Abstract
Background
Circadian rhythm involved with physiology has been reported to affect pharmacokinetics or pharmacodynamics. We hypothesized that circadian variations in physiology disturb anesthesia and eventually affect recovery after anesthesia.
Methods
A retrospective cohort study initially included 107,406 patients (1 June 2016–6 June 2021). Patients were classified into morning or afternoon surgery groups. The primary outcome was daytime variation in PACU (post-anesthesia care unit) recovery time and Steward score. Inverse probability weighting (IPW) approach based on propensity score and univariable/multivariable linear regression were used to estimate this outcome.
Results
Of 28,074 patients, 13,418 (48%) patients underwent morning surgeries, and 14,656 (52%) patients underwent afternoon surgeries. LOWESS curves and IPW illustrated daytime variation in PACU recovery time and Steward score. Before adjustment, compared to morning surgery group, afternoon surgery group had less PACU recovery time (median [interquartile range], 57 [46, 70] vs. 54 [43, 66], p < 0.001) and a higher Steward score (5.62 [5.61, 5.63] vs. 5.66 [5.65, 5.67], p < 0.001). After adjustment, compared to morning surgery group, afternoon surgery group had less PACU recovery time (58 [46, 70] vs. 54 [43, 66], p < 0.001). In multivariable linear regression, morning surgery is statistically associated with an increased PACU recovery time (coefficient, −3.20; 95% confidence interval, −3.55 to −2.86).
Conclusion
Among non-cardiac surgeries, daytime variation might affect recovery after general anesthesia. These findings indicate that the timing of surgery improves recovery after general anesthesia, with afternoon surgery providing protection.
In this retrospective cohort study of 28,074 participants, the afternoon surgery group has a higher Steward score than the morning surgery group.
In multivariable linear regression, morning surgery is statistically associated with an increased PACU recovery time.
Among non-cardiac surgeries, daytime variation affects the recovery after general anesthesia, with afternoon surgery providing protection.
KEY MESSAGES
1. Introduction
The 2017 Nobel Prize in Physiology or Medicine was awarded to three scientists (Jeffrey Hall, Michael Rosbash, and Michael Young) for their contribution to the molecular circadian clock. Circadian rhythmicity is produced endogenously by genetically encoded molecular clocks, displaying cyclic variation in physiology and behaviors to respond to daily recurring environmental changes [Citation1]. At present, basic studies have shown that powerful biological rhythms affect the organism in many aspects, including influencing cell maturation [Citation2], regulating lymphocyte migration [Citation3], interacting with metabolic systems [Citation4–6], and modulating the microbiota-gut-brain axis [Citation7]. This evidence presents the physiological importance of the circadian clock.
A large number of clinical studies have also found that biological rhythms influence human health in many ways. A randomized controlled trial reported that the incidence of adverse cardiovascular events in the afternoon surgery group was lower than that in the morning surgery group [Citation8]. Additionally, circadian misalignment increases metabolic and cardiovascular diseases in humans [Citation9,Citation10]. Therefore, time of day variation significantly impacts clinical outcomes and prognosis, which suggests daytime variation is a vital factor that cannot be ignored in clinical work.
Furthermore, alterations in the circadian rhythm impact renal functions, including glomerular filtration rate, renal plasma flow, diuresis, and tubular transport activities [Citation11], which interfere with the pharmacokinetics and/or pharmacodynamics of various drugs. Simultaneously, circadian regulation of hepatic functions, such as lipid and bile acid metabolism, amino acid and polyamine metabolism, glucose/carbohydrate metabolism, cell processes, and xenobiotic detoxification [Citation12], results in circadian variations in drug metabolism/detoxification and efficacy.
General anesthesia, including intravenous/inhalational anesthesia, through the use of sedatives, analgesics, and muscle relaxants, puts patients in a state of sleep, analgesia, and unconsciousness, which is convenient for surgeries. During surgery, daytime variation can induce circadian variations in liver, renal, and cardiac function [Citation13], potentially leading to fluctuations in the metabolism of anesthetic drugs in the human body.
Hence, based on this evidence, we boldly speculated that daytime variation in physiology could disturb the elimination of anesthetics in the body and ultimately affect the recovery after general anesthesia. This study retrospectively analyzed surgeries in the past six years and explored whether there was daytime variation (morning or afternoon surgery) in the recovery after general anesthesia.
