Abstract
Objective
The coronavirus disease 2019 (COVID-19) pandemic has led to a global restriction of public behavior due to lockdowns in various major cities. Lifestyle changes and reduced rates of outpatient lifestyle guidance/consulting may have had some impact on glycemic control in patients with type 2 diabetes. This study analyzed the impact of changes in the frequency of nutritional guidance/consulting (NGC) during the COVID-19 pandemic on outpatient care for type 2 diabetes.
Methods
Among 785 patients, 67 who received regular NGC during the COVID-19 pandemic were assigned to the continuation group (CG), 143 whose NGC was discontinued after the pandemic were assigned to the discontinuation group (DG), and 575 who did not receive regular NGC regardless of the COVID-19 pandemic status were assigned to the irregular NGC group (IGG). The three groups were followed up for two years. Analyses among the three categories were performed using the chi-square test or an analysis of covariance.
Results
The number of diabetes medications after the declaration of the COVID-19 emergency did not markedly increase in the CG (2.0±1.4 to 2.1±1.5, p＞0.05) but significantly increased from 2.2±1.4 to 2.6±1.4 in the DG (p＜0.005) and from 2.2±1.4 to 2.4±1.4 in the IGG (p＜0.005). The increase in HbA1c adjusted for confounders was unchanged at 0.12±1.06% for the CG and -0.07±1.29% for the IGG but was significantly increased at 0.19±1.49% for the DG (p＜0.05).
Conclusion
In patients with type 2 diabetes mellitus, regular nutritional guidance may be important for maintaining good glycemic control, even during the COVID-19 pandemic.
Keywords: type 2 diabetes mellitus, coronavirus disease 2019, nutritional guidance, pandemic, glycemic control
Introduction
Type 2 diabetes mellitus is a disease caused by a genetic or environmental predisposition to insulin deficiency, which leads to various complications due to hyperglycemia. It is widely known that the continuation of regular medical care is essential in the treatment of type 2 diabetes mellitus (1). Personalized nutritional guidance/consulting (NGC) during outpatient visits can help maintain good glycemic control (2). A decrease in the frequency of outpatient visits to medical institutions may lead to an increase in HbA1c levels by decreasing the chance to receive patient-centered glycemic management, such as NGC.
Owing to the coronavirus disease 2019 (COVID-19) pandemic in Japan, a state of emergency was declared nationwide from April 16 to May 14, 2020 (3). The COVID-19 pandemic in Japan has led to a decrease in the number of outpatient tests for patients with diabetes (4). In other countries, diabetes management has been reported to have worsened due to the reduced frequency of outpatient visits after the lockdown (5). At the time of the first emergency declaration issued by the Ministry of Health, Labour and Welfare in Japan, the number of COVID-19 patients was 13.0 per 100,000 population.
The present retrospective study determined how the interruption of NGC due to the COVID-19 pandemic influenced glycemic control in patients with type 2 diabetes mellitus.
Materials and Methods
Study population and patient preparation
A total of 5,127 patients eligible for the present study diagnosed with type 2 diabetes mellitus visited the diabetes outpatient clinic at Kawasaki School Hospital from December 1, 2019 to October 31, 2021. The study protocol, including opt-out informed consent, was approved by the Institutional Review Board of Kawasaki Medical School (No. 5362-00). This study was conducted in accordance with the principles of the Declaration of Helsinki.
The flow of study participants in this study is shown in Fig. 1. We initially selected 930 patients with type 2 diabetes who attended regularly as outpatients for 3 consecutive years between December 1, 2019, and October 31, 2021. Patients with malignancies, autoimmune diseases and those using corticosteroids were excluded from the analysis (145 patients). Finally, 785 patients with type 2 diabetes were divided into 3 groups based on NGC habits. Of the 210 patients who received regular NGC before the COVID-19 epidemic, the 67 who received NGC at least twice a year between December 2019 and March 2021 were defined as the continuation group (CG), and the 143 who had not received NGC for at least one year since December 2019 were defined as the discontinuation group (DG). The 575 patients who did not receive regular NGC at least twice a year, regardless of the COVID-19 epidemic, were defined as the irregular NGC group (IGG). NGC included face-to-face instruction with a dietitian and delivery of dialysis prevention instructions by a nurse and dietitian. NGC at least twice a year was defined as regular guidance. We analyzed the annual trends in weight and blood test results measured every four months.
