evidence provides some support that nighttime versus morning dosing of antihypertensive medications improves ambulatory BP and reduces the incidence of major cardiovascular events [25,27–30]. There is a need to extend this chronomedicine paradigm to identify an optimal time-of-day to engage in cardio-protective health behaviors such as eating, exercise, and sleep (i.e., ‘chronobehaviors’) that promotes reduced BP.

To date, there is a growing body of research examining the association between time-of-day for eating, exercise and sleep on BP and hypertension control. In the context of time-of-day for eating, the effects of different feeding/fasting regimes, such as intermittent fasting and time restricted eating with BP and hypertension control have been considered. Intermittent fasting alternates between a period of fasting and a period of typical eating. Common intermittent fasting regimes include the 5:2 plan (i.e., eating ad-libitum for five days, and 75–80% energy restriction or fasting for two days) and the alternate day fasting plan (i.e., fast day followed by an eating as usual day) [31]. Time restricted eating is a version of intermittent fasting that consistently limits eating windows to a period of 4–12 hours, thus requiring 12–20 hours/day of fasting [32]. Restricting energy consumption to daylight hours, independent of diet quality or caloric intake, has been shown to significantly reduce systolic BP [33]. More recently, there have been investigations into more specific time-of-day for eating effects on cardiometabolic outcomes. For example, delayed eating (e.g., skipping breakfast) is common in the general population [34] and has been independently associated with hypertension [35]. While eating dinner at 22:00 versus 18:00 was associated with lower levels of fat burning and higher blood sugar [36]. Meta-analytic data have shown that greater caloric intake earlier in the day was associated with greater short-term weight loss compared to eating later in the day in the context of an energy-reduced diet [37].

Associations between time-of-day for exercise and sleep with BP and hypertension control have also been investigated, with mixed results. Some studies suggest morning exercise is associated with improved cardiovascular markers (independent of exercise duration and intensity), [38] while others suggest afternoon exercise is better [39]. In terms of time-of-day for sleep, bedtimes ≥23:00 are reported in 40–80% of adults [40–42] and have been independently associated with elevated resting BP [43] (vs. earlier bedtimes). Collectively, while these studies provide a scientific basis to consider the role of time-of-day for health behaviors such as eating, exercise and sleep on BP, there has yet to be a systematic review of this literature that synthesizes results. To address this gap, the current review sought to identify and synthesize findings from studies examining the independent associations between estimated time-of-day for eating, exercise, and sleep with measured BP in adults with hypertension or elevated BP. It is expected that the synthesis of rigorous, extant literature in this area will directly inform hypotheses about the effects of time-of-day for eating, exercise and sleep on BP that can be subsequently tested in controlled studies [44]. We ultimately expect this line of inquiry to and guide the development of chronobehavioral approaches to improve BP control and reduce CVD risk.

Methods

Search Methodology

This systematic literature review was registered with the International Prospective Register of Systematic Reviews (PROSPERO; Registration no. CRD42023407779) and completed using the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [45]. An expert health science librarian conducted a comprehensive search of six bibliographic databases: PubMed, CINAHL Plus with Full Text, Scopus, SPORTDiscus with Full Text, Web of Science, and APA PsycINFO (ProQuest). The search was conducted in February 2023, and updated in July 2023. The search strategy covered dates from January 2013 through July 2023 with an English language limit applied. The publication date range was limited to the previous ten years to ensure that only the most recent studies, reporting on current recommendations for gold-standard assessments of eating, exercise, and sleep, were included. No other publication restrictions were applied. The search strategy was developed in PubMed and was adapted to fit the constraints of the other databases. Keywords and database specific controlled vocabulary (e.g., PubMed’s Medical Subject Headings, CINAHL Headings, SPORTDiscus’ Sport Thesaurus, and PsycINFO’s Thesaurus of Psychological Index Terms) were utilized to select terms related to the following: hypertension or blood pressure, time-of-day or chronotype, and health behaviors (diet, exercise, or sleep). Additionally, handsearching of the references of included articles was performed to identify additional eligible studies. For a complete search strategy, see Supplementary Table 1.

Eligibility Criteria

We used the PICOS (Participants, Intervention/Exposure, Comparisons, Outcomes, Study design) framework to define the search process and study concepts (see Table 1 for listing).

Table 1:

Study Inclusion and Exclusion Criteria

Inclusion Criteria

1. Published in a peer-reviewed scholarly journal between January 1, 2013, and July 31, 2023.

2. Published in English

3. Study reports on a primary data collection in study participants who are adults (aged >18 years) and have at least elevated blood pressure (BP ≥120/80 mmHg).

4. Study reports on objective or “gold standard” estimation (i.e., using accelerometry) of time-of-day for exercise and/or sleep for 4 or more days; dietary recall methods for 1 or more days.