2. Materials and methods
2.1. Study design and data resources
This retrospective cohort study employed surgical data from Lu’an Hospital Affiliated with Anhui Medical University, Lu’an People’s Hospital, which adheres to the applicable STROBE guidelines. Data from patients were obtained from the surgical anesthesia information management system (AIMS).
2.2. Study approval
This retrospective study was granted by the Ethics Committee of Lu’an Hospital Affiliated with Anhui Medical University, Lu’an People’s Hospital (approval number: 2021LLKS003).
2.3. Participants
All patients were identified who underwent surgeries in Lu’an People’s Hospital and required general anesthesia between 1 June 2016 and 6 June 2021.
Inclusion criteria: all non-cardiac surgical patients with general anesthesia (inhalation anesthesia, intravenous anesthesia, and intravenous-inhalation combined anesthesia), who were transferred to PACU (post-anesthesia care unit) after surgery, ASA (American Society of Anesthesiologists) grade: I–IV.
Exclusion criteria: transferred to ICU (intensive care unit) or the ward after surgery; whereabouts after surgery not recorded; missing demographic information; no extubation; missing information about comorbidities surgery, anesthesia, and recovery in PACU; delayed recovery after anesthesia (Patients with PACU recover time > 90 min were defined as delayed recovery after anesthesia); PACU recovery time < 30 min (Patients transferred to PACU for monitoring and observation for at least 30 min); general anesthesia effect rating (evaluating the effect and quality of general anesthesia) = III (the worst); the surgical time span including morning and afternoon.
Patients accepting multiple surgeries within different periods were also included in the initial cohort. A diagram of surgeries included and cohort build was shown in . ICD-10 (International Classification of Diseases, Tenth Revision) definitions for comorbidities were in Supplemental Table 1.
Figure 1. Diagram of surgeries included and cohort build.
Table 1. Baseline characteristics of patients accepting morning or afternoon surgery.
2.4. Exposure of interest
The exposure of interest in the present study was the surgery performed in the afternoon. Patients were regarded as the exposure to surgery in the afternoon according to the anesthetic record sheets in AIMS. We determined which time period of surgery (morning or afternoon) subjects were exposed to according to the surgery start time and end time. Moring surgery was defined as the surgical time ranging from 7:30 to 12:00 a.m. Then, afternoon surgery was defined as the surgical time ranging from 12:00 a.m. to 19:00 p.m.
2.5. Outcome
The main outcome was the daytime variation (morning or afternoon) in PACU recovery time and Steward score when patients left PACU. Steward scoring system was used to assess the quality and level of recovery after general anesthesia (0, 1, or 2 points given to each of the three categories: consciousness, airway control, and movement) [Citation14,Citation15], which was extensively applied in clinical practice [Citation16]. A high Steward score means better recovery from anesthesia.
2.6. Statistical analysis
Continuous variables of baseline demographic characteristics were recorded as the mean (standard deviation [SD]) or median (interquartile range [IQR]). Categorical variables of demographic data were summarized with frequency and percentage. Wilcoxon rank sum test and t-test were used to examine the difference in time and quality of post-anesthetic recovery between groups. The LOWESS (locally weighted scatterplot smoothing) is a curve fitting method that combines multiple regression models in a k-nearest-neighbor-based meta-model used in time series data analysis [Citation17]. The LOWESS curve fitting was performed by Stata 16.0 to predict the overall trend.
Patients in the non-exposed (morning surgery) and exposed (afternoon surgery) groups were systematically different owing to confounders. Surgeries implemented in the morning or afternoon were not randomly assigned. Consequently, patient baseline characteristics were imbalanced between the morning and afternoon groups.
Firstly, we employed an inverse probability weighting (IPW) approach based on propensity score to control confounding variables [Citation18–20]. The propensity score matching was evaluated by a logistic regression model with receipt of the surgery in the afternoon as the dependent variable. Independent variables in the multivariable regression model included age, sex, duration of surgery, year of surgery, surgery grading, elective/emergent surgery, type of surgery, using antibiotics during surgery, duration of anesthesia, ASA grade, Steward score when entering PACU, type of intubation, general anesthesia effect rating, post-operative analgesia, and comorbidities. After calculating propensity scores, surgeries in the afternoon group were used to match in a 1:1 ratio to those in the morning group, which constructed a matched sample. Propensity score matching and IPW were conducted by using psmatch2 and teffects ipw package in Stata 16.0, respectively. Patients were weighted by IPW with propensity score to estimate ATE (average treatment effect in population), which reflects their actual average treatment. The balance of covariates pre-and post-weighting was assessed via estimating standardized differences (Std Diff), and the Std Diff of ≥10% is a meaningful imbalance [Citation18].
