Brachial-Ankle Pulse Wave Velocity is Associated with Incident Dementia in Patients with Cerebral Small-Vessel Disease

Abstract

Aims: Increased arterial stiffness is associated with the severity of cerebral small-vessel disease (SVD) and may predict incident dementia. This study investigated the predictive value of brachial-ankle pulse wave velocity (ba-PWV) for dementia and cognitive decline.

Methods: Data were obtained from a Japanese cohort of 478 patients who underwent ba-PWV measurement. Magnetic resonance imaging (MRI) was used to evaluate SVD severity. The Mini-Mental State Examination (MMSE) and the Japanese version of the Montreal Cognitive Assessment (MoCA-J) were used to assess the cognitive function. The primary outcome was the incidence of dementia. The secondary outcome was cognitive change during three years of follow-up.

Results: The median age was 71 years old, 61% were men, and the median ba-PWV was 1787 cm/s. Dementia was diagnosed in 23 patients during a mean follow-up of 4.8 years. A Cox proportional hazard regression analysis revealed that the highest quartile (ba-PWV ≥ 2102 cm/s) was associated with a significantly higher risk of dementia than the first to third quartiles (ba-PWV ≤ 2099 cm/s) after adjusting for risk factors, the mean blood pressure, the MoCA-J score, and SVD severity (adjusted HR, 3.40; 95% CI, 1.24-9.34; P=0.018). Longitudinal cognitive changes in 192 patients indicated that ba-PWV was negatively related to changes in the MoCA-J score (r=-0.184, P=0.011). The decline in the MoCA-J score in the highest quartile was greater than that in the first to third quartiles after adjusting for risk factors, SVD severity, and baseline MoCA-J score (P=0.017).

Conclusions: ba-PWV was associated with incident dementia and cognitive decline, independent of age, risk factors, the baseline cognitive function, and the SVD severity.

Introduction

The large artery has a powerful cushioning function that protects the microvasculature from harmful fluctuations in pressure and blood flow. Large-artery stiffening, which occurs with aging and various pathological states, impairs the cushioning function, has important consequences on vascular health, and predicts cardiovascular events¹⁾.

The influence of arterial stiffness on cognitive impairment and dementia has attracted attention, although whether or not it is a true risk or confounding factor remains unknown²⁾. Brachial-ankle pulse wave velocity (ba-PWV), which is obtained easily and noninvasively without specialized technical expertise, is a simple and reliable method for evaluating the stiffness of large- to middle-sized arteries³⁾. Vascular risk factors such as age and hypertension strongly influence arterial stiffness, which is associated with imaging evidence of cerebral small-vessel disease (SVD)^{4, 5)}.

A meta-analysis revealed that cerebral SVD, especially moderate or severe white matter hyperintensities (WMH), can predict incident dementia⁶⁾. These confounding factors make the causal relationship between arterial stiffness and dementia unclear.

Aim

This study investigated the predictive value of ba-PWV for incident dementia and cognitive decline, considering the SVD severity and baseline cognitive function in patients with cerebral SVD.

Methods

Study Design and Patients

Data were derived from a prospective study conducted by the Tokyo Women’s Medical University Cerebral Vessel Disease (TWMU CVD) Registry (registration URL: https://upload.umin.ac.jp/cgi-open-bin/icdr_e/ctr_view.cgi?recptno=R000030620; UMIN000026671). All of the participants provided their written informed consent. This study was approved by the Institutional Review Board of Tokyo Women’s Medical University (approval number 3621).

The research protocol and inclusion criteria of the TWMU CVD registry have been previously described in detail⁷⁾. In brief, this study included consecutive patients ≥ 40 years old who presented with CVD on magnetic resonance imaging (MRI) and with ≥ 1 cerebrovascular risk factor, such as arterial hypertension, diabetes mellitus, dyslipidemia, coronary artery disease, atrial fibrillation, or smoking. Patients underwent the Mini Mental State Examination (MMSE) and Japanese version of the Montreal Cognitive Assessment (MoCA-J)⁸⁾ within 1 year of the MRI examination. The exclusion criteria for the registry were any type of aphasia, evidence of dementia (Clinical Dementia Rating⁹⁾ ≥ 1), and dependence in activities of daily living and walking. In this registry, we gathered as much predefined clinical information as possible, including data concerning the motor function, willingness state, blood test values, electrocardiogram, carotid ultrasound, echocardiography, ba-PWV, ankle brachial pressure index, flow-mediated vasodilation, and other examinations at enrollment. This study excluded patients who experienced vascular events within one month of enrollment.

The protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki, in line with the Ethical Guidelines for Epidemiological Research by the Japanese government and the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. The TWMU CVD enrolled 1011 outpatients from October 2015 to July 2019. MRI is generally performed to investigate lesions in patients with a history of stroke or suspicious neurological symptoms (e.g. headache, vertigo, dizziness, numbness, syncope, or subjective memory impairment).

MRI Protocol and Assessments

Each participant underwent brain MRI within one year before entry into the registry. The MRI assessment included white matter hyperintensities (WMH), consisting of periventricular hyperintensities (PVH), deep white matter hyperintensities (DWMH), and lacunar infarctions (LI) for SVD^{10, 11)}. Axial fluid-attenuated inversion recovery (FLAIR) images were used to visually assess WMH severity. PVH and DWMH were scored 0-3 based on the Fazekas scale (0=none; 1=mild; 2=moderate; 3=severe). Lesions in the basal ganglia, internal capsule, centrum semiovale, or brainstem with hypointensity on T1-weighted imaging (T1WI), hyperintensity on T2-weighted imaging (T2WI), and a hyperintense rim around the cavity on FLAIR were defined as LI, with sizes ranging from 3 to 15 mm. One point was awarded for WMH if PVH Fazekas 3 (extending into the deep white matter) and/or DWMH Fazekas 2-3 (confluent or early confluent) were present¹¹⁾. If ≥ 1 asymptomatic lesion was present, 1 point was assigned for LI¹¹⁾. Two trained board-certified neurologists (M. H. and M.S.) who were blinded to the clinical details rated all cerebral SVD-related results. The inter-rater κ for each SVD feature or MTA score was between 0.80 and 0.85. A third rater (K.K.) was consulted in cases of disagreement.