5. Study uses statistical methods to test the association between time-of-day for one or more of the health behaviors of eating, exercise, or sleeping with blood pressure.

6. Resting or ambulatory systolic and/or diastolic blood pressure is reported as a study outcome.

7. Study uses a controlled intervention, observational cohort/cross-sectional or before-after (Pre-Post) design.

Exclusion Criteria

8. Study conducted in children or women who are pregnant.

9. Study design is a review or meta-analysis.

10. Studies on animals.

11. Reports on a meeting abstract or unpublished material.

Participants:

Adults (>18 years) with at least elevated BP (BP ≥120/80 mmHg). Adults with elevated BP were included because of the graded risk with BP above optimal levels with CVD risk and mortality [46].

Intervention/Exposure:

The exposure of interest is time-of-day for one or more of the following health behaviors: eating, exercise, sleep. To be included in this review, an objective (i.e., using accelerometry) or “gold standard” assessment technique for time-of-day (e.g., onset, offset, or midpoint) for exercise or sleeping, or dietary recall records for eating was required.

Comparisons:

Studies included in this review were required to test the association between the time-of-day for one or more of the health behaviors (eating, exercising, sleeping) with BP.

Outcomes:

The outcomes of interest were mean systolic and diastolic BP, measured at rest or using ambulatory methodologies.

Study design:

Original research studies included in this review were required to use one of the following designs: (1) Controlled Intervention Studies, (2) Observational Cohort and Cross-sectional, or (3) Before-After (Pre-Post) with no control group. Studies of multiple design types (i.e., RCT, cohort, cross-sectional) were included in this review so as to provide the most comprehensive review of literature examining this emergent area of inquiry. Precedent for including multiple types of study designs was provided by previous systematic reviews in related areas [47–49].

Studies excluded were those that did not meet the inclusion criteria, were conducted in children, women who are pregnant, or animals, reported on a meeting abstract or unpublished material, and were published in a language other than English. Gray literature (i.e., unpublished material, meeting abstracts) were not searched because of the lack of peer-review in this type of literature and evidence suggesting that when included, gray literature represents only 2% of studies included in systematic reviews [50]. A systematic review with a narrative synthesis of findings was used since this body of research could not be summarized quantitatively (i.e., via a meta-analysis) due to the heterogeneity of research designs, statistical analyses, and outcome variables [51,52].

Data Management, Extraction and Analysis

Citations and abstracts of studies retrieved by the search strategy were downloaded to Covidence and screened independently by two reviewers (TK and FP) to identify studies that potentially met the inclusion criteria. The full text versions of these potentially eligible studies were retrieved and independently assessed for eligibility by two review team members. A standardized form was used to extract data from the included studies for assessment of study quality and evidence synthesis. Extracted information included study setting; study population, participant demographics, baseline characteristics; study design; time-of-day assessment methodology for targeted health behaviors, assessment methodology for BP (including time of BP assessment when resting BP was measured), and results (unadjusted associations between time-of-day for health behaviors with BP, adjusted associations between time-of-day for health behaviors in controlled multivariable models). Two reviewers extracted data independently, discrepancies were identified, and resolved.

Study Quality, Data Extraction and Synthesis

Two reviewers (TK and FP) independently assessed the methodological quality of the selected studies with the National Heart, Lung, and Blood Institute’s Quality Assessment Tools for Controlled Intervention Studies, Observational Cohort and Cross-Sectional Studies, Before-After (Pre-Post) Studies with no Control Group, depending on the study design (https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools).

Studies were grouped and examined by the independent variable being considered (i.e., eating times, exercise times, sleep times). Quality rating was based on study design, following the guidelines of study quality assessment as outlined by the NHLBI study quality assessment tool. Study characteristics extracted from each study included study aim, country conducted, study design, population, time-of-day for which independent variable being considered, and assessment methodologies for both the independent and dependent variables. The results of each study were also extracted. Key variables considered included the covariates considered in multivariable modeling, the reported magnitude of adjusted association between the independent and dependent variable, and an overview of key study findings.

Results

Study Selection

The electronic search strategy yielded 789 references from six databases; 333 duplicate references were identified and removed. The remaining 456 references were screened for inclusion, with 22 being identified as eligible for full-text review. Nine of the 22 full-text articles met study eligibility. (see Figure 1 for PRISMA diagram). The reference list in each of the eligible articles were screened for study inclusion. This hand-search provided one additional eligible study to yield a final study sample of ten articles.

Figure 1:

PRISMA Diagram to show Process of Identification of Studies via Databases

Time-of-day for Eating

Study Quality and Characteristics

Four studies examined the association between time-of-day for eating and BP [35,53–55]. Three used a cross-sectional design [53–55] and one used an observational cohort design [35]. All three cross-sectional studies earned a quality rating of fair. The observational cohort study [35] earned a quality rating of good (see Supplementary Table 2b).