Secondly, univariable and multivariable linear regression was also utilized to control covariates, with the estimation of coefficients (β) and 95% confidence interval (CI). The independence among variables was evaluated via the Durbin–Watson test. A p-value < 0.05 was regarded as statistically significant. The protocol of a retrospective cohort study and statistical analysis plan was in Supplemental Files.
3. Results
3.1. Patients
Between 1 June 2016 and 6 June 2021, 107,406 patients were initially enrolled in this study. Of 107,406 patients, 28,074 patients were finally included in the data analysis. There were 13,418 patients in the morning surgery group and 14,565 patients in the afternoon surgery group (). After propensity score matching, baseline covariates were well-balanced in morning and afternoon surgery groups (), with an Std Diff of no more than 10% ().
Figure 2. (A) Standardized difference of all covariates before or after propensity score matching. (B) Graphical representation of results after propensity score matching. The covariate with Std Diff of ≥10% is a meaningful imbalance. ASA: American Society of Anesthesiologists; PACU: post-anesthesia care unit.
3.2. LOWESS curve
We used the start time of surgery as a time variable to draw a fitting curve, as shown in . The LOWESS curves suggest that patients in the afternoon surgery group displayed less PACU recovery time and higher Steward score when compared to the morning surgery group. Both LOWESS curves illustrated daytime variation in PACU recovery time and Steward score.
Figure 3. Daytime variation in PACU recovery time and Steward score. (A) Shows the daytime variation in PACU recovery time of patients undergoing non-cardiac surgeries with general anesthesia. (B) Shows the daytime variation in Steward score of patients undergoing non-cardiac surgeries with general anesthesia. The start time of surgery was used as a time variable to draw LOWESS curves.
3.3. Outcomes
The comparison data between groups before and after adjustment was presented in . Before adjustment, when compared to the morning surgery group, patients in the afternoon surgery group had less PACU recovery time (57 [46, 70] vs. 54 [43, 66], p < 0.001) and a higher Steward score (5.62 [5.61, 5.63] vs. 5.66 [5.65, 5.67], p < 0.001).
Table 2. Time and quality of post-anesthetic recovery in patients accepting morning or afternoon surgery.
After adjustment, when compared to the morning surgery group, patients in the afternoon surgery group had less PACU recovery time (58 [46, 70] vs. 54 [43, 66], p < 0.001). Additionally, there is no difference in Steward score between groups (5.65 [5.64, 5.66] vs. 5.66 [5.65, 5.67], p = 0.07).
3.4. ATE
After IPW based on the propensity score, ATE of PACU recovery time in the afternoon surgery group was −3.45 (95% CI, −3.80 to −3.12), compared with the morning surgery group. It suggests that patients undergoing surgery in the morning were delayed by an average of 3.45 min in PACU recovery time compared with patients undergoing surgery in the afternoon. Furthermore, ATE of Steward score in the afternoon surgery group was 0.0094 (95% CI, 0.0050–0.014), compared with the morning surgery group. It suggests, on average, that patients undergoing afternoon surgery were 0.0094 points higher than patients undergoing morning surgery ().
3.5. Univariable and multivariable linear regression
To further estimate daytime variation in the recovery after general anesthesia, univariable and multivariable linear regression was employed. Above all, univariable linear regression included all baseline variables in . The variable with p < 0.05 in the model was further taken in multivariable linear regression. In multivariable linear regression, the morning surgery is statistically associated with an increased PACU recovery time (β, −3.20; 95% CI, −3.55 to −2.86). After controlling other variables, patients accepting morning surgery were delayed 3.20 min in PACU recovery time compared with patients accepting afternoon surgery ().
Figure 4. Univariate linear regression to examine the factors correlated with PACU recovery time. ASA: American Society of Anesthesiologists; PACU: post-anesthesia care unit.
Table 3. Multivariable linear regression to examine the factors correlated with PACU recovery time.
3.6. Subgroup analyses
Subgroup analyses of PACU recovery time still found group differences when patients were stratified by sex and elective or emergent surgeries (Supplemental Table 2).