Vascular Risk Factors

Hypertension was defined as a blood pressure of ≥ 140/90 mmHg on measurements taken at least twice at intervals of 2 min and on another day or on the intake of antihypertensive medications. Diabetes was defined as fasting plasma glucose level ≥ 126 mg/dL, HbA1c level ≥ 6.5%, or receiving antidiabetic therapies. Dyslipidemia was defined as a low-density lipoprotein cholesterol level ≥ 140 mg/dL, total cholesterol level ≥ 220 mg/dL, triglyceride level ≥ 150 mg/dL, or intake of cholesterol-lowering therapies. Smoking status was evaluated based on current habits. The baseline kidney function was indicated by the estimated glomerular filtration rate (eGFR), and chronic kidney disease (CKD) was defined as an eGRF ＜60 mL/min or proteinuria.

Measurement of the ba-PWV

Medical technologists certified by the Japanese Association of Medical Technologists measured ba-PWV according to a standardized methodology using an automated device (VP-1000; Collin Co Ltd., Komaki, Japan), as described previously¹²⁾. This device simultaneously measures the bilateral brachial and posterior tibial arterial pulse waveforms and arterial blood pressure using a sociometric method. The ba-PWV was automatically calculated as the transmission distance divided by the transmission time. All technologists involved were fully trained in ba-PWV measurements, and each of them performed ≥ 250 ba-PWV examinations per year. Higher ba-PWV values between the right and left sides were used in the analysis, according to our previous study¹²⁾. The mean values of the right and left ba-PWV were used in the analysis. Systolic, diastolic, and mean blood pressures were measured in the supine position during the ba-PWV measurement.

Follow-Up

All patients were invited to receive the second global cognitive function assessment three years after enrollment. The patients were followed-up until March 2023, and all-cause dementia was evaluated. Patients who withdrew consent and those lost to follow-up were censored at the last visit.

The Diagnosis of Dementia

Patients visited outpatient clinics to control for risk factors every three months to prevent stroke and vascular events. Changes in patients’ general medical conditions were obtained annually through medical records and interviews. Furthermore, several aspects of daily cognitively driven functioning were assessed at each clinical visit to rate patients on the Clinical Dementia Rating (CDR). Final follow-up data were collected from April 2022 to March 2023. Neurologists periodically investigated the patients with suspected cognitive decline during the follow-up period. Clinically significant cognitive impairment was defined as an MMSE score of ＜24 or a decline of ≥ 1.5 standard deviation (SD) of the change in score, as previously reported¹³⁾. This corresponded to a decline of ≥ 3 points in the MMSE score in this study. In addition, patients were considered to have probable dementia if they had two consecutive semiannual CDR scores of ≥ 1 and did not revert to normal cognition. The medical records of all participants were continuously monitored at our clinic and other clinics to obtain information on the diagnosis of dementia and avoid missing incident dementia cases. Furthermore, a phone interview to collect clinical data was conducted with the patient and caregiver whenever possible, for patients who could not visit the clinic. Finally, an independent committee of neurologists reviewed all potential dementia cases with all available information to reach a consensus on the diagnosis according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5)¹⁴⁾. Dementia subtypes were diagnosed according to standardized criteria^{15, 16)}. Time to dementia was defined as the time between the baseline visit and date of the dementia diagnosis. In addition, patients were followed up until death or refusal to undergo further participation. Patients who did not progress to dementia were censored at the last visit.

Statistical Analyses

The JMP 14 Pro software program (SAS Institute, Cary, NC, USA) was used for all analyses. Quantitative variables are expressed as the mean±SD or median with the interquartile range (IQR) for normally and non-normally distributed data, respectively, whereas categorical variables are expressed as frequencies and percentages. A one-way analysis of variance or Kruskal–Wallis test for quantitative variables and χ² test for qualitative variables were used for comparisons among multiple groups, as appropriate. The patients were categorized into 4 groups based on ba-PWV quartiles to investigate the effect of ba-PWV on dementia and cognitive decline: first (≤ 1525 cm/s), second (1526-1786 cm/s), third (1788-2099 cm/s), and fourth (≥ 2102 cm/s). The Kaplan–Meier method was used to estimate event rates, and the log-rank test was used to assess intergroup differences. Cox proportional hazard regression models were used to evaluate the association between ba-PWV and the risk of incident dementia by calculating the hazard ratios (HRs) with 95% confidence intervals (CIs). A linear regression analysis was used to investigate the relationship between ba-PWV and changes in MMSE or MoCA-J scores during the three years of follow-up. Changes in MMSE or MoCAJ scores were then further compared between the ba-PWV-quartile groups by adjusting for age, sex, risk factors, SVD, and baseline MoCA-J. Statistical significance for all analyses was set at a P-value of ＜0.05. All variables with a P-value of ＜0.10 in the univariate analysis, including age, sex, hypertension, mean blood pressure at ba-PWV measurement, diabetes mellitus, CKD, atrial fibrillation, baseline MoCA-J score, WMH, and presence of lacunae, were entered in the multivariate models.

Results

The TWMU CVD study enrolled 1,011 outpatients from October 2015 to June 2019 (Fig.1). The analysis excluded patients who refused testing for MMSE (n=28), had MMSE scores ＜24 (n=77) and modified Rankin Scale (mRS) score ≥ 2 (n=46), and did not undergo a ba-PWV examination (n=382) because they did not agree to receive ba-PWV measurement or could not fit it in their schedule. All analyses were based on 478 patients who underwent a ba-PWV examination at enrollment with complete baseline data (Fig.1). Among the 478 patients, 192 underwent a second MMSE and MoCA-J examination 3 years after enrollment. During the mean follow-up period of 4.8 years, 26 patients who died, 55 patients who withdrew consent primarily because of the transfer of care to another provider, and 34 patients who were lost to follow-up were censored at the last visit.

Fig.1. Flowchart of the study

MMSE: minimental state examination; mRS: modified Rankin scale.