Time-of-day for eating was evaluated by 1- [54], 2- [53], or 3-day [55] dietary recall or 5-day photo-assisted diet diaries [35]. Resting BP was assessed in all studies.

Associations Between Time-of-Day for Eating and Blood Pressure

Two of the four studies reported that later food consumption was associated with a higher BP [35,53] and greater likelihood of having [53] or developing [35] hypertension in multivariable models. Leech et al. (2018) [53] (n=4482) showed that women who consumed food at later times (lunch 1 h later than 12:00–1:00 PM) had 2.45 mmHg greater SBP (β = 2.45 mmHg, CI = 0.05 – 4.84, p ≤ 0.05), 1.69 mmHg greater DBP (β = 1.69 mmHg, CI = 0.25 – 3.13, p ≤ 0.05), and a 45% greater likelihood of having hypertension (OR = 1.45, CI = 1.00 – 2.22, p ≤ 0.05).

Almoosawi et al. (2013) [35] (n=1152) demonstrated that participants in their prospective cohort study consuming a greater proportion of total energy intake in the late evening (no specific time window defined) had 5.09 mmHg greater SBP (β = 5.09, CI = 1.25 – 8.93, p ≤ 0.05), 2.80 mmHg greater DBP (β = 2.80, CI = 0.27 – 5.32, p ≤ 0.05), and a 55% greater likelihood of developing hypertension after 10 years in comparison to those who consumed the least proportion of total energy intake in the late evening (OR=1.55, CI = 0.93 – 2.61, p ≤ 0.05). This study also showed that consuming a greater proportion of total daily energy intake at breakfast was cross-sectionally associated with 30% lower odds of hypertension compared to consuming the least proportion of total energy intake at breakfast (OR=0.70, CI = 0.45 – 1.08, p ≤ 0.01).

Shim and colleagues (2021) [54] (n=13,361) and Dote-Montero et al. (2023) [55] (n=118) found no significant associations between caloric midpoint and eating mid-point, respectively and BP in multivariable models.

Time-of-day for Exercise

Study Quality and Characteristics

Five studies examined the association between time-of-day for exercise and BP [56–60]. Of these, two studies used a randomized controlled trial [56,59], and three used a randomized crossover design that were evaluated as pre-post designs for quality review [57,58,60]. Both randomized trials earned a study quality rating of ‘poor. The three randomized crossover studies were rated as ‘poor’ (n=1) [57], ‘fair’ (n=1) [58], and ‘good’ (n=1) [60] (see Supplementary Table 2c). Time-of-day for exercise was operationalized via supervised exercise training sessions performed in the morning (7:00–9:00 [56]; 9:00–10:00 [57]; 7:00–11:30 [58]; 6:30–8:30 [59]; 7:00–7:30 [60]), afternoon (13:00–13:30 [60]), or evening (18:00–20:00 [56] [59]; 18:30–19:15 [57]; 17:00–21:00 [58]; 19:00–19:30 [60]) as part of interventions lasting 3 days [60], 6 days [57,58], 10 weeks [56], or 12 weeks [59]. Resting BP (n=2/5) or 24-hour ambulatory methods (n=3/5) were utilized in these studies.

Association Between Time-of-Day for Exercise and Blood Pressure

Two of the four studies reported that evening exercise was associated with significantly lower SBP than exercise conducted in the morning [56,57]. Specifically, Brito et al (2019) [56] (n=50) reported that in evening evaluations of resting BP, evening (i.e., 18:00–20:00) exercise was associated with significantly greater reductions in SBP as compared to morning (7:00–9:00) exercise (d = −0.61, CI = −1.33 - +0.13) [56]. Similarly, Brito et al. (2018) [57] (n=13) reported that there were significantly greater reductions in SBP following evening exercise (18:30 and 19:15; 116 ± 11 mmHg) versus a seated control condition (120 ± 10 mmHg) [57]. By contrast, one study conducted by Brito et al. (2015) [58] (n=16) reported that morning exercise between the hours of 7:30–11:30 significantly reduced SBP by an additional 4 mmHg as compared to evening exercise conducted between the hours of 17:00–21:00 [58].

Fairbrother et al. (2014) [60] (n=20) showed no statistically significant differences between acute bouts of morning and evening exercise with BP. However, they did demonstrate that a 30-minute bout of exercise performed at 07:00 in comparison to a 30-minute bout of exercise at 13:00 reduced nighttime SBP (107 ± 4.8 vs 114.8 ± 5.2 mmHg; p = 0.05), nighttime DBP (60.8 ± 0.89 vs 65.6 ± 0.99 mmHg; p = 0.038), and evoked greater nocturnal SBP (16.67% ± 0.93% vs 12.3% ± 1.2%; p = 0.034) and DBP (20.9% ± 1.15% vs 15.4% ± 1.6%; p = 0.031) dipping. Additionally, a 30-minute bout of exercise at 19:00 in comparison to a 30-minute bout of exercise performed at 13:00 significantly reduced nighttime DBP (61.8 ± 0.98 vs 65.6 ± 0.84 mmHg; p ≤ 0.05) and evoked a greater nocturnal DBP dip (21.1% ± 1.9% vs 15.4% ± 1.6%; p ≤ 0.05) in adults with pre-hypertension.