Figure 1.
The flowchart of the participants and exclusions in this study.
Statistical analyses
Data are expressed as mean and standard deviation. The primary endpoint was the impact of changes in HbA1c levels among the three groups (CG, DG and IGG) and in each category. The secondary endpoint was the impact of changes in the body mass index (BMI), estimated glomerular filtration rate (eGFR), and alanine aminotransferase (ALT) on the frequency of NCG following the advent of the novel COVID-19 pandemic on outpatient care for type 2 diabetes mellitus.
Analyses among the three groups were performed using the chi-square test or an analysis of covariance (ANCOVA) with adjusting for the age, gender and disease duration, using Tukey's method for post hoc tests. An ANCOVA with adjusting for the age, gender, disease duration, BMI, and number of HbA1c tests per year was also performed for HbA1c. Continuous variables were compared every four months among the three groups.
Differences in HbA1c, BMI, eGFR, ALT and medication changes between the beginning and end of the study were evaluated using paired t-tests. To evaluate the number of drugs used in this study, diabetes medications were evaluated using seven factors: insulin secretagogues, such as sulfonylurea/glinide; alpha glucosidase inhibitors; thiazolidines; biguanides; dipeptidyl peptidase-4 inhibitors (DPP4is); sodium-glucose cotransporter-2 inhibitors (SGLT2is); and injectable formulations. Injectable formulations were defined as injectable insulin and/or GLP-1 receptor agonists (GLP-1RAs).
Analyses were performed using the JMP software program (ver. 16.0.1; SAS Institute Japan, Tokyo, Japan) and Microsoft EXCEL for Mac software program (ver. 16.58; Microsoft Japan, Tokyo, Japan) for tabulation.
Results
Clinical characteristics of study participants at baseline
The clinical characteristics of the patients are summarized in Table 1. The mean age of the participants was 65.7±12.3 years old, and the mean HbA1c was 7.14±0.97% in 2019. The frequency of NGC sessions performed between January 2020 and December 2021 was 10.6±1.1 for the CG, 3.7±1.1 for the DG and 2.9±1.7 for the IGG; the value in the CG group was significantly higher than in the other groups (p＜0.05). The number of HbA1c tests per year was 5.9±0.6 for the CG, 5.7±0.4 for the DG and 4.9±0.2 for the IGG, with the value being significantly higher in the CG and DG than in the IGG (p＜0.05). There were no significant differences in HbA1c levels among the 3 groups at study entry after adjusting for the age, gender and duration of diabetes (7.00±0.74% for the CG, 7.19±1.00% for the DG and 7.14±0.98% for the IGG). Patients in the CG had significantly greater body weights than those in the other groups (p＜0.05). The liver, kidney and lipid parameters did not differ markedly among the groups.
Table 1.
Comparison of Various Parameters among Patients with CG, DG, and IGG in 2019, 2020, and 2021.