Table 1 summarizes the baseline characteristics of the study participants. The median patient age was 71 years old, and 61% of the patients were male. Hypertension, diabetes mellitus, dyslipidemia, atrial fibrillation, CKD, current smoking, history of stroke or transient ischemic attack (TIA), and cardiovascular disease history were reported in 67%, 33%, 53%, 12%, 48%, 8%, 56%, and 14% of the patients, respectively. The median MMSE and MoCA-J scores were 28 and 26, respectively, and 22% and 47% of patients had severe or moderate WMH and at least 1 lacuna, respectively. The median ba-PWV was 1787 cm/s.

Table 1.Patient demographics in reference to incident dementia

	Total	All-cause dementia	No dementia	P
Patient Number	478	23	455
Age, median (IQR), years	71 (63-77)	80 (73-83)	70 (63-76)	＜0.001
Sex, male, %	61.3	38.7	61.3	0.966
Risk factors
Hypertension, %	66.5	69.6	66.2	0.746
Diabetes mellitus, %	32.6	27.2	32.9	0.583
Dyslipidemia, %	53	39.1	53.7	0.173
Atrial fibrillation, %	12	31.8	10.9	0.003
Chronic kidney disease, %	47.9	69.6	46.8	0.033
Current smoking, %	7.9	8.7	8.1	0.883
A history of stroke or TIA, %	55.6	52.2	55.7	0.738
A history of cardiovascular disease, %	14.2	8.7	14.5	0.436
Medication, statin use, %	54.1	47.8	54.4	0.536
Anti-hypertensive drug use, %	66.2	65.2	66.3	0.915
Anti-platelet use, %	58.6	56.5	58.7	0.915
Anti-Coagulant use, %	15.7	34.8	14.7	0.838
MMSE score	28 (27-30)	27 (25-27)	28 (27-30)	0.002
MoCA-J score	24 (21-27)	24 (21-27)	22 (19-24)	＜0.001
Education years	16 (12-16)	16 (12-16)	16 (12-16)	0.898
MRI findings
WMH 0/1, %	21.8	47.8	20.4	0.002
Lacunes 0/1, %	46.8	52.1	46.6	0.6
ba-PWV (cm/s), median (IQR)	1787 (1526-2100)	2232 (1893-2584)	1774 (1517-2057)	＜0.001
Mean blood pressure at ba-PWV measurement, mmHg	101 (92-111)	112 (94-117)	101 (92-110)	0.084
Follow-up duration, years	4.69 (4.04-5.47)	3.31 (2.14-4.87)	4.71 (4.08-5.50)	＜0.001

TIA, transient ischemic attack; MMSE, Mini-Mental State Examination; MoCA-J, Japanese version of the Montreal Cognitive Assessment; WMH, white matter hyperintensities; ba-PWV, brachial-ankle pulse wave velocity

Table 2 summarizes the baseline characteristics of the ba-PWV-quartile groups. Higher ba-PWV-quartile groups were associated with age, sex, hypertension, mean blood pressure at ba-PWM measurement, diabetes mellitus, CKD, history of cardiovascular disease, MMSE and MoCA-J scores, and WMH. Supplemental Table 1 summarizes the association between ba-PWV as a continuous variable and the risk factors, cognitive function, and MRI results. Older age, hypertension, the mean blood pressure, diabetes mellitus, atrial fibrillation, CKD, a history of stroke or TIA, low MMSE and MoCA-J scores, the presence of severe or moderate WMH, and lacunae were significantly associated with ba-PWV.

Table 2.Baseline characteristics in reference to the ba-PWV-quartile groups

ba-PWV Quartile	1st quartile (979–1525 cm/ sec, N = 119)	2nd quartile (1526–1786 cm/sec, N = 120)	3rd quartile (1788–2099 cm/sec, N = 120)	4th quartile (2102–4663 cm/sec, N = 119)	P
Age, median (IQR), years	62 (51-69)	70 (65-74)	74 (67-78)	76 (71-81)	＜0.001
Sex, male %	58.0	65.0	59.2	63.0	0.654
Hypertension, %	49.6	63.9	71.4	80.7	＜0.001
Mean blood pressure at ba-PWV measurement, mmHg	95 (84-101)	100 (91-108)	102 (96-115)	110 (102-121)	＜0.001
Diabetes mellitus, %	23.3	27.7	37.0	42.4	0.007
Dyslipidemia, %	51.3	52.1	47.9	60.5	0.250
Atrial fibrillation, %	6.1	14.3	13.5	13.7	0.174
Chronic kidney disease, %	37.0	45.0	54.2	55.5	0.013
Current smoker, %	10.9	9.2	6.1	5.2	0.321
A history of stroke or TIA, %	50.4	58.3	50.0	63.6	0.102
A history of cardiovascular disease, %	9.2	15.8	21.7	10.1	0.020
Medication, statin use, %	46.2	55.8	52.5	61.9	0.106
MMSE score	29 (28-30)	29 (27-30)	28 (27-30)	27 (26-29)	＜0.001
MoCA-J score	26 (23-28)	24 (23-28)	24 (21-26)	23 (20-25)	＜0.001
Education, years	16 (12-16)	16 (12-16)	14 (12-16)	15 (12-16)	0.074
WMH 0/1, %	16.0	19.2	19.2	32.8	0.008
Lacunes 0/1, %	42.0	45.0	43.7	56.8	0.093

TIA, transient ischemic attack; MMSE, Mini-Mental State Examination; MoCA-J, Japanese version of the Montreal Cognitive Assessment; WMH, white matter hyperintensities; ba-PWV, brachial-ankle pulse wave velocity