4. Discussion
In this large retrospective cohort of patients undergoing surgery, a clinically significant and statistically important association between daytime variation of surgery and the recovery after general anesthesia was reported. ATE and LOWESS curves showed that patients accepting morning surgery suffered delayed PACU recovery time and decreased Steward score compared to afternoon surgery. Moreover, univariable and multivariable linear regression also demonstrated the same results consistent with the IPW strategy.
From the evidence of basic and clinical research, the influence of circadian rhythm on biological processes is subsistent. The previous study used clinical trials combined with animal experiments that reported that time-of-the-day variation was associated with perioperative myocardial injury, which is transcriptionally orchestrated by the circadian clock [Citation8]. Meanwhile, it further confirmed that afternoon surgery could provide a perioperative myocardial protective effect and better patient outcomes, which seem to be in agreement with our findings. However, a sizeable propensity-matched cohort study that included 7148 patients undergoing aortic valve replacement found no daytime-dependent effect on perioperative myocardial injuries [Citation21]. A similar retrospective study that included 2720 patients showed that morning or afternoon surgery was not associated with statistical differences in risk-adjusted morbidity and mortality [Citation22].
Likewise, another retrospective cohort study included 9734 patients who had aortic valve, mitral valve, and/or coronary artery bypass graft surgery and also described that there was no difference in mortality and cardiac complications between morning and afternoon cardiac surgeries [Citation23]. The reasons for inconsistencies between other studies and our results were as follows. Firstly, due to the small number of patients undergoing cardiac surgery, our study only included non-cardiac surgeries. Secondly, our topics only focused on the clinically protective effect of daytime variation on recovery after anesthesia instead of mortality or cardiac complications. Finally, these recent clinical trials and retrospective studies presented disputed conclusions about daytime variation in the cardioprotective effect of the afternoon compared with morning surgery. Considering the controversy about daytime variation in the cardioprotective effect, we excluded cardiac surgeries, reducing their potential impacts and capturing more accuracy in our results.
Furthermore, Szolnoki et al. in their retrospective study (data from 2340 procedures in children), have also illustrated that daytime variation significantly affects PACU recovery time in children having brain imaging under general anesthesia [Citation24], which was consistent with this present study. In detail, among children older than 3 or 5 years, the PACU recovery time was increased with an observed maximum increase of 18 or 19 min (<9 a.m. vs. >6 p.m.) [Citation24]. Whereas, in our study, the PACU recovery time was decreased with an observed average decrease of 3.5 min. The reasons for the difference in results between the two studies were summarized as follows. Firstly, our study included surgical populations of all ages, as well as different types of surgeries. By contrast, their study only collected data on children under 18 years old who underwent general anesthesia for brain magnetic resonance imaging (MRI), which highly differs from our study procedures. Secondly, Szolnoki et al. did not apply some statistical methods to control covariates or confounding variables in data analysis, which might influence the main outcome. In contrast, we employed propensity score matching and univariable/multivariable linear regression. Thirdly, the sample size in that study, especially in subgroup analyses, was relatively small. At last, the grouping in the two studies was still dissimilar, which also resulted in various conclusions.
The present study confirmed the association between circadian rhythms and recovery after general anesthesia. We suspected that the underlying mechanism behind this association was involved in an interactive dialogue between general anesthesia and cyclic variation in physiology and behavior regulated by circadian rhythms. On the one hand, anesthetics, such as propofol, can affect the circadian rhythm of plasma melatonin, produce strong effects on neurotransmitter systems linked with circadian control [Citation25], and impair synchronization of the circadian rest-activity rhythm [Citation26,Citation27]. Moreover, from transcriptomes, Zhang et al. used RNA-seq and DNA arrays in the animal study. They discovered that anesthesia-related drugs, such as diazepam, halothane, isoflurane, ketamine, lidocaine, midazolam, phenobarbital, and propofol, target the product of a circadian gene [Citation28], suggesting their potential impact on circadian rhythms (Supplemental Table 3).
On the other hand, circadian variations in the physiological processes and variables are related to drugs’ absorption, distribution, metabolism, and excretion [Citation29–31]. In addition, circadian rhythm also disturbs the efficacy and functions of drugs in vivo by affecting the physiology, absorption, metabolism, transport, and secretion of the liver and kidney [Citation11,Citation32]. Eventually, anesthetics interfere with the body’s circadian clock. Conversely, the circadian clock also feeds back and regulates physiological functions and behavioral alteration. With such mutual crosstalk, a balance is finally reached at the molecular, physiological, and behavioral levels.
