Attention-deficit hyperactivity symptoms and risk of alcohol use relapse

Abstract

Aim

Alcohol use disorder (AUD) is frequently accompanied by comorbid attention-deficit hyperactivity disorder (ADHD). Comorbid ADHD has been reported to increase the severity of AUD. We investigated whether ADHD severity also influences AUD relapse risk at baseline and after inpatient treatment.

Methods

In this study, 187 AUD patients admitted to Narimasu Kosei Hospital from October 2019 to March 2021 were included in the analysis. According to the Adult ADHD Self-Report Scale (ASRS), participants were divided into two groups: ASRS+ with ADHD characteristics (n = 43) and ASRS− with low/no ADHD characteristics (n = 144). Groups were compared for AUD relapse risk at the start of treatment (baseline) and before hospital discharge using the multidimensional Alcohol Relapse Risk Scale (ARRS). The change in relapse risk during hospitalization was also compared by assessment of the interaction between groups (ASRS+ vs. ASRS−) and time (at discharge vs. baseline).

Results

The total ARRS score and dimension subscores for stimulus-induced vulnerability and emotionality problems were significantly higher in the ASRS+ group at baseline and before discharge compared to the ASRS− group. There was a significant group × time interaction indicating less improvement of stimulus-induced vulnerability during inpatient treatment among the ASRS+ group compared to the ASRS− group.

Conclusions

Our findings suggest that AUD patients with ADHD characteristics have a higher risk of relapse both at baseline and after inpatient treatment. Stimulus-induced vulnerability to relapse is less likely to improve with treatment in patients with ADHD characteristics.

INTRODUCTION

Alcohol use disorder (AUD) is characterized by an inability to stop drinking alcohol in excess despite deleterious physical, psychological, and social consequences.^1–3 Further, AUD is difficult to treat due to a constant craving for alcohol, impaired control of drinking, withdrawal symptoms, tolerance, and a drinking-centered lifestyle.^1,4 A substantial portion of AUD patients (6.6%–21.3%) also exhibit symptoms of attention-deficit hyperactivity disorder (ADHD),⁵ a neurodevelopmental disease characterized by inattention, hyperactivity, and impulsivity most often diagnosed in childhood but persisting into adulthood in up to two-thirds of cases.⁶ It is speculated that the disabling characteristics of ADHD in adulthood may contribute to the onset or exacerbation of other psychiatric comorbidities, including substance use disorders.^7,8

Numerous studies have reported that patients with comorbid ADHD and substance use disorders exhibit greater substance use and more severe psychiatric symptoms,^9–11 including more intense cravings that could increase the risk of relapse.¹² A study comparing AUD relapse risk between subjects with and without ADHD traits showed a higher risk of relapse in the ADHD group.¹³ However, the sample size was small and did not investigate whether changes in the relapse risk during the treatment differed between the groups.

Because ADHD is prone to relapse in the long term even after remission,¹⁴ the following should be noted in adulthood: Comorbid ADHD is associated with high impulsivity and risk-taking behavior.^15,16 This situation is expected to increase the risk of relapse drinking in AUD patients. Impulsive responses to stimuli, immediate rewards, and high expectations and motivation for favorite things^17,18 may make individuals more likely to choose drinking for current gratification than to tolerate drinking for future benefits. Executive dysfunction,¹⁹ including a poor outlook on negative consequences, can make it difficult to imagine the various risks associated with relapsing into drinking. There is weak emotional control,^20–23 and patients may be prone to relapse to distract from negative emotions such as anxiety.

Thus, we hypothesized that AUD patients with strong ADHD characteristics would be at higher risk of AUD relapse and that this risk would be less likely to improve after inpatient treatment. To test this hypothesis, we compared relapse risk at baseline and after inpatient treatment between patients with and without clinically significant ADHD characteristics using a well-validated relapse risk assessment tool.

METHODS Procedure and participants

This study was conducted at Tokyo Alcohol Medical Center, Narimasu Kosei Hospital, from October 2019 to March 2021. The center has 56 beds specializing in the treatment of AUD and offers an alcohol rehabilitation treatment program. The program includes motivational interviewing, lectures on AUD, group meetings, occupational therapy, and participation in self-help groups (Table 1).