Supplemental Table 1.Association between ba-PWV and risk factors, baseline cognitive function, and MRI findings

	r or median (IQR)	P
Age, years	0.431	＜0.001
Sex, male/female	1790 (1503-2087)/1786 (1531-2112)	0.587
Risk factors
Hypertension, Y/N	1860 (1611-2189)/1707 (1420-1923)	＜0.001
Mean blood pressure at ba-PWV measurement	0.509	＜0.001
Diabetes mellitus, Y/N	1891 (1622-2214)/1747 (1509-2031)	＜0.001
Dyslipidemia, Y/N	1802 (1537-2147)/1774 (1524-2023)	0.117
Atrial fibrillation, Y/N	1846 (1668-2368)/1784 (1517-2084)	0.033
Chronic kidney disease, Y/N	1843 (1603-2160)/1735 (1492-2034)	0.012
Current smoking, Y/N	1592 (1452-1917)/1790 (1527-2108)	0.379
a history of stroke or TIA, Y/N	1795 (1376-2141)/1783 (1510-2030)	0.019
Previous cardiovascular disease, Y/N	1819 (1673-2033)/1783 (1513-2126)	0.632
MMSE score	-0.197	＜0.001
MoCA-J score	-0.212	＜0.001
MRI findings
WMH, Y/N	1892 (1647-2717)/1772 (1510-2021)	＜0.001
Lacues, Y/N	1812 (1559-2215)/1769 (1509-2021)	0.006

Abbreviations: ba-PWV = brachial-ankle pulse wave velocity, TIA = transient ischemic attack, MMSE = mini-mental state examination, MoCA-J = Japanese version of the Montreal Cognitive Assessment, WMH = white matter hyperintensities

Dementia was diagnosed in 23 patients during a mean follow-up period of 4.57±1.38 years. Among the 23 dementia events, 15, 6, and 2 involved Alzheimer’s disease, vascular, and mixed dementia, respectively. The follow-up durations were 4.73±1.33, 4.43±1.51, 4.67±1.35, and 4.43±1.34 years in the first (ba-PWV: 979-1525 cm/s, n=119), second (ba-PWV: 1526-1786 cm/s, n=120), third (ba-PWV: 1788-2099 cm/s, n=120), and fourth quartiles (ba-PWV: 2102-4663 cm/s, n=119), respectively. The follow-up durations were similar in all four groups. The median follow-up duration until dementia was 3.31 years and significantly shorter than that to the last visit in the non-dementia group (Table 1). During the follow-up period, 25 patients experienced stroke. Among them, 2 patients developed vascular dementia later, but 21 of 23 patients with incident dementia did not experience stroke during the follow-up period. Age, atrial fibrillation, CKD, MMSE and MoCA-J scores, and WMH severity were significantly associated with the incidence of dementia (Table 1). A Kaplan-Meier analysis indicated that the likelihood of incident dementia was significantly higher in the highest ba-PWV group than in the other three groups when compared individually (Fig.2A, P＜0.001) and when the first to third ba-PWV groups were combined (Fig.2B, P＜0.001). The results were similar when we used the mean values of the right and left ba-PWV in the analysis (Supplemental Fig.1A and 1B). We included 56 patients with atrial fibrillation in whom the reliability of ba-PWV measurement was expected to be diminished. In addition, we investigated the association between ba-PWV and incident dementia in 422 patients without atrial fibrillation. A Kaplan-Meier analysis again indicated that the likelihood of incident dementia was significantly higher in the highest ba-PWV group than in the other three groups in this population (Supplemental Fig.1C and 1D).

Fig.2.

(A, B) Incident dementia according to the ba-PWV-quartile groups. A Kaplan-Meier analysis of the dementia-free survival in the first (ba-PWV ≤ 1525 cm/s), second (ba-PWV: 1526-1786 cm/s), third (ba-PWV: 1788-2099 cm/s), and fourth (ba-PWV ≥ 2102 cm/s)-quartile groups (A), and in the first to third (ba-PWV ≤ 2099 cm/s) and fourth (ba-PWV ≥ 2102 cm/s)-quartile groups (B).

(C, D) Relationship between ba-PWV and changes in MMSE (C) and MoCA-J scores (D) over three years.

Supplemental Fig.1.

(A, B) Incident dementia according to the mean ba-PWV quartile groups. Kaplan-Meier analysis of dementia-free survival in the first (mean ba-PWV ≤ 1520 cm/s), second (mean ba-PWV: 1524 -1785 cm/s), third (mean ba-PWV: 1786-2084 cm/s), and fourth (mean ba-PWV ≥ 2088 cm/s) quartile groups (A), and in the first to third (mean ba-PWV ≤ 2084 cm/s) and fourth (mean ba-PWV ≥ 2088 cm/s) quartile groups (B).

(C, D) Incident dementia according to the ba-PWV quartile groups in 422 patients without atrial fibrillation.

Tables 3 and Supplemental Tables 2 and 3 show the Cox proportional hazards regression analysis for dementia. The risk of incident dementia was significantly higher only in the fourth ba-PWV group than in the first ba-PWV group in the crude analysis; however, the significance disappeared after adjusting for age and sex. Owing to the small number of events in each group, we combined the first, second, and third groups and used them as references. The risk of incident dementia was significantly higher in the highest ba-PWV group than the first to third ba-PWV groups used as references after adjusting for the age, risk factors, mean blood pressure, WMH severity, lacunae, and baseline MoCA-J (Table 3; adjusted HR: 3.40; 95% CI: 1.24-9.34, P=0.018). The results were similar when we used the mean values of the right and left ba-PWV in the analysis (Supplemental Table 2). However, in 422 patients without atrial fibrillation, the significance between the first and third ba-PWV groups and the highest ba-PWV group disappeared after adjusting for age and sex (Supplemental Table 3).