Time-of-day for Sleep

Study Quality and Characteristics

One cross-sectional study meeting study eligibility criteria examined the association between the time-of-day for sleep and BP [61]. This study earned a quality rating of fair. (see Supplementary Table 2b). The sample for this study was n=2156 Hispanic adults. Time-of-day for sleep was assessed using 7-day wrist accelerometry. The variables of average sleep onset, offset, and midpoint across seven days were estimated. Resting BP was measured using an automated sphygmomanometer [61].

Associations Between Time-of-Day for Sleep and Blood Pressure

Results from this study suggest that each 1-h delay in sleep midpoint was associated with an increase in SBP and DBP of 0.73 (β = 0.73, CI = 0.30 – 1.17, p < 0.01) and 0.53 (β = 0.53, CI = 0.17 – 0.90, p < 0.01) mmHg, respectively [61]. The associations remained significant after adjusting for age, sex, Hispanic/Latino ethnicity, study site, income, acculturation, education, sleep, duration, and apnea-hypopnea index but lost significance after the inclusion of shift work status as a covariate. Post-hoc stratified analysis showed that the association between time of day for sleep and BP was only significant in multivariable models in non-shift workers [61].

Discussion

The purpose of this systematic review was to identify and synthesize peer-reviewed literature that examined the independent role of time-of-day for eating, exercise, and sleep behaviors on BP in participants with elevated BP or hypertension. The key findings indicated by this review include: (1) Later time-of-day for caloric consumption was associated with higher BP in two of the four fair-quality studies reviewed; (2) Either morning or evening exercise was associated with lower BP in five studies that were rated low-quality and; (3) Later sleep midpoint was associated with higher BP in one study of fair quality. Together, these data support a proof-of-concept that time-of-day for eating, exercise and sleep may be independently associated with BP, and as such help inform hypotheses for future studies to test the impact of chronobehavioral interventions on BP and CVD risk in adults with elevated BP or hypertension.

Time-of-Day for Eating

According to the studies reviewed here, later time-of-day for caloric consumption (e.g., an eating window midpoint after 14:00h) was associated with significantly higher BP [35,53]. There are several metabolic mechanisms that may help explain why eating later in the day and closer to nighttime might be associated with higher BP. Later time-of-day for dietary intake (i.e. during sleep time) evokes metabolic dysfunction (e.g., alterations in adipose, hepatic, and skeletal muscle insulin sensitivity/glucose metabolism) that promotes disease development, such as metabolic syndrome [62–64]. The feeding/fasting cycle is a potent contributor to the circadian regulation of many cardiometabolic biomarkers (e.g., lipid metabolism [65]) that are important for risk factors associated with the development of hypertension [66]. Interestingly, the misalignment of time-of-day for eating in relation to the light/dark and sleep/wake cycles also may evoke the development of metabolic syndrome as demonstrated in animal models. This effect that could be even further exacerbated by other circadian disruption (i.e. misaligned sleep/wake cycles) [62]; however, the effect of multiple misaligned systems was not a criterion for inclusion in this review. Specifically, delayed dietary intake exacerbates desynchronization of central and peripheral clocks, which can induce disruptions in metabolic function known to influence BP regulation (e.g., weight gain, reductions in insulin sensitivity) [64]. The current findings converge with the results of previous reviews showing greater caloric intake in the evening/night to be significantly associated with impaired metabolic profiles [67,68], and extends this work by showing its relevance to the cardiovascular metric of hypertension and BP. Additionally, although 2/4 studies included [53,55] variables to adjust for diet quality/type, future studies should address this as a potential confounder to the relationship between time-of-day for eating and BP due to the known effects of some specific nutrients (e.g., sodium, potassium) on BP. Although ‘later’ time-of-day for eating is generally associated with elevated BP, our review of literature only identified four articles explicitly testing this relationship, one of which reported null findings. Not yet clear is the optimal time-of-day for caloric consumption in the 24-hr day that is most strongly associated with reduced BP.

Time-of-Day for Exercise

Studies from this review suggest that the association between the time-of-day for exercise with BP may follow an inverted-U pattern. Specifically, morning and evening exercise may be associated with lower BP in comparison to afternoon exercise, however, significant work is needed to confirm this relationship. Exercise is known to be a powerful zeitgeber (environmental cue capable of influencing circadian rhythmicity) through alterations in skeletal muscle activity, a primary element of the peripheral clock [69,70]. Morning exercise may be associated with lower BP through decreased cardiac output, perhaps secondary to sympathovagally-mediated reductions in heart rate [58]. Conversely, evening exercise may relate to lower BP because of a decrease in vasomotor sympathetic modulation and reductions in systemic vascular resistance [56]. Evening exercise may also accelerate melatonin release that may alter cardiac autonomic function, such that parasympathetic tone predominates [71–73].