TABLE 1 Alcohol rehabilitation treatment program

Program component	Content
Motivational interviewing	Patients increase their motivation to stay sober by discussing their ambivalence with staff
Lectures on AUD	Staff educate patients about the physical, mental, and social effects caused by alcohol
Group meetings	Patients share their problems with each other; discuss drinking triggers, coping strategies, etc., and build supportive relationships
Occupational therapy	Patients experience the joys of sobriety through activities such as exercising, practicing relaxation techniques, listening to music, drawing pictures, and growing plants
Participation in self-help groups	Patients meet others who already have a long history of sobriety to help them visualize their own recovery and give them hope

Abbreviation: AUD, alcohol use disorder.

Inclusion criteria were as follows: (1) patients meeting the criteria for AUD of the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5),²⁴ as assessed by expert clinicians based on a comprehensive psychiatric interview; (2) patients admitted to the center for alcohol rehabilitation; (3) patients at least 20 years of age; and (4) patients who voluntarily agreed to participate in writing after receiving an explanation of the study and fully understanding it during the first 2 weeks after admission. Exclusion criteria were the following: (1) patients with severe restlessness, (2) patients with a severe physical illness, (3) patients with alcohol withdrawal delirium, (4) patients with dementia, and (5) patients with intellectual disability. Thirty-four individuals met the exclusion criteria. In addition, 4 patients were discharged for personal reasons before the study was explained to them. The study was explained to 21 patients, but they refused to participate. A total of 200 AUD patients participated 13 were discharged voluntarily before completion of inpatient treatment; thus, data from 187 patients were included in the analysis. A flowchart of participant recruitment is shown in Figure 1.

The study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Tokyo Medical and Dental University Faculty of Medicine (Approval number M2019-161) and Narimasu Kosei Hospital (Approval number 204).

Measures Adult ADHD Self-Report scale (ASRS)

The Adult ADHD Self-Report Scale version 1.1 (ASRS-v1.1) is a self-administered symptom checklist for ADHD in adulthood developed by the World Health Organization.²⁵ The ASRS measures the severity of various ADHD symptoms in adults based on 6 items in Part A and 12 items in Part B. Each item is scored on a five-point Likert scale ranging from 0 (never) to 4 (very often), with a higher total score indicating more severe ADHD symptoms. In the present study, ADHD symptoms were assessed with Part A only as it has been shown to better reflect ADHD in adulthood.^25,26 The Japanese version of the ASRS has been shown to have sensitivity, specificity, positive predictive value, and negative predictive value of 0.69 or higher.²⁷ In the current study, Cronbach's alpha coefficient for ASRS Part A was 0.82.

To exclude the effects of alcohol withdrawal symptoms on ADHD, participants completed the ASRS Part A between days 15 and 28 after hospital admission (more than 2 weeks after their last drink). Following the study by Kessler et al.,²⁵ one point was assigned to the responses “sometimes,” “often,” and “very often” for items 1–3, and to the responses “often” and “very often” for items 4–6. A total score of 4 or more was defined as ASRS+ and a lower score as ASRS−. The ASRS Part A score was not used in subsequent analyses but was summarized in Table 2 to show the characteristics of the subjects.

TABLE 2 Sociodemographic characteristics and clinical characteristics of ASRS+ and ASRS− patients

[Table omitted. See PDF]

Note: Student's t-test was used to compare normally distributed continuous variables. Mann–Whitney U test was used to compare non-normally distributed continuous variables. Chi-squared test was used to compare categorical variables. *p < 0.05.

Abbreviations: ADHD, attention-deficit hyperactivity disorder; ASRS, Adult ADHD Self-Report Scale; AUDIT, Alcohol Use Disorders Identification Test; IQR, interquartile range; SD, standard deviation.

In addition to the ASRS, some subjects were diagnosed with ADHD during inpatient treatment by a physician based on the DSM-5 diagnostic criteria and were prescribed ADHD medication. Thus, the number of subjects clinically diagnosed with ADHD, the number of DSM-5 symptom items per subject, the number of subjects using ADHD medication, and the duration of medication use in the ASRS+ group were also described.