Table 3.A Cox regression hazard analysis for incident dementia

	Number	Event	Crude HR	Model 1 Adjusted HR	Model 2 Adjusted HR	Model 3 Adjusted HR
ba-PWV 1st quartile (979-1525 cm/sec)	119	1	1.00 (Reference)	1.00 (Reference)	1.00 (Reference)	1.00 (Reference)
ba-PWV 2nd quartile (1526-1786 cm/sec)	120	2	2.16 (0.20-23.86)	1.29 (0.11-14.32)	1.08 (0.09-12.58)	0.86 (0.07-10.70)
			P=0.529	P=0.838	P=0.948	P=0.908
ba-PWV 3rd quartile (1788-2099 cm/sec)	120	6	6.02 (0.72-50.09)	2.68 (0.31-23.00)	2.03 (0.22-18.29)	1.97 (0.20-19.11)
			P=0.097	P=0.369	P=0.530	P=0.557
ba-PWV 4th quartile (2102-4663 cm/sec)	119	14	15.83 (2.08-120.46)	5.16 (0.63-41.89)	4.28 (0.47-39.05)	4.80 (0.52-44.47)
			P=0.008	P=0.125	P=0.198	P=0.167
ba-PWV 1st to 3rd quartile (979-2099 cm/sec)	359	9	1.00 (Reference)	1.00 (Reference)	1.00 (Reference)	1.00 (Reference)
ba-PWV 4th quartile (2102-4663 cm/sec)	119	14	5.14 (2.22-11.90)	2.66 (1.10-6.41)	2.64 (1.01-7.03)	3.40 (1.24-9.34)
			P＜0.001	P=0.029	P=0.049	P=0.018

Model 1: Adjusted for age and sex; Model 2: Adjusted for model 1 and hypertension, MBP, diabetes mellitus, atrial fibrillation, chromic kidney disease, a history of stroke or TIA; Model 3: Adjusted for model 2, WMH and lacunae, and MoCA-J.

ba-PWV, brachial-ankle pulse wave velocity; MBP, mean arterial blood pressure at ba-PWV measurement; TIA, transient ischemic attack; WMH, white matter hyperintensities; MoCA-J, Japanese version of the Montreal Cognitive Assessment

Supplemental Table 2.Cox regression hazard analysis for association between mean ba-PWV and the risk of incident dementia

	Number	Event	Crude HR	Model 1 Adjusted HR	Model 2 Adjusted HR	Model 3 Adjusted HR
mean ba-PWV 1st quartile	120		1.00 (Reference)	1.00 (Reference)	1.00 (Reference)	1.00 (Reference)
(978 – 1520 cm/sec)
mean ba-PWV 2nd quartile	119	2	2.17 (0.20-23.91)	1.21 (0.11-13.62)	0.86 (0.07-10.30)	0.91 (0.07-11.00)
(1524 – 1785 cm/sec)			P = 0.527	P = 0.877	P = 0.908	P = 0.940
mean ba-PWV 3rd quartile	119	5	5.37 (0.63-46.00)	2.20 (0.14-14.37)	1.44 (0.15-14.37)	1.64 (0.16-16.74)
(1786 – 2084 cm/sec)			P = 0.097	P = 0.756	P = 0.756	P = 0.675
mean ba-PWV 4th quartile	120	15	17.12 (2.26-129.74)	5.31 (0.64-44.04)	4.21 (0.43-41.18)	4.51 (0.46-43.91)
(2088 – 4563 cm/sec)			P = 0.006	P = 0.122	P = 0.216	P = 0.194
mean ba-PWV 1st to 3rd	358	8	1.00 (Reference)	1.00 (Reference)	1.00 (Reference)	1.00 (Reference)
quartile (978 - 2084 cm/sec)
mean ba-PWV 4th quartile	120	15	6.11 (2.58-14.43)	3.17 (1.29-7.81)	3.46 (1.25-9.58)	3.40 (1.20-9.62)
(2088 - 4563 cm/sec)			P＜0.001	P = 0.012	P = 0.017	P = 0.021

Model 1: Adjusted for age and sex, Model 2: Adjusted for model 1 and hypertension, MBP, diabetes mellitus, atrial fibrillation, chromic

kidney disease, a history of stroke or TIA, Model 3: Adjusted for model 2, and WMH and lacune, and MoCA-J. Abbreviations: ba-PWV = brachial-ankle pulse wave velocity, MBP = mean arterial blood pressure at ba-PWV measurement, TIA = transient ischemic attack, WMH = white matter hyperintensities, MoCA-J = Japanese version of the Montreal Cognitive Assessment

Supplemental Table 3.Cox regression hazard analysis for incident dementia in 422 patients without atrial fibrillation

	Number	Event	Crude HR	Model 1 Adjusted HR	Model 2 Adjusted HR	Model 3 Adjusted HR
ba-PWV 1st quartile	114		1.00 (Reference)	1.00 (Reference)	1.00 (Reference)	1.00 (Reference)
(979 – 1525 cm/sec)
ba-PWV 2nd quartile	101	2	2.44 (0.22-26.93)	1.49 (0.13-16.86)	1.64 (0.14-19.38)	2.22 (0.18-27.36)
(1526 – 1786 cm/sec)			P= 0.466	P= 0.748	P= 0.694	P= 0.532
ba-PWV 3rd quartile	102	3	3.56 (0.37-34.28)	1.49 (0.14-15.34)	1.63 (0.14-18.68)	2.29 (0.19-27.00)
(1788 – 2099 cm/sec)			P= 0.271	P= 0.739	P= 0.693	P= 0.510
ba-PWV 4th quartile	105	10	12.18 (1.56-95.27)	3.97 (0.45-35.49)	4.29 (0.37-49.84)	5.33 (0.44-64.76)
(2102 – 4663 cm/sec)			P= 0.017	P= 0.217	P= 0.244	P= 0.189
ba-PWV 1st to 3rd quartile	317	6	1.00 (Reference)	1.00 (Reference)	1.00 (Reference)	1.00 (Reference)
(979 - 2099 cm/sec )
ba-PWV 4th quartile	105	10	5.40 (1.96-14.90)	2.84 (0.96-8.37)	2.74 (0.79-9.51)	2.55 (0.69-9.42)
(2102 - 4663 cm/sec)			P= 0.001	P= 0.058	P= 0.112	P= 0.159

Model 1: Adjusted for age and sex, Model 2: Adjusted for model 1 and hypertension, MBP, diabetes mellitus, atrial fibrillation, chromic

kidney disease, a history of stroke or TIA, Model 3: Adjusted for model 2, and WMH and lacune, and MoCA-J. Abbreviations: ba-PWV= brachial-ankle pulse wave velocity, MBP = mean arterial blood pressure at ba-PWV measurement, TIA = transient ischemic attack, WMH = white matter hyperintensities, MoCA-J = Japanese version of the Montreal Cognitive Assessment