| Parameter | All subjects | CG | DG | IGG |
|---|---|---|---|---|
| Male/female * | 460/325 | 38/29 | 67/76 | 355/220 |
| Age (years) | 65.7±12.3 | 60.3±13.7 * | 63.9±13.1 | 66.8±11.8 |
| Duration of diabetes (years) | 15.0±9.4 | 14.2±11.5 | 14.2±9.3 | 15.2±9.1 |
| Body weight (kg) | 67.2±15.6 | 73.6±19.8 * | 68.6±18.0 | 67.6±14.9 |
| BMI (kg/m2) | 25.8±5.1 | 28.0±6.0 * | 26.4±6.0 | 25.3±4.6 |
| Family history of diabetes (with/without/unknown) | 415/264/106 | 36/22/9 | 78/42/23 | 301/200/79 |
| Smoking history (current/past/never) * | 118/256/396 | 6/22/36 | 16/39/86 | 96/192/272 |
| Drinking history (more/less than 4 days per week) * | 174/595 | 9/58 | 19/121 | 146/416 |
| Number of nutritional guidance from 2017 to 2019 | 7.2±1.0 | 15.7±1.6 * | 6.1±1.2 * | 2.9±1.3 |
| Number of nutritional guidance from 2020 to 2021 * | 6.7±1.0 | 10.6±1.1 * | 3.7±1.1 * | 2.9±1.7 |
| Number of HbA1c tests per year* | 5.1±0.2 | 5.9±0.6 | 5.7±0.4 | 4.9±0.2 |
| Diabetic neuropathy (with/without/unknown) * | 278/439/68 | 22/41/4 | 66/61/16 | 190/337/48 |
| Diabetic retinopathy (with/without/unknown) * | 170/554/61 | 17/43/7 | 47/88/8 | 106/423/46 |
| Diabetic nephropathy (≥stage 2) (with/without/unknown) * | 257/494/34 | 20/46/1 | 60/78/5 | 177/370/28 |
| History of cerebrovascular disease (with/without/unknown) | 85/684/16 | 7/60/0 | 18/124/1 | 60/500/15 |
| History of coronary artery disease (with/without/unknown) | 75/694/16 | 6/61/0 | 20/119/4 | 49/513/13 |
| Systolic blood pressure (mmHg) | 128.4±15.8 | 130.1±22.0 | 126.6±16.4 | 128.6±14.9 |
| Diastolic blood pressure (mmHg) | 74.6±11.2 | 77.0±11.4 | 73.9±13.4 | 74.5±10.6 |
| HbA1c (%, NGSP) | 7.14±0.97 | 7.02±0.74 | 7.19±1.00 | 7.14±0.97 |
| AST (U/L) | 24.4±13.6 | 23.9±6.8 | 27.2±21.7 * | 23.8±11.1 |
| ALT (U/L) | 25.0±16.9 | 28.9±13.3 * | 29.0±26.6 * | 23.4±13.3 |
| eGFR (mL/min/1.73m2) | 67.9±20.2 | 70.2±22.3 | 67.9±19.4 | 67.5±20.0 |
| UA (mg/dL) | 5.1±1.3 | 5.2±1.3 | 5.1±1.3 | 5.1±1.4 |
| Total cholesterol (mg/dL) | 191.9±50.0 | 205.0±49.0 | 179.3±47.5 | 179.3±47.5 |
| Triglyceride (mg/dL) | 135.0±89.2 | 133.2±64.5 | 149.0±131.3 | 131.8±77.5 |
| LDL cholesterol (mg/dL) | 95.5±26.1 | 96.3±25.5 | 95.6±27.0 | 95.4±26.0 |
| BMI, eGFR and ALT values in September 2020 and September 2021 | ||||
| BMI in September 2020 (kg/m2) | 25.9±5.2 | 27.6±6.0 * | 25.7±5.4 | 25.5±4.9 |
| BMI in September 2021 (kg/m2) | 25.4±5.0 | 27.0±5.9 | 26.4±6.0 | 25.3±4.6 |
| eGFR in September 2020 (mL/min/1.73m2) | 66.2±19.5 | 67.1±24.3 | 63.0±17.1 | 66.9±17.1 |
| eGFR in September 2021 (mL/min/1.73m2) | 64.5±22.2 | 66.9±25.5 | 66.0±19.3 | 63.8±22.4 |
| ALT in September 2020 (U/L) | 24.8±16.4 | 29.7±18.8 * | 26.8±22.3 | 23.6±13.8 |
| ALT in September 2021 (U/L) | 24.2±15.8 | 27.2±23.3 | 25.4±17.6 | 23.5±14.1 |
Data are presented as mean±standard deviation. CG: continuation group, DG: discontinuation group, IGG: irregular guidance group, BMI: body mass index, AST: aspartate aminotransferase, ALT: alanine aminotransferase, eGFR: estimated glomerular filtration rate, UA: uric acid, LDL-cholesterol: low-density lipoprotein cholesterol. *: p<0.05 chi-square test, and ANCOVA compared to category of IGG adjusted for age, gender, and disease duration.