Alcohol relapse risk scale (ARRS)

Risk of relapse was measured by the ARRS, a self-administered questionnaire developed by the Tokyo Metropolitan Institute of Medical Science to predict the risk of AUD relapse based on multiple dimensions of alcohol abuse.²⁸ The ARRS had demonstrated good psychometric properties in Japanese AUD patients. The questionnaire consists of 32 items and assesses the risk of relapse based on the following five-dimension subscores: (1) stimulus-induced vulnerability (nine items), (2) emotionality problems (eight items), (3) compulsivity for alcohol (three items), (4) lack of negative expectancy for alcohol (four items), and (5) positive expectancy for alcohol (three items). Since the other five items concern the strength of insight into illness, we only used the above five-dimension subscores in this study. Each item was rated on a 3-point Likert scale as follows: “strongly disagree and disagree” (1 point), “neither agree nor disagree” (2 points), and “strongly agree and agree” (3 points). According to the ARRS manual,²⁹ each dimension subscale score was evaluated by calculating the mean score on dimension-related items, and the mean score on all five subscales was used as the ARRS total score with higher score indicating a higher risk of relapse.

To investigate changes in relapse risk during hospitalization, participants completed the ARRS twice, once 15–28 days after admission and again from the day of discharge to 2 weeks prior to discharge. In this study, Cronbach's alpha coefficients for individual dimensions ranged from 0.64 (compulsivity for alcohol) to 0.86 (stimulus-induced vulnerability) at baseline, and from 0.72 (compulsivity for alcohol) to 0.89 (stimulus-induced vulnerability) before discharge, indicating that subscales had fair to good internal consistency.

Alcohol use disorders identification test (AUDIT)

The AUDIT is a screening test for alcohol problems. This self-administered questionnaire scale was developed by the World Health Organization^30,31 and consists of 10 items scored between 0 and 4 according to the responses to each item, with higher total scores indicating a greater likelihood of alcohol consumption and alcohol-related problems. Although the AUDIT is a screening tool rather than a direct indicator of AUD severity, we collected AUDIT scores at the time of admission as a reference for background information on each subject. Cronbach's alpha coefficient was 0.71.

Statistical analyses

We compared the sociodemographic and clinical characteristics, AUDIT scores, ASRS Part A scores, frequency of self-help group participation during hospitalization, and duration of hospital stay between the ASRS+ and ASRS− groups. Normally distributed continuous variables were compared by Student's t-test, non-normally distributed continuous variables by Mann–Whitney U test (including ARRS total and subscale scores), and categorical variables by chi-squared test.

To examine whether the change in ARRS from baseline to post-treatment also differed between ASRS+ and ASRS− groups, we conducted an analysis of the group × time interaction using the linear model “$\mathit{\text{ARRS}}\left(2\mathit{nd}-1\mathit{st}\right)=\alpha +{\beta }_1\times \mathit{\text{ARRS}}1\mathit{st}+{\beta }_2\times \mathit{\text{ASRS}}+{\beta }_3\times \mathit{\text{ARRS}}1\mathit{st}\times \mathit{\text{ASRS}}$.” If ${\beta }_3$ is significant, the second test score minus the first test score divided by the first test score (or relative change) will differ significantly between groups.

An additional analysis of interactions was also conducted within the ASRS+ group to determine whether the change in ARRS score differed between patients receiving ADHD medications and those receiving no ADHD drugs using the linear model “$\begin{array}{c}\mathit{\text{ARRS}}\left(2\mathit{nd}-1\mathit{st}\right)=\alpha +{\beta }_1\times \mathit{\text{ARRS}}1\mathit{st}+{\beta }_2\times \mathit{\text{ADHD medication}}+{\beta }_3\times \mathit{\text{ARRS}}1\mathit{st}\hfill \end{array}$$\begin{array}{c}\times \mathit{\text{ADHD medication}}\hfill \end{array}$.” Again, if ${\beta }_3$ is significant, the second ARRS score minus the first score divided by the first ARRS score (or relative change) will differ significantly between patients taking ADHD medications and unmedicated patients.

All statistical analyses were performed using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan),³² and R (The R Foundation for Statistical Computing, Vienna, Austria, version 4.0.2). A p ≤ 0.05 was considered significant for all tests.

RESULTS Sociodemographic description

The 187 AUD inpatients participating in this study were first stratified into two groups according to the severity of ADHD characteristics as assessed by the ASRS, an ASRS+ group with greater ADHD symptoms and an ASRS− group with milder or no ADHD symptoms. Sociodemographic and clinical characteristics of each group are summarized in Table 2. The ASRS+ group was significantly younger than the ASRS− group and had significantly higher AUDIT scores. There were no significant differences between the groups regarding the gender ratio, presence of a cohabitant, education duration, occupational status, frequency of participation in self-help groups during hospitalization, and duration of hospital stay.