In 192 patients who underwent the second MMSE and MoCA-J examinations 3 years later, the ba-PWV was negatively related to changes in MoCA-J (r=-0.184, P=0.011) but not to changes in MMSE (Fig.2C and 2D). Differences in the changes in MoCA-J among the four groups were significant (Table 4A). Statistical significance disappeared after adjusting for age and sex but reappeared after adjusting for the age, sex, risk factors, mean blood pressure, baseline MoCA-J and WMH severity, and presence of lacunae (Table 4A). Comparisons of changes in MoCA-J scores between the first to third-quartile group and fourth-quartile group revealed that the decline in MoCA-J was higher in the fourth-quartile group than in the other 3 groups, even after adjustment (P=0.002) (Table 4B). Among 192 patients, 5 showed a decline of more than 10 points in the MoCA-J score during the 3-year follow-up. Among these five patients, two had a history of stroke, and one experienced cardiogenic embolism during the follow-up period.

Table 4.Adjusted change in the MoCA-J score over three years among the ba-PWV-quartile groups

A	Unadjusted	Adjusted model 1	Adjusted model 2
ba-PWV 1st quartile	0.38±0.54	-0.33±0.59	1.38±0.84
ba-PWV 2nd quartile	-0.35±0.54	-0.39±0.54	0.90±0.69
ba-PWV 3rd quartile	0.17±0.54	0.22±0.53	1.24±0.65
ba-PWV 4th quartile	-1.96±0.54	-1.67±0.56	-1.07±0.75
	p= 0.011	p= 0.091	P= 0.022
B
ba-PWV 1st to 3rd quartile	-0.06 ±0.31	-0.16±0.31	1.15±0.53
ba-PWV 4th quartile	-1.96 ±1.54	-1.69±0.55	-1.01±0.74
	p= 0.001	p= 0.017	p= 0.002

Model 1: Adjusted for age and sex

Model 2: Adjusted for model 1 and hypertension, MBP, diabetes mellitus, atrial fibrillation, chronic kidney disease, history of stroke or TIA, MoCA-J, WMH, and lacunae

ba-PWV, brachial-ankle pulse wave velocity; MBP, mean arterial blood pressure at ba-PWV measurement; TIA, transient ischemic attack; MoCA-J, Japanese version of the Montreal Cognitive Assessment; WMH, white matter hyperintensities.

Discussion

This study revealed that a marker of arterial stiffness, ba-PWV, predicted incident dementia independently of the risk factors, baseline cognition, and SVD severity. The strengths of our study are as follows: (1) all patients underwent MRI and MMSE examinations at enrollment, and (2) most patients visited our neurology clinics every 3 months, with approximately 40% undergoing a second cognitive function test after 3 years.

Arterial stiffness is known to be related to age and several vascular risk factors and to predict future vascular events such as myocardial infarction and stroke^{3, 17)}. An increase in arterial stiffness leads to a pressure load to small penetrating arteries in the kidney and brain due to a decrease in the cushion function in large arteries¹⁾. Cerebral circulation leads to an increase in blood-brain permeability, thrombotic occlusion, hemosiderin extravasation, and elongation of penetrating arteries, which cause representative radiological findings in cerebral SVD, such as white matter hyperintensities, lacunar infarction, cerebral microbleeds, and enlarged perivascular space, respectively^{4, 5, 18, 19)}. SVD, especially WMH, is a strong predictor of incident dementia⁶⁾, so arterial stiffness has been suggested to cause cognitive decline and dementia through SVD progression^{2, 20, 21)}. Therefore, the potential presence of an independent association between arterial stiffness and incident dementia should be investigated, considering the cerebral SVD severity on MRI.

Several cross-sectional studies have revealed that ba-PWV is associated with cognitive impairment and dementia^22-24), this association disappeared after adjusting for SVD severity in one study²⁴⁾. Poels et al. conducted a population-based cohort study and found no independent association between ba-PWV and cognitive decline or incident dementia²⁵⁾. Our findings disagree with their findings; however, several aspects could explain this difference. First, they included ≥ 3000 patients with a mean follow-up of 4.4 years; 527 of these patients died, and 156 cases of incident dementia occurred. Those authors followed patients every three years, so several cases of incident dementia may have been overlooked before death. Second, they did not consider MRI findings at baseline. Baseline SVD severity influences the risk of incident dementia, regardless of arterial stiffness. In the Framingham Offspring study, Pase et al. revealed that aortic stiffness was associated with incident mild cognitive impairment and dementia, but the association disappeared after adjusting for risk factors²⁶⁾. The weakness of their study was that they did not consider the baseline cognitive function or MRI findings to be confounding factors for incident dementia. In contrast, several cohort studies have consistently reported a significant association between arterial stiffness and cognitive decline^27-34). We revealed that the MoCA-J score tended to decline progressively in the highest ba-PWV-quartile group compared to the other three groups, even after adjusting for SVD severity. The MoCA was considered a more sensitive marker of cognitive function than the MMSE and was adopted in recent randomized studies for longitudinal cognitive decline³⁵⁾.

Several mechanisms other than SVD progression may underlie the association between arterial stiffness and dementia incidence. Arterial stiffness may accelerate amyloid β deposition³⁶⁾ and increase tau burden³⁷⁾. Vessel stiffness in the venous system may cause jugular vein reflux, which can result in cognitive decline and incident dementia³⁸⁾. Decreased cerebral blood flow and endothelial dysfunction may underlie the association between arterial stiffness and dementia incidence.

Several limitations associated with the present study warrant mention. First, this was a single-center cohort study with a single ethnic population and had a selection bias, such as a lower percentage of coronary artery disease than in cases with a history of stroke because of the requirement of a brain examination and cognitive examination. Thus, the results are not generalizable. Second, the ba-PWV was measured only at enrollment. Changes in the ba-PWV during follow-up were therefore not considered in this study. Third, the small number of dementia incidents did not allow us to analyze the effect of ba-PWV on each subtype of dementia. Fourth, we included 56 patients with atrial fibrillation, in whom the reliability of ba-PWV measurement would diminisihed³⁾. Similar results were obtained in 422 patients without atrial fibrillation in the Kaplan-Meier analysis, although the significance disappeared after adjusting for age and sex in this group, probably because of the small number of events (Supplemental Fig.1 and Supplemental Table 3). Fifth, 7% of the patients were lost to follow-up; thus, we may have missed incidents of dementia in these patients.