Diabetes medications during the observational period
The medications used for diabetes are listed in Table 2. At the beginning of the study, there were no marked differences in drug use between the groups. The overall trend during the 2-year observation period for all patients showed a significant increase in the use of injectable formulations (p＜0.05), DPP4is (p＜0.005), biguanides (p＜0.0005), and SGLT2is (p＜0.0005) and a significant decrease in the use of insulin secretagogues (p＜0.0005). In the CG, drug use did not change significantly. The difference in the number of diabetes medications used at the beginning and end of the study was 0.07±0.23 (from 2.0±1.4 to 2.1±1.5, p=n.s.) for the CG, 0.38±0.16 (from 2.2±1.4 to 2.6±1.4, p＜0.005) for the DG and 0.08±0.19 (from 2.2±1.4 to 2.4±1.4, p＜0.005) for the IGG, with the number of medications significantly increased in the DG and IGG.
Table 2.
Diabetes Medication Usage Rates in Subjects with CG, DG and IGG at Various Time Points.
| Parameter | All subjects | CG | DG | IGG |
|---|---|---|---|---|
| (n=785) | (n=67) | (n=143) | (n=575) | |
| At the beginning of the study (2019) | ||||
| Injectable formulation (%) | 17.2 | 22.4 | 21.0 | 15.7 |
| Insulin secretagogue (%) | 26.8 | 17.9 | 28.0 | 27.7 |
| Alpha glucosidase inhibitor (%) | 12.1 | 10.4 | 10.5 | 12.7 |
| Thiazolidine (%) | 28.2 | 23.9 | 23.1 | 29.9 |
| Biguanide (%) | 50.1 | 46.3 | 51.7 | 50.1 |
| Dipeptidyl peptidase 4 inhibitor (%) | 56.2 | 41.8 | 53.1 | 58.6 |
| Sodium-glucose cotransporter-2 inhibitor (%) | 29.4 | 37.3 | 32.2 | 27.8 |
| At the end of the study (2021) | ||||
| Injectable formulation (%) | 21.9 | 25.4 | 30.1 | 19.5 |
| Insulin secretagogue (%) | 26.5 | 16.4 | 26.6 | 27.7 |
| Alpha glucosidase inhibitor (%) | 10.6 | 3.0 | 7.7 | 12.2 |
| Thiazolidine (%) | 22.7 | 16.4 | 16.1 | 25.0 |
| Biguanide (%) | 52.7 | 47.8 | 59.4 | 51.7 |
| Dipeptidyl peptidase 4 inhibitor (%) | 61.5 | 53.7 | 62.2 | 62.3 |
| Sodium-glucose cotransporter-2 inhibitor (%) | 42.0 | 44.8 | 51.0 | 39.5 |
| Number of diabetes drugs used | ||||
| At the beginning of the study (2019) | 2.2±1.4 | 2.0±1.4 | 2.2±1.4 | 2.2±1.4 |
| At the end of the study (2021) | 2.4±1.4 | 2.1±1.5 | 2.6±1.4 | 2.4±1.4 |
| Difference between beginning and end of study | 0.22±0.10 | 0.07±0.23 | 0.38±0.16 * | 0.08±0.19 |
Data are presented as mean±standard deviation. CG: continuation group, DG: discontinuation group, IGG: irregular guidance group. *: p<0.05 ANCOVA compared to category of IGG.
Time course of HbA1c in the CG, DG, and IGG
To evaluate the influence of the COVID-19 pandemic on glycemic control, we measured HbA1c values every four months in outpatients with type 2 diabetes (Fig. 2). After the first emergency declaration, the DG had higher HbA1c levels than the CG and IGG (p＜0.05), and there was a significant difference in values between the DG and IGG (p＜0.05). The differences in the HbA1c levels at the beginning (2019) and end of the study (2021) (ΔHbA1c) for the CG, DG, and IGG groups are shown in Fig. 3; the ΔHbA1c for the CG, DG and IGG were 0.12±1.06%, 0.19±1.49%, and -0.07±1.29%, respectively. Only the ΔHbA1c in the DG was significantly elevated (p＜0.05). Even after further adjusting for the BMI and number of HbA1c tests per year, the significance of glycemic control in the DG remained (Fig. 4).
Figure 2.
HbA1c measured every four months after January 2020. *p<0.05 (ANCOVA adjusted for age, gender and disease duration) between the discontinuation group (DG) and irregular guidance group (IGG). 1st: 1st state of emergency from April 16 to May 14, 2020. 2nd: 2nd state of emergency declaration from June 1 to June 20, 2021.