The present study has several significant strengths. At first, to the best of our knowledge, this retrospective cohort study was the first large sample of patients and incorporated a wide variety of surgeries, focusing on exploring circadian variations in the recovery after general anesthesia. From a statistical perspective, both propensity score matching and univariable/multivariable linear regression were utilized to control the covariates, which helped capture more accuracy and efficacy in data analyses. Furthermore, to avoid imbalance and non-randomization of observation data in the study, ATE was harnessed as a treatment-effects estimator to examine the main outcome. Finally, we eliminated a small number of patients with missing information, which prevented their influence on results. All cases used in the study were stored in AIMS and strictly reviewed by clinicians, which guaranteed the correctness of the data.
This study has underlying limitations needed to be addressed. Above all, it was not likely to determine the precise time boundary in a grouping setting. According to the previously published study and the hospital’s working procedures [Citation8], we set up two groups, which might limit the capability to examine the maximum difference in PACU recovery time between groups. In addition, this retrospective cohort analysis was an observational study, which may be subject to unmeasured confounders. A large number of participants and the exclusion of patients with incomplete data can reduce such an impact. Other possible limitations were several influencing factors that were quite difficult to control. For example, the characteristics of drug metabolism in the patient and the evaluation of organs’ functions related to anesthetic effects were not documented. Regardless, we sought to strictly validate our results and decrease the influence of other factors. Last, data in this study confined to a single hospital has a possible influence on our generalization. Nevertheless, our findings still discovered the significance of circadian rhythm in clinical practice. Ultimately, prospective cohort studies still need to be designed and implemented to support the findings of this study.
5. Conclusions
Our study has reported that the PACU recovery time was decreased with an observed average decrease of 3.45 min in the afternoon surgery group. The clinical significance of the 3.45 min delay in recovery after anesthesia for a specific patient appears to be less important. However, in a PACU filled with many patients recovering from anesthesia, the cumulative effect of this delay in recovery after anesthesia can reach a long time. Thus, the cumulative effect of this delay has clinical relevance. Ultimately, we provide physicians with an essential piece of evidence that the circadian clock affects recovery from anesthesia. Among patients undergoing non-cardiac surgeries with general anesthesia, daytime variation might affect the speed and quality of the recovery after general anesthesia. These findings indicate that the timing of non-cardiac surgery might improve the recovery after general anesthesia, with afternoon surgery offering a protective effect.
Author contributions
Feng Xu, Qingtong Zhang, and Xiangdong Chen: drafting of the paper. All authors: conception and design, analysis and interpretation of the data, revising it critically for intellectual content, and the final approval of the version to be published.
Supplemental Material
Download MS Word (683.5 KB)Data availability statement
The data that support the findings of this study are available upon reasonable request through contact with the corresponding author (Prof. Chen).
Disclosure statement
No potential conflict of interest was reported by the author(s).
Additional information
Funding
References
- Patke A, Young MW, Axelrod S. Molecular mechanisms and physiological importance of circadian rhythms. Nat Rev Mol Cell Biol. 2020;21(2):67–84.
- Du Pré BC, Van Veen TA, Young ME, et al. Circadian rhythms in cell maturation. Physiology. 2014;29(1):72–83.
- Druzd D, Matveeva O, Ince L, et al. Lymphocyte circadian clocks control lymph node trafficking and adaptive immune responses. Immunity. 2017;46(1):120–132.
- Huang W, Ramsey KM, Marcheva B, et al. Circadian rhythms, sleep, and metabolism. J Clin Invest. 2011;121(6):2133–2141.
- Chaix A, Lin T, Le HD, et al. Time-restricted feeding prevents obesity and metabolic syndrome in mice lacking a circadian clock. Cell Metab. 2019;29(2):303–319.e4.
- Sinturel F, Gerber A, Mauvoisin D, et al. Diurnal oscillations in liver mass and cell size accompany ribosome assembly cycles. Cell. 2017;169(4):651–663.e14.
- Teichman EM, O’Riordan KJ, Gahan CGM, et al. When rhythms meet the blues: circadian interactions with the microbiota-gut-brain axis. Cell Metab. 2020;31(3):448–471.