The optimal time of day for exercise as it relates to BP may also be influenced by an individual’s circadian preference (i.e., individual preference for morning or evening activities and for sleep/wake timing [74]). For example, in a parallel-group cross-over study where 30 adults with type-2 diabetes completed an exercise intervention at a time-of-day aligned with their circadian preference versus a time-of-day that was not aligned, results showed that completing exercise at the individual’s preferred time-of-day was associated with better glucose metabolism, lipid profile, and functionality in adults with type 2 diabetes [75]. In line with this, data from an observational study of 116 healthy adults, showed that when participants exercised during their preferred time-of-day, their frequency of exercising across a 4-week monitoring period was higher as compared to adults who did not exercise during their preferred time-of-day [76]. Another important consideration are various exercise characteristics that are known to impact BP (i.e. type intensity, duration, frequency) which varied between the included studies (Table 2). These data underscore the potential moderating effects of time-of-day preference for exercise on the relationship between the time-of-day for exercise and BP that warrant consideration in the defining of chronobehavioral approaches for BP management and control.

Table 2:

Study Characteristics

Reference	Study Aim	Country	Study Design	Study Population	Time-of-Day for Which Health Behavior being Considered (i.e., Independent Variable)	Assessment Methodology for Independent Variable	Assessment Methodology for Dependent Variable	Time of BP Measurement
Eating Timing
Leech et al. 2018 [53]	Examine associations of eating patterns with blood pressure.	Australia	Cross-sectional	4482 Adults ≥ 19 yrs old	“Conventional” eating pattern “Later lunch” eating pattern “Grazing” pattern	2 non-consecutive 24-h diet recall	Resting blood pressure (SBP & DBP)	Not Reported
Almoosawi et al. 2013 [35]	Examine the associations of eating occasion timing and blood pressure.	England, Scotland, Wales	Observational Prospective Cohort	1152 Adult men and women	Energy quintiles based on energy intake distribution across time	5-day diet diary	Resting blood pressure (SBP & DBP)	Not Reported
Shim et al. 2021 [54]	Examine the association of late eating and blood pressure and cardiometabolic risk factors.	Korea	Cross-sectional	13,361 Adult men and women with hypertension	Early eating: eating midpoint before 13:45 Late eating: eating midpoint after 13:45	1-day 24-h diet recall	Resting blood pressure (SBP & DBP)	Not Reported
Dote-Montero et al. 2023 [55]	Examine the association of meal timing with BMI and cardiometabolic risk factors.	Spain	Cross-sectional	118 Young adult men and women	Caloric midpoint Time interval between midsleep point and first food intake	3 non-consecutive 24-h diet recall	Resting blood pressure (Mean BP: DBP + 1/3(SBP – DBP))	Not Reported
Exercise Timing
Brito et al. 2019 [56]	Compare effects of 10 weeks of morning vs. evening aerobic training.	Brazil	RCT	50, 30–65 yr old men with treated hypertension	Morning Exercise: (7:00–9:00) Evening Exercise: (18:00–20:00)	Supervised cycling, 30–45 min., 3x/week for 10 weeks at moderate intensity	Resting and 24-h ambulatory blood pressure (SBP & DBP)	Morning (7:00–9:00) & Evening (18:00–20:00) 24-hr post-evening assessment
Brito et al. 2018 [57]	Examine blood pressure response to morning and evening exercise.	Brazil	Randomized Crossover	13, 20–45 yr old prehypertensive men	Morning Exercise: (9:00–9:45) Evening Exercise: (18:30–19:15)	Supervised cycling, 45 min., 1 session at 50% VO2peak and 1 control session at each time-of-day	24-h ambulatory blood pressure (SBP, DBP, mean BP)	24-hr, beginning 1-hr after morning, evening, and control session
De Brito et al. 2015 [58]	Compare the effect of morning vs. evening exercise on post-exercise hypotension.	Brazil	Randomized Crossover	16, 20–45 yr old prehypertensive men	Morning Exercise: (7:30–11:30) Evening Exercise: (17:00–21:00)	Supervised cycling, 45 min., 1 session at 50% VO2peak and 1 control session at each time-of-day	Resting blood pressure (SBP, DBP, mean BP)	Morning & Evening (45–60 min. after exercise session)
Arciero et al. 2022 [59]	Examine sex differences in physiological response and performance to morning vs. evening exercise.	United States	Randomized Trial	47 Adult men and women	Morning Exercise: (6:30–8:30) Evening Exercise: (18:00–20:00)	Supervised exercise (RISE – Resistance, Interval, Stretching, Endurance), 4x/week for 12 weeks, varying intensity/duration based on type of exercise	Resting blood pressure (SBP & DBP)	Not Reported
Fairbrother et al. 2014 [60]	Examine the impact of morning, afternoon, and evening exercise on ambulatory blood pressure and sleep.	United States	Randomized Crossover	20, 30–60 yr old prehypertensive men and women	Morning Exercise: (7:00–7:30) Afternoon Exercise: (13:00–13:30) Evening Exercise: (19:00–19:30)	Supervised treadmill exercise, 30 min., 1 session at 65% of heart rate obtained at VO2peak at each time-of-day	24-h ambulatory blood pressure (SBP & DBP)	24-hr, beginning immediately post-exercise
Sleep Timing
Abbott et al. 2019 [61]	Examine the association between sleep-wake timing and stability with cardiometabolic risk factors.	United States	Cross-sectional	2,156 Hispanic/Latino adults	Sleep onset Sleep offset Sleep midpoint	7-day wrist accelerometry	Resting blood pressure (SBP & DBP)	Morning (specific time not reported)