Figure 3.
Differences in HbA1c between the beginning and end of the study (ΔHbA1c). *p<0.05, paired t-test between the discontinuation group (DG) and irregular guidance group (IGG).
Figure 4.
HbA1c measured every 4 months after January 2020. *p<0.05 (ANCOVA adjusted for age, gender, disease duration, BMI and number of HbA1c tests per year) between the discontinuation group (DG) and irregular guidance group (IGG). 1st: 1st state of emergency from April 16 to May 14, 2020. 2nd: 2nd state of emergency declaration from June 1 to June 20, 2021.
Time course of the BMI, eGFR and ALT in the CG, DG, and IGG
The BMI, eGFR, and ALT levels at study entry in 2019, September 2020 and September 2021 are shown in Table 1. The BMI at the beginning of the study was significantly higher in the CG than in the DG and IGG (p＜0.005, respectively). At the beginning of the study, the ALT levels were significantly higher in the CG and DG than in the IGG (p＜0.005, respectively). At the end of the study, there were no marked differences in the BMI, eGFR, or ALT levels among the groups. The differences in the BMI, eGFR and ALT levels at the beginning and end of the study did not differ markedly among the groups.
Discussion
These retrospective data showed that glycemic control worsened only in patients with diabetes who discontinued regular NGC under the declared state of emergency of COVID-19. This is the first study to analyze the frequency of NGC and HbA1c trends in Japan, and the data in this study indicate the importance of continuing regular NGC even during the COVID-19 pandemic.
The COVID-19 pandemic has caused lockdowns worldwide and restricted the actions of the public. In Japan, a state of emergency was declared between April 16 and May 14, 2020, which greatly affected the lives of citizens. The curtailment of outings due to COVID-19 had a negative impact on diabetes care by worsening diabetes management, such as decreased hospital visits and increased obesity (5-7). In the present study, the number of medications was increased in the DG and IGG, but not in the CG, to maintain HbA1c levels. Furthermore, HbA1c levels were also increased in the DG despite this group receiving an increased number of medications. These results suggest that discontinuation of NGC was associated with an increase in HbA1c and the number of medications. HbA1c levels in the IGG were similar to the CG, but the number of medications was significantly increased in the IGG, but not in the CG. This suggests that NGC potentially inhibited the increase in HbA1c. It seems necessary to continue NGC even during the COVID-19 pandemic.
In a study examining medical care in Japan during the COVID-19 pandemic, the number of prescriptions and hospital visits for chronic conditions, such as diabetes, decreased temporarily during the first state of emergency compared to the second one (8). In addition, a study reported that the effective use of telemedicine and clinic visits was associated with improved HbA1c levels in patients with diabetes under a declared state of emergency in Japan (9). There are also reports that remote NGC by dietitians has effectively improved the care of adult patients with type 2 diabetes (10). We believe that the development of such remote NGC may improve diabetes management during the COVID-19 pandemic.
Although there have been several reports of weight gain due to the COVID-19 pandemic, the ΔBMI did not change markedly among the three groups in this study. It has been reported that weight gain in patients with type 2 diabetes during the COVID-19 pandemic was caused by increased snacking due to a decrease in eating out (11). One possible reason for the invariance in the BMI in this study may be the significant increase in the use of SGLT2is and GLP-1RAs and the significant decrease in the use of insulin secretagogues in all groups. Unfortunately, only 210 of 785 participants in this study received regular NGC prior to the COVID-19 pandemic. As this study showed that regular NGC may improve diabetes control and prevent patients becoming overweight, future efforts should be focused on increasing the implementation rate of regular NGC.
Several limitations associated with the present study warrant mention. First, this was a single-center, retrospective study. Since most of the patients in this study were residents of Kurashiki City, Okayama Prefecture, the generalizability may be limited. In addition, the medical insurance system, clinical style, differences in NGC style and pandemic situation may not apply to other countries. Second, the analysis did not separate NGC delivered by dietitians from that in which nurses were also present. Next, the CG/DG had received NGC prior to the COVID-19 outbreak, while the IGG had not. Patients' motivation for health may therefore have differed among groups. Finally, administrative infection control and self-defense measures related to the COVID-19 epidemic may also have affected the physical activity of patients. Physical activity, such as going out and exercising, may have thus been reduced in patients who avoided NGC and did not visit the hospital, but these possibilities were not considered in this study. In addition, actual dietary changes were not objectively evaluated.