Conclusions

The ba-PWV was associated with incident dementia and cognitive decline independent of age, risk factors, baseline cognitive function, and SVD severity in patients with cerebral SVD. Attention and regular check-ups for the cognitive function are recommended for patients with high ba-PWV values. Furthermore, interventions for the improvement of ba-PWV, such as lowering blood pressure, warrant more attention for cognitive function maintenance and incident dementia prevention.

Acknowledgements

None.

Grant Support

None.

Conflicts of Interest

Dr. Kitagawa received personal fees from Kyowa Kirin and Kowa, grants and personal fees from Daiichi Sankyo, and grants from Bayer and Dainihon Sumitomo outside of the submitted work. The authors declare that they have no conflicts of interest.

References

• 1)  Chirinos JA, Segers P, Hughes T and Townsend R. Large-Artery Stiffness in Health and Disease: JACC State-of-the-Art Review. J Am Coll Cardiol, 2019; 74: 1237-1263
• 2)  Iulita MF, Noriega de la Colina A and Girouard H. Arterial stiffness, cognitive impairment and dementia: confounding factor or real risk? J Neurochem, 2018; 144: 527-548
• 3)  Tomiyama H, Matsumoto C, Shiina K and Yamashina A. Brachial-Ankle PWV: Current Status and Future Directions as a Useful Marker in the Management of Cardiovascular Disease and/or Cardiovascular Risk Factors. J Atheroscler Thromb, 2016; 23: 128-146
• 4)  Henskens LH, Kroon AA, van Oostenbrugge RJ, Gronenschild EH, Fuss-Lejeune MM, Hofman PA, Lodder J and de Leeuw PW. Increased aortic pulse wave velocity is associated with silent cerebral small-vessel disease in hypertensive patients. Hypertension, 2008; 52: 1120-1126
• 5)  Poels MM, Zaccai K, Verwoert GC, Vernooij MW, Hofman A, van der Lugt A, Witteman JC, Breteler MM, Mattace-Raso FU and Ikram MA. Arterial stiffness and cerebral small vessel disease: the Rotterdam Scan Study. Stroke, 2012; 43: 2637-2342
• 6)  Debette S, Schilling S, Duperron MG, Larsson SC and Markus HS. Clinical Significance of Magnetic Resonance Imaging Markers of Vascular Brain Injury: A Systematic Review and Meta-analysis. JAMA Neurol, 2019; 76: 81-94
• 7)  Kubota M, Iijima M, Shirai Y, Toi S and Kitagawa K. Association Between Cerebral Small Vessel Disease and Central Motor Conduction Time in Patients with Vascular Risk. J Stroke Cerebrovasc Dis, 2019; 28: 2343-2350
• 8)  Hosoya M, Toi S, Seki M, Saito M, Hoshino T, Yoshizawa H and Kitagawa K. Association between total cerebral small vessel disease score and cognitive function in patients with vascular risk factors. Hypertens Res, 2023; 46: 1326-1334
• 9)  Hughes CP, Berg L, Danziger WL, Coben LA and Martin RL. A new clinical scale for the staging of dementia. Br J Psychiatry, 1982; 140: 566-572
• 10)  Duering M, Biessels GJ, Brodtmann A, Chen C, Cordonnier C, de Leeuw FE, Debette S, Frayne R, Jouvent E, Rost NS, Ter Telgte A, Al-Shahi Salman R, Backes WH, Bae HJ, Brown R, Chabriat H, De Luca A, deCarli C, Dewenter A, Doubal FN, Ewers M, Field TS, Ganesh A, Greenberg S, Helmer KG, Hilal S, Jochems ACC, Jokinen H, Kuijf H, Lam BYK, Lebenberg J, MacIntosh BJ, Maillard P, Mok VCT, Pantoni L, Rudilosso S, Satizabal CL, Schirmer MD, Schmidt R, Smith C, Staals J, Thrippleton MJ, van Veluw SJ, Vemuri P, Wang Y, Werring D, Zedde M, Akinyemi RO, Del Brutto OH, Markus HS, Zhu YC, Smith EE, Dichgans M and Wardlaw JM. Neuroimaging standards for research into small vessel disease-advances since 2013. Lancet Neurol, 2023; 22: 602-618
• 11)  Staals J, Makin SD, Doubal FN, Dennis MS and Wardlaw JM. Stroke subtype, vascular risk factors, and total MRI brain small-vessel disease burden. Neurology, 2014; 83: 1228-1234
• 12)  Ishizuka K, Hoshino T, Shimizu S, Shirai Y, Mizuno S, Toi S, Maruyama K, Uchiyama S and Kitagawa K. Brachial-ankle pulse wave velocity is associated with 3-month functional prognosis after ischemic stroke. Atherosclerosis, 2016; 255: 1-5
• 13)  Miwa K, Tanaka M, Okazaki S, Furukado S, Yagita Y, Sakaguchi M, Mochizuki H and Kitagawa K. Chronic kidney disease is associated with dementia independent of cerebral small-vessel disease. Neurology, 2014; 82: 1051-1057
• 14)  Regier DA, Kuhl EA and Kupfer DJ. The DSM-5: Classification and criteria changes. World Psychiatry, 2013; 12: 92-98
• 15)  McKhann GM, Knopman DS, Chertkow H, Hyman BT, Jack CR Jr, Kawas CH, Klunk WE, Koroshetz WJ, Manly JJ, Mayeux R, Mohs RC, Morris JC, Rossor MN, Scheltens P, Carrillo MC, Thies B, Weintraub S and Phelps CH. The diagnosis of dementia due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement, 2011; 7: 263-269