- Montaigne D, Marechal X, Modine T, et al. Daytime variation of perioperative myocardial injury in cardiac surgery and its prevention by Rev-Erbα antagonism: a single-centre propensity-matched cohort study and a randomised study. Lancet. 2018;391(10115):59–69.
- Morris CJ, Purvis TE, Hu K, et al. Circadian misalignment increases cardiovascular disease risk factors in humans. Proc Natl Acad Sci USA. 2016;113(10):E1402–E1411.
- Scheer FA, Hilton MF, Mantzoros CS, et al. Adverse metabolic and cardiovascular consequences of circadian misalignment. Proc Natl Acad Sci USA. 2009;106(11):4453–4458.
- Firsov D, Bonny O. Circadian rhythms and the kidney. Nat Rev Nephrol. 2018;14(10):626–635.
- Mukherji A, Bailey SM, Staels B, et al. The circadian clock and liver function in health and disease. J Hepatol. 2019;71(1):200–211.
- Zhang L, Jain MK. Circadian regulation of cardiac metabolism. J Clin Invest. 2021;131(15):e148276.
- Steward DJ. A simplified scoring system for the post-operative recovery room. Can Anaesth Soc J. 1975;22(1):111–113.
- Aun CS, Short TG, O’Meara ME, et al. Recovery after propofol infusion anaesthesia in children: comparison with propofol, thiopentone or halothane induction followed by halothane maintenance. Br J Anaesth. 1994;72(5):554–558.
- Bhardwaj N, Yaddanapudi S. A randomized trial of propofol consumption and recovery profile with BIS-guided anesthesia compared to standard practice in children. Paediatr Anaesth. 2010;20(2):160–167.
- Cleveland WS. Robust locally weighted regression and smoothing scatterplots. J Am Stat Assoc. 1979;74(368):829–836.
- Haukoos JS, Lewis RJ. The propensity score. JAMA. 2015;314(15):1637–1638.
- Austin PC. The performance of different propensity-score methods for estimating differences in proportions (risk differences or absolute risk reductions) in observational studies. Stat Med. 2010;29(20):2137–2148.
- Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011;46(3):399–424.
- Kenney PS, Nielsen PH, Modrau IS. Daytime-dependent cardioprotection in cardiac surgery: a large propensity-matched cohort study. Ann Thorac Surg. 2020;110(5):1629–1636.
- Götte J, Zittermann A, Deutsch MA, et al. Daytime variation in aortic valve surgery and clinical outcome: a propensity score-matched analysis. Ann Thorac Surg. 2020;110(2):558–566.
- Hijazi RM, Sessler DI, Liang C, et al. Association between in-hospital mortality and low cardiac output syndrome with morning versus afternoon cardiac surgery. Anesthesiology. 2021;134(4):552–561.
- Szolnoki J, Polaner DM, Eckle T. Diurnal variations in recovery times after general anaesthesia in children. Br J Anaesth. 2018;121(4):776–786.
- Poulsen RC, Warman GR, Sleigh J, et al. How does general anaesthesia affect the circadian clock? Sleep Med Rev. 2018;37:35–44.
- Dispersyn G, Touitou Y, Coste O, et al. Desynchronization of daily rest-activity rhythm in the days following light propofol anesthesia for colonoscopy. Clin Pharmacol Ther. 2009;85(1):51–55.
- Dispersyn G, Pain L, Touitou Y. Propofol anesthesia significantly alters plasma blood levels of melatonin in rats. Anesthesiology. 2010;112(2):333–337.
- Zhang R, Lahens NF, Ballance HI, et al. A circadian gene expression atlas in mammals: implications for biology and medicine. Proc Natl Acad Sci USA. 2014;111(45): 16219–16224.
- Labrecque G, Bélanger PM. Biological rhythms in the absorption, distribution, metabolism and excretion of drugs. Pharmacol Ther. 1991;52(1):95–107.
- Dong D, Yang D, Lin L, et al. Circadian rhythm in pharmacokinetics and its relevance to chronotherapy. Biochem Pharmacol. 2020;178:114045.
- Sukumaran S, Almon RR, DuBois DC, et al. Circadian rhythms in gene expression: relationship to physiology, disease, drug disposition and drug action. Adv Drug Deliv Rev. 2010;62(9–10):904–917.
- Tahara Y, Shibata S. Circadian rhythms of liver physiology and disease: experimental and clinical evidence. Nat Rev Gastroenterol Hepatol. 2016;13(4):217–226.