Comparisons of the ARRS at baseline and before discharge

At baseline (BL), total score on the ARRS as well as dimension scores for stimulus-induced vulnerability and emotionality problems were significantly higher in the ASRS+ group compared to the ASRS− group. In contrast, the ASRS+ group scored significantly lower on the lack of negative expectancy for alcohol dimension compared to the ASRS− group. There were no significant differences in compulsivity for alcohol and positive expectancy for alcohol dimension scores between groups. Before discharge (BD), the ASRS+ group exhibited a similar scoring pattern as at baseline, with a significantly higher total ARRS score, significantly higher stimulus-induced vulnerability, emotionality problems, and compulsivity for alcohol dimension scores, significantly lower lack of negative expectancy for alcohol dimension score, and a similar positive expectancy for alcohol dimension score compared to the ASRS− group (Figure 2, Table 3).

TABLE 3 ARRS in ASRS+ and ASRS− patients at baseline and before discharge

	ASRS+ (n = 43)		ASRS− (n = 144)		ASRS+ vs. ASRS−
	Median [IQR]		Median [IQR]		p-Value
	Baseline	Before discharge	Baseline	Before discharge	Baseline	Before discharge
ARRS	1.78 [1.63, 1.96]	1.74 [1.52, 1.94]	1.63 [1.44, 1.81]	1.52 [1.37, 1.64]	<0.001*	<0.001*
SV	1.78 [1.44, 2.00]	1.67 [1.22, 1.94]	1.50 [1.22, 1.89]	1.33 [1.11, 1.78]	0.005*	0.006*
EP	2.13 [1.88, 2.38]	2.00 [1.81, 2.25]	1.63 [1.38, 2.00]	1.50 [1.25, 1.88]	<0.001*	<0.001*
CA	1.00 [1.00, 1.33]	1.00 [1.00, 1.33]	1.00 [1.00, 1.33]	1.00 [1.00, 1.00]	0.195	0.042*
NE	1.75 [1.50, 2.00]	1.75 [1.50, 2.12]	2.22 [1.75, 2.50]	2.00 [1.50, 2.50]	<0.001*	0.016*
PE	1.67 [1.33, 2.00]	1.67 [1.00, 2.33]	1.67 [1.00, 1.75]	1.33 [1.00, 1.67]	0.069	0.134

Note: Mann–Whitney U test was used to compare the ASRS+ group with the ASRS− group. *p < 0.05.

Abbreviations: ADHD, attention-deficit hyperactivity disorder; ARRS, Alcohol Relapse Risk Scale; ASRS, Adult ADHD Self-Report Scale; CA, Compulsivity for alcohol; EP, Emotionality problems; IQR, interquartile range; NE, Lack of negative expectancy for alcohol; PE, Positive expectancy for alcohol; SV, Stimulus-induced vulnerability.

Comparisons of changes in the ARRS

Analysis of the interaction between groups (ASRS+ vs. ASRS−) and time (BL vs. BD) revealed differences in changes in the risk of relapse (Table 4). Specifically, the ASRS+ group reported significantly less improvement in stimulus-induced vulnerability dimension score during inpatient treatment than the ASRS− group, while improvements in other dimensions and the total ARRS score were similar.

TABLE 4 Analysis of interactions to determine if the rate of change in ARRS differs between the ASRS+ and ASRS− groups

Interaction	Regression coefficient, estimate	Std. error	t-Value	p-Value
ARRS 1st × ASRS	0.254	0.138	1.843	0.067
SV 1st × ASRS	0.301	0.148	2.038	0.043*
EP 1st × ASRS	0.109	0.159	0.689	0.492
CA 1st × ASRS	−0.035	0.113	−0.311	0.756
NE 1st × ASRS	−0.070	0.164	−0.424	0.672
PE 1st × ASRS	0.194	0.144	1.353	0.178

Note: We analyzed the interactions with the linear model “$\mathit{\text{ARRS}}\left(2\mathit{nd}-1\mathit{st}\right)=\alpha +{\beta }_1\times \mathit{\text{ARRS}}1\mathit{st}+{\beta }_2\times \mathit{\text{ASRS}}+{\beta }_3\times \mathit{\text{ARRS}}1\mathit{st}\times \mathit{\text{ASRS}}$.” *p < 0.05.