In conclusion, discontinuation of regular NGC was associated with increased HbA1c levels and an increased number of medications during the COVID-19 pandemic. These data clearly indicate that it is very important to continue regular NGC in order to maintain good glycemic control, even during the COVID-19 pandemic, in patients with type 2 diabetes mellitus.
Author's disclosure of potential Conflicts of Interest (COI).
Kohei Kaku: Advisory role, Sanwa Kagaku; Honoraria, Novo Nordisk Pharma, Sanwa Kagaku, Takeda, Taisho Pharma, MSD, Kowa, Sumitomo Pharma, Mitsubishi Tanabe Pharma, AstraZeneca, Boehringer Ingelheim, Daiichi Sankyo and Sanofi. Hideaki Kaneto: Honoraria, Novo Nordisk Pharma, Sanofi, Eli Lilly, Boehringer Ingelheim, Taisho Pharma, Sumitomo Dainippon Pharma, Takeda Pharma, Ono Pharma, Daiichi Sankyo, Mitsubishi Tanabe Pharma, Kissei Pharma, MSD, AstraZeneca, Astellas, Novartis, Kowa and Abbott.
Acknowledgement
The abstract of this report was presented at the 65th Annual Meeting of the Japan Diabetes Society (Hyogo).
References
- 1. Karter AJ, Parker MM, Moffet HH, et al. Missed appointments and poor glycemic control: an opportunity to identify high-risk diabetic patients. Med Care 42: 110-115, 2004. [DOI] [PubMed] [Google Scholar]
- 2. Evert AB, Dennison M, Gardner CD, et al. Nutrition therapy for adults with diabetes or prediabetes: a consensus report. Diabetes Care 42: 731-754, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Prime Minister's Office in JapanThe declaration of the lifting of the state of emergency in response to the novel coronavirus disease [Internet]. [cited 2020 Dec 28]. Available from: https://japan.kantei.go.jp/ongoingtopics/_00027.html
- 4. Ikesu R, Miyawaki A, Sugiyama T, Nakamura M, Ninomiya H, Kobayashi Y. Trends in diabetes care during the COVID-19 outbreak in Japan: an observational study. J Gen Intern Med 36: 1460-1462, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Pellegrini M, Ponzo V, Rosato R, et al. Changes in weight and nutritional habits in adults with obesity during the “lockdown” period caused by the COVID-19 virus emergency. Nutrients 12: 2016, 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Yamada M, Kimura Y, Ishiyama D, et al. The influence of the COVID-19 pandemic on physical activity and new incidence of frailty among initially non-frail older adults in Japan: a follow-up online survey. J Nutr Health Aging 25: 751-756, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Ojo O, Wang XH, Ojo OO, et al. The effects of COVID-19 lockdown on glycaemic control and lipid profile in patients with type 2 diabetes: a systematic review and meta-analysis. Int J Environ Res Public Health 19: 1095, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Yamaguchi S, Okada A, Sunaga S, et al. Impact of COVID-19 pandemic on healthcare service use for non-COVID-19 patients in Japan: retrospective cohort study. BMJ Open 12: e060390, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Onishi Y, Yoshida Y, Takao T, et al. Diabetes management by either telemedicine or clinic visit improved glycemic control during the coronavirus disease 2019 pandemic state of emergency in Japan. J Diabetes Investig 13: 386-390, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Benson GA, Sidebottom A, Hayes J, et al. Impact of ENHANCED (dietitians helping patients care for diabetes) telemedicine randomized controlled trial on diabetes optimal care outcomes in patients with type 2 diabetes. J Acad Nutr Diet 119: 585-598, 2019. [DOI] [PubMed] [Google Scholar]
- 11. Takahara M, Watanabe H, Shiraiwa T, et al. Lifestyle changes and their impact on glycemic control and weight control in patients with diabetes during the coronavirus disease 2019 pandemic in Japan. J Diabetes Investig 13: 375-385, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]