Time-of-Day for Sleep

One article met our inclusion criteria that examined the relationship between the time-of-day for sleep and BP. Results from this study indicate that delayed sleep timing was associated with increased systolic and diastolic BP in a sample of Latinx adults [61]. Given that only one study was found, no synthesis of evidence is possible. These results do align with previous studies suggesting that later time-of-day for sleep predicts cardiovascular events in adults [40–42] and hypertension in adolescents [43]. The associations between sleep characteristics and multiple cardiovascular outcomes, and the subsequent lack of evidence for the association between the time-of-day for sleep and BP posit a need for observational and longitudinal studies elucidating this paradigm.

Implications for Population Health and Research

The results of this systematic review highlight several important considerations for future research and population health. Considering the effect of time-of-day for daily health behaviors (eating, exercise, and sleep) on BP in adults with elevated BP is especially relevant in contributing to efforts that aim to optimize hypertension treatment and prevention. Current pharmacologically focused efforts to curtail the rapidly increasing prevalence of hypertension are inadequate [9,10], and therefore strategies that reduce BP through the enhancement of behavioral tactics by specific factors that include time-of-day prescriptions are needed. These preliminary data inform hypotheses for future research that should focus on identifying the independent role of time-of-day for eating, exercise, and sleep on BP through fully powered, high-quality clinical trials and longitudinal observations of free-living populations with elevated BP or hypertension that utilize validated assessments (e.g. 24-hour ambulatory BP) in the context of gold-standard, randomized controlled trials. Another line of inquiry that warrants further consideration is the role of circadian preference and the extent to which the association between time-of-day for engaging in eating, exercise, or sleep on BP may be modified by circadian preference. It is plausible that aligning the time-of-day for these lifestyle behaviors with circadian preference may optimize health benefits [75].

Strengths and Limitations

This review has several strengths and limitations that should be considered when interpreting results. A particular strength is the comprehensive search strategy performed by a highly experienced research librarian across multiple databases to identify eligible studies. Moreover, the use of rigorous methodology for screening titles/abstracts and full-text articles, data extraction, and quality rating using the National Heart, Lung, and Blood Institute quality rating tools specific to study design by two independent researchers enhanced the methodological rigor. Regarding the methodology, the inclusion criteria requiring device-based estimates or direct observation of sleep or exercise may have limited the studies included in this review; however, due to their superior validity and reliability as compared to other self-report methods, this was done to support increased rigor in study synthesis [77–79]. A key limitation of the identified studies was the considerable methodological heterogeneity. Sources of variability included different methodologies to assess and report time-of-day for the lifestyle behaviors. For example, self-report and device-estimates of sleep behaviors were found and time-of-day for lifestyle behaviors were sometimes reported as a continuous clock times and other times as a categorical range of times. Time intervals of intervention for observation varied across studies, as did time periods of follow-up. Moreover, the method of BP assessment varied between resting, and 24-hour ambulatory methods and criteria used for sample stratification was inconsistent. These variations in methodology limited our ability to make direct comparisons between studies.

Conclusion

This review found hypothesis-generating, preliminary evidence to support the role of time-of-day for eating, exercise, and sleep on BP in adults with elevated BP. These data suggest that delayed time-of-day for eating was associated with increased BP, and morning/evening exercise was more strongly associated with lowering BP than afternoon exercise. One study showed later time-of-day for sleep to be associated with increased BP. The few studies examining the role of time-of-day for each behavior on BP in adults with elevated BP demonstrated high variability in research methods, rigor, and quality. Further research that prospectively assesses the role of time-of-day for eating, sleep and exercise on BP using rigorous and validated assessments are needed so that chronobehavioral strategies to optimize current hypertension prevention and management approaches can be identified.