Abbreviations: ADHD, attention-deficit hyperactivity disorder; ARRS, Alcohol Relapse Risk Scale; ASRS, Adult ADHD Self-Report Scale; CA, Compulsivity for alcohol; EP, Emotionality problems; NE, Lack of negative expectancy for alcohol; PE, Positive expectancy for alcohol; Std. Error, standard error; SV, Stimulus-induced vulnerability.

ASRS+ participants with confirmed ADHD diagnoses

In the ASRS+ group, 11 subjects met the DSM-5 criteria, but insufficient information was available for the other ASRS+ subjects. The median [interquartile range] number of DSM-5 symptom items per patient for the 11 patients with confirmed diagnoses was 5.0 [5.0, 6.0] for inattentive symptoms and 3.0 [2.5, 4.5] for hyperactive-impulsive symptoms. Five of the ASRS+ subjects used ADHD medications, with a median and interquartile range of medication duration of 19.0 [16.0, 35.0] days. Additional analyses to investigate the influence of ADHD medications revealed a significant interaction effect on positive expectancy for alcohol (Regression coefficient estimate = −1.750, p = 0.013), indicating that ASRS+ group patients receiving ADHD medications (n = 5) were more likely to report significantly improved positive expectancy for alcohol during inpatient treatment than unmedicated ASRS+ group patients (n = 38). There were no significant differences in improvement on other ARRS measures between medicated and unmedicated ASRS+ patients.

DISCUSSION

We found that AUD inpatients with more severe ADHD symptoms as measured by the ASRS (ASRS+ group) demonstrated greater relapse risk both at baseline and after inpatient treatment compared to patients with mild or no ADHD symptoms (ASRS− group). This increased risk was largely due to greater stimulus-induced vulnerability and emotionality problems, while other risk dimensions were similar between ASRS+ and ASRS− groups. Further, the ASRS+ group exhibited less improvement in stimulus-induced vulnerability during inpatient treatment (ie, from BL to BD) than the ASRS− group, while the dimensions emotionality problems, compulsivity for alcohol, lack of negative expectancy for alcohol, and positive expectancy for alcohol were improved to a similar degree in both groups. These results support the original hypothesis that AUD patients with ADHD characteristics are at higher risk of relapse, in part due to treatment-refractory stimulus-induced vulnerability. Therefore, inpatient treatment programs for AUD patients with strong ADHD characteristics should place greater emphasis on suppressing stimulus-induced vulnerability.

Several previous studies have reported that AUD patients with comorbid ADHD exhibit more intense alcohol craving, lower behavioral inhibition, and greater impulsivity,^12,33,34 behavioral characteristic that collectively may increase the risk of relapse. In accordance with these findings, the present study suggests that AUD patients with strong ADHD characteristics are at higher risk of relapse due to greater baseline stimulus-induced vulnerability, a dimension strongly associated with behavioral impulsivity and inhibition. For instance, one stimulus-induced vulnerability dimension item is “If the alcohol is placed in front of me, I would drink it.” Therefore, the current results support previous findings that AUD patients with ADHD characteristics are more prone to relapse due to poorer behavioral control. In addition, several studies have suggested that impulsivity is an important mediator of the association between ADHD and alcohol use.^33,35–37

In contrast to stimulus-induced vulnerability, all other dimensions of the ARRS were improved to the same degree in patients with and without ADHD tendencies. We found that the changes in total ARRS scores during hospitalization did not differ between the ASRS+ and ASRS− groups against our hypothesis. In other words, during AUD inpatient treatment, changes in the risk of relapse were similar, regardless of the presence or absence of ADHD characteristics. In addition, only positive expectancy for alcohol was more likely to improve with the initiation of ADHD medication in the ASRS+ group. Our results suggest that pharmacological treatment for ADHD patients could augment AUD inpatient treatment in part, but this finding requires verification, given the small sample size. From this perspective, AUD inpatient treatment may be beneficial to ADHD individuals because patients with comorbid ADHD could benefit from AUD inpatient treatment as much as those without comorbidities. Nevertheless, because the total ARRS scores before discharge were significantly higher in the ASRS+ group, this group may still be more likely to relapse after discharge. The total ARRS scores correlate with relapse within 1 month of evaluation²⁸ and may be predictive of postdischarge relapse to some extent, but this study did not confirm actual postdischarge relapse; hence, caution should be exercised for this interpretation.