Table 3:

Association between Time of Day for Exercise, Eating, and Sleep with Blood Pressure

Reference	Variables Adjusted in Multivariable Model	Magnitude of Association between Independent and Dependent Variable	Results Summary	Quality Rating
Eating Timing
Leech et al. 2018 [53]	• Education • Country of birth • Smoking status • Physical activity • Sleep duration • Sedentary time • Currently dieting • Total daily energy intake • Diet quality • BMI	Women: Associations of SBP between ‘later lunch’ and ‘conventional’ eating pattern (β = 2.45 mmHg, CI = 0.05 – 4.84)* Associations of DBP between ‘later lunch’ and ‘conventional’ eating pattern (β = 1.69 mmHg, CI = 0.25 – 3.13)* Associations of HTN prevalence between ‘later lunch’ and ‘conventional’ eating pattern (OR = 1.45, CI = 1.00 – 2.22)*	Men No association between eating patterns and SBP or DBP. Women A later eating pattern was associated with higher SBP and DBP (2.45 and 1.69 mmHg, respectively) when compared to a conventionally timed eating pattern. A later eating pattern was associated with a 45% higher likelihood of having HTN as compared to a conventional eating pattern.	Fair
Almoosawi et al. 2013 [35]	• Occupational Social Class • Smoking Status • Region of Residence • BMI	Hypertensives vs nonhypertensives: proportion of energy consumed in late evening (8.2% vs 7.3%)** Highest vs lowest quintile of energy intake at breakfast: HTN prevalence (OR = 0.70, CI = 0.45 – 1.08)** Highest vs lowest quintile of energy intake in the late evening: HTN incidence (OR = 1.55, CI = 0.93 – 2.61)*; SBP (β = 5.09, CI = 1.25 – 8.93)*; DBP (β = 2.80, CI = 0.27 – 5.32)*;	A greater proportion of energy intake at breakfast was associated with a 30% decreased likelihood of having HTN as compared to a lower proportion of energy intake at breakfast. A greater proportion of energy intake in the late-evening was associated with a 55% increased risk of developing HTN after 10 years and higher SBP and DBP (5.09 and 2.80 mmHg, respectively) compared to a lower proportion of energy intake in the late-evening.	Good
Shim et al. 2021 [54]	• Gender • Age • Shift work • Smoking • Drinking • Walking • BMI • Comorbid status • Medication adherence	Crude Model: Late eaters vs Early eaters: SBP (−0.61)*; DBP (3.20)** Adjusted Model: No significant associations	Simple regression: late eaters had a significantly higher DBP** (3.20 mmHg) than early eaters. Adjusted regression: no significant association between eating timing and blood pressure.	Fair
Dote-Montero et al. 2023 [55]	• Sex • A priori Mediterranean diet pattern • Light physical activity • Midsleep point or sleep duration • BMI	No significant association between caloric midpoint and blood pressure: (β = 0.054 mmHg, CI = not published) Crude Model: Time interval between midsleep point and mean blood pressure: (β = 0.197 mmHg, CI = not published)* Adjusted Model: No significant associations between meal timing and blood pressure.	Caloric midpoint was not significantly associated with blood pressure for the total sample, men, or women. A positive association between the average duration of time between midsleep point and first food intake and mean blood pressure. This association became insignificant after adjustment for covariates.	Fair
Exercise Timing
Brito et al. 2019 [56]	• N/A	Morning Evaluation: SBP for ET vs C (Cohen’s d = −0.63, CI = −1.03 – +0.07)*; SBP for ET vs MT (Cohen’s d = −0.43)* Evening Evaluation: SBP for ET vs C (Cohen’s d = −0.50, CI = −1.17 − +0.19)*; SBP for ET vs MT (Cohen’s d = −0.61, CI = −1.33 – +0.13)* Decreased 24-h DBP and asleep DBP: ET vs C*; ET vs MT*	At the morning evaluation, SBP was reduced following ET and MT. ET reduced SBP to a greater extent than MT (Cohen’s d = −0.43) At the evening evaluation, SBP and DBP decreased following ET only. ET reduced SBP to a greater extent than MT (Cohen’s d = −0.61) 24-h and asleep DBP decreased only after ET. ET reduced 24-h and asleep DBP to a greater extent than MT.	Poor