The effect of comorbid ADHD on the risk of AUD relapse has been previously assessed using the ARRS.¹³ The results suggested that those with comorbid ADHD had a higher relapse risk, which is in line with our findings. However, there were differences in the ARRS subscores between this previous study and our current study. In the previous study, those with strong ADHD symptoms exhibited higher scores for positive expectancy for alcohol than those with weaker ADHD symptoms, and the scores for lack of negative expectancy for alcohol were the same between the two groups. In contrast, in the present study, those with strong ADHD symptoms had lower scores for lack of negative expectancy for alcohol than those with weaker ADHD symptoms, and the scores for positive expectancy for alcohol were the same between the two groups. Furthermore, there was no difference between the ASRS+ and ASRS− groups in either the baseline score for compulsivity for alcohol or the change in that score during hospitalization. The questions on the compulsivity for alcohol include, for example, “I would do almost anything in order to drink alcohol.” Since this is only an indication of a high level of obsession with drinking and does not necessarily reflect the impulsivity commonly seen in ADHD, it may explain the reason why neither the baseline score for compulsivity for alcohol nor the change in that score during hospitalization differed between the two groups.

This study has several limitations. First, results may not be applicable to outpatient treatment or other inpatient facilities. Therefore, multicenter studies are required for confirmation. Second, all patients were relatively healthy because severely ill patients with alcohol withdrawal delirium requiring many additional treatments were excluded.^38,39 Therefore, the results may be most applicable to spontaneously admitted patients without delirium. Inclusion of patients with alcohol withdrawal delirium may provide useful information about the effects of ADHD on relapse risk after delirium has resolved. Third, this study used a rating scale rather than a diagnosis of ADHD. Since AUDs have been linked to elevated ADHD symptoms,⁴⁰ we cannot rule out the possibility that our present study included patients who had increased ADHD symptoms due to severe AUDs. Fourth, this study used a self-administered rating scale. There are concerns regarding the validity of self-reported ADHD symptoms, including discrepancies between self-assessment and assessment by others.^41,42 Therefore, future studies using ADHD diagnoses based on collected self- and informant ratings are desirable. Fifth, this study did not correct for multiple test comparisons in Tables 3 and 4; because of concerns about a Type 1 error, the results should be interpreted with caution and followed up with a different sample. Sixth, this study did not examine confounding factors for relapse other than ADHD characteristics. Therefore, our study results should be interpreted with caution. The risk of relapse should be reexamined using a study design that includes appropriate confounding factors, such as the number of psychiatric comorbidities and associated medications and executive function.

CONCLUSION

Alcohol use disorder patients with strong ADHD characteristics are at higher risk of relapse at baseline and following inpatient treatment due to persistent stimulus-induced vulnerability.

AUTHOR CONTRIBUTIONS

TK, GS, YK, MT, MM, EM, and TT conceived and designed the study. GS, YK, EM, TT, and HT supervised the study. TK, YK, and MT recruited participants and collected data. TK and GS analyzed the data. TK, GS, and HT drafted the manuscript. All authors provided critical comments to the manuscript.

ACKNOWLEDGMENTS

The authors are very grateful to the patients who participated in this study. We would also like to thank the ward staff for patient care.

FUNDING INFORMATION

This work was supported in part by Brain/MINDS Beyond program from Japan Agency for Medical Research and Development (AMED) (JP18dm0307008).

CONFLICT OF INTEREST

The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT

We did not obtain informed consent for the release of our data and therefore cannot share the data with a third party.

APPROVAL OF THE RESEARCH PROTOCOL BY AN INSTITUTIONAL REVIEWER BOARD

The research project protocol was approved by the Institutional Review Board of Tokyo Medical and Dental University Faculty of Medicine (Approval number M2019-161) and Narimasu Kosei Hospital (Approval number 204) and conforms to the provisions of the Declaration of Helsinki.

INFORMED CONSENT

Written informed consent was obtained from all subjects.

REGISTRY AND THE REGISTRATION NO. OF THE STUDY/TRIAL

Not applicable.