Brito et al. 2018 [57]	• N/A	Evening Experiments: EE vs EC (nighttime SBP: 116 ± 11 vs 120 ± 10 mmHg)*	Following ME, all ABPM values were similar to C. Following EE, nighttime SBP was 4 mmHg lower than C on average.	Poor
De Brito et al. 2015 [58]	• N/A	Morning Experiments: ME vs MC (SBP: 118 ± 3 vs 124 ± 9 mmHg)*, (DBP: 84 ± 8 vs 87 ± 6 mmHg)*, (MBP: 100 ± 7 vs 96 ± 8 mmHg)* Evening Experiments: ME vs MC (SBP: 117 ± 4 vs 119 ± 9 mmHg)*, (DBP: 83 ± 8 vs 86 ± 7 mmHg)*, (MBP: 95 ± 6 vs 97 ± 7 mmHg)* ME vs EE: SBP Reductions (−7 ± 3 vs −3 ± 4 mmHg)*	ME reduced SBP by 6 mmHg, DBP by 3 mmHg, and MBP by 4 mmHg. EE reduced SBP by 2 mmHg, DBP by 3 mmHg, and MBP by 3 mmHg. Reductions in SBP were around 3 mmHg greater after ME compared to EE.	Fair
Arciero et al. 2022 [59]	• N/A	Women: AM vs PM (SBP: −13 vs −3 mmHg)*; (DBP: −10 vs −5 mmHg) Men: AM vs PM (SBP: −3 vs −15 mmHg)*	Following AM exercise training, women experienced reductions in SBP and DBP (10 mmHg and 5 mmHg, respectively) greater than PM exercise training. Following PM exercise training, men experienced reductions in DBP (13 mmHg) greater than AM exercise training.	Poor
Fairbrother et al. 2014 [60]	• N/A	Morning Exercise vs. Afternoon Exercise: Reduced mean nighttime SBP (107 ± 4.8 vs 114.8 ± 5.2 mmHg)*, Reduced mean nighttime DBP (60.8 ± 0.89 vs 65.6 ± 0.99 mmHg)*, Nocturnal SBP dipping (16.67% ± 0.93% vs 12.3% ± 1.2%)**, Nocturnal DBP dipping (20.9% ± 1.15% vs 15.4% ± 1.6%)* Evening Exercise vs. Afternoon Exercise: Reduced mean nighttime DBP (61.8 ± 0.98 vs 65.6 ± 0.84 mmHg)*, Nocturnal DBP dipping (21.1% ± 1.9% vs 15.4% ± 1.6%)*	In comparison to a 30-minute bout of aerobic exercise performed at 13:00, a 30-minute bout of aerobic exercise at 7:00 reduced nighttime SBP* and DBP* to a greater extent. Additionally, exercise at 7:00 evoked greater nocturnal dipping of SBP and DBP compared to exercise at 13:00. In comparison to a 30-minute bout of aerobic exercise performed at 13:00, a 30-minute bout of aerobic exercise at 19:00 reduced nighttime DBP* to a greater extent. Additionally, exercise at 19:00 evoked greater nocturnal dipping of DBP compared to exercise at 13:00.	Good
Sleep Timing
Abbott et al. 2019 [61]	• Age1,2,3 • Sex1,2,3 • Hispanic/Latino background1,2,3 • Study site1,2,3 • Income2,3 • Acculturation2,3 • Education2,3 • Sleep duration2,3 • Apnea-hypopnea index2,3 • Shift work status3	Model 1: Associations between Sleep midpoint and SBP (β = 0.73 mmHg, CI = 0.30 – 1.17)** and DBP (β = 0.53 mmHg, CI = 0.17 – 0.90)** Model 2: Associations between Sleep midpoint and SBP (β = 0.55 mmHg, CI = 0.12 – 0.99)* and DBP (β = 0.41 mmHg, CI = 0.06 – 0.76)* Model 3: Associations between Sleep midpoint and SBP (β = 0.27 mmHg, CI = −0.21 – 0.74) and DBP (β=0.25 mmHg, CI = −0.13 – 0.63) (Associations not significant) Stratified Analysis: Non-shift workers: Associations between sleep midpoint and SBP (β=0.65 mmHg, CI = 0.01 – 1.28)* and DBP (β=0.52 mmHg, CI = 0.02 – 1.02)* Shift workers: Associations between sleep midpoint and SBP (β=0.34 mmHg, CI = −0.38 – 1.06) and DBP (β=0.24 mmHg, CI = −0.30 – 0.78) (Associations not significant)	For every 1-h delay in sleep midpoint, a 0.73** mmHg and 0.53** mmHg increase in SBP and DBP, respectively was observed after adjustment for age, sex, Hispanic/Latino background, and study site. For every 1-h delay in sleep midpoint, a 0.55* mmHg and 0.41* mmHg increase in SBP and DBP, respectively was observed after adjustment for age, sex, Hispanic/Latino background, study site, income, acculturation, education, sleep duration, and apnea-hypopnea index. After including shift work status as a covariate, the associations lost significance. For every 1-h delay in sleep midpoint, a 0.65* mmHg and 0.52* mmHg increase in SBP and DBP, respectively was observed in non-shift workers after adjustment for age, sex, Hispanic/Latino background, study site. Associations between sleep midpoint and SBP and DBP were not significant.	Fair

^*p≤0.05

^**p≤0.01