• 16)  Gorelick PB, Scuteri A, Black SE, Decarli C, Greenberg SM, Iadecola C, Launer LJ, Laurent S, Lopez OL, Nyenhuis D, Petersen RC, Schneider JA, Tzourio C, Arnett DK, Bennett DA, Chui HC, Higashida RT, Lindquist R, Nilsson PM, Roman GC, Sellke FW, Seshadri S; American Heart Association Stroke Council, Council on Epidemiology and Prevention, Council on Cardiovascular Nursing, Council on Cardiovascular Radiology and Intervention, and Council on Cardiovascular Surgery and Anesthesia. Vascular contributions to cognitive impairment and dementia: a statement for healthcare professionals from the american heart association/american stroke association. Stroke, 2011; 42: 2672-2713
• 17)  Mitchell GF, Hwang SJ, Vasan RS, Larson MG, Pencina MJ, Hamburg NM, Vita JA, Levy D and Benjamin EJ. Arterial stiffness and cardiovascular events: the Framingham Heart Study. Circulation, 2010; 121: 505-511
• 18)  Ohmine T, Miwa Y, Yao H, Yuzuriha T, Takashima Y, Uchino A, Takahashi-Yanaga F, Morimoto S, Maehara Y and Sasaguri T. Association between arterial stiffness and cerebral white matter lesions in community-dwelling elderly subjects. Hypertens Res, 2008; 31: 75-81
• 19)  Riba-Llena I, Jiménez-Balado J, Castañé X, Girona A, López-Rueda A, Mundet X, Jarca CI, Álvarez-Sabin J, Montaner J and Delgado P. Arterial Stiffness Is Associated With Basal Ganglia Enlarged Perivascular Spaces and Cerebral Small Vessel Disease Load. Stroke, 2018; 49: 1279-1281
• 20)  van Sloten TT, Protogerou AD, Henry RM, Schram MT, Launer LJ and Stehouwer CD. Association between arterial stiffness, cerebral small vessel disease and cognitive impairment: A systematic review and meta-analysis. Neurosci Biobehav Rev, 2015; 53: 121-130
• 21)  Li X, Lyu P, Ren Y, An J and Dong Y. Arterial stiffness and cognitive impairment. J Neurol Sci, 2017; 380: 1-10
• 22)  Hanon O, Haulon S, Lenoir H, Seux ML, Rigaud AS, Safar M, Girerd X and Forette F. Relationship between arterial stiffness and cognitive function in elderly subjects with complaints of memory loss. Stroke, 2005; 36: 2193-2197
• 23)  Meyer ML, Palta P, Tanaka H, Deal JA, Wright J, Knopman DS, Griswold ME, Mosley TH and Heiss G. Association of Central Arterial Stiffness and Pressure Pulsatility with Mild Cognitive Impairment and Dementia: The Atherosclerosis Risk in Communities Study-Neurocognitive Study (ARIC-NCS). J Alzheimers Dis, 2017; 57: 195-204
• 24)  Robert C, Ling LH, Tan ESJ, Gyanwali B, Venketasubramanian N, Lim SL, Gong L, Berboso JL, Richards AM, Chen C and Hilal S. Effects of Carotid Artery Stiffness on Cerebral Small-Vessel Disease and Cognition. J Am Heart Assoc, 2022; 11: e027295
• 25)  Poels MM, van Oijen M, Mattace-Raso FU, Hofman A, Koudstaal PJ, Witteman JC and Breteler MM. Arterial stiffness, cognitive decline, and risk of dementia: the Rotterdam study. Stroke, 2007; 38: 888-892
• 26)  Pase MP, Beiser A, Himali JJ, Tsao C, Satizabal CL, Vasan RS, Seshadri S, and Mitchell GF. Aortic Stiffness and the Risk of Incident Mild Cognitive Impairment and Dementia. Stroke, 2016; 47: 2256-2261
• 27)  Waldstein SR, Rice SC, Thayer JF, Najjar SS, Scuteri A and Zonderman AB. Pulse pressure and pulse wave velocity are related to cognitive decline in the Baltimore Longitudinal Study of Aging. Hypertension, 2008; 51: 99-104
• 28)  Zeki Al Hazzouri A, Newman AB, Simonsick E, Sink KM, Sutton Tyrrell K, Watson N, Satterfield S, Harris T, Yaffe K and Health ABC Study. Pulse wave velocity and cognitive decline in elders: the Health, Aging, and Body Composition study. Stroke, 2013; 44: 388-393
• 29)  Taniguchi Y, Fujiwara Y, Nofuji Y, Nishi M, Murayama H, Seino S, Tajima R, Matsuyama Y and Shinkai S. Prospective Study of Arterial Stiffness and Subsequent Cognitive Decline Among Community-Dwelling Older Japanese. J Epidemiol, 2015; 25: 592-599
• 30)  Scuteri A, Tesauro M, Guglini L, Lauro D, Fini M and Di Daniele N. Aortic stiffness and hypotension episodes are associated with impaired cognitive function in older subjects with subjective complaints of memory loss. Int J Cardiol, 2013; 169: 371-377
• 31)  Yamamoto N, Yamanaka G, Ishikawa M, Takasugi E, Murakami S, Yamanaka T, Ishine M, Matsubayashi K, Hanafusa T and Otsuka K. Cardio-ankle vascular index as a predictor of cognitive impairment in community-dwelling elderly people: four-year follow-up. Dement Geriatr Cogn Disord, 2009; 28: 153-158
• 32)  Tsao CW, Himali JJ, Beiser AS, Larson MG, DeCarli C, Vasan RS, Mitchell GF and Seshadri S. Association of arterial stiffness with progression of subclinical brain and cognitive disease. Neurology, 2016; 86: 619-626
• 33)  Yukutake T, Yamada M, Fukutani N, Nishiguchi S, Kayama H, Tanigawa T, Adachi D, Hotta T, Morino S, Tashiro Y, Aoyama T and Arai H. Arterial Stiffness Predicts Cognitive Decline in Japanese Community-dwelling Elderly Subjects: A One-year Follow-up Study. J Atheroscler Thromb, 2015; 22: 637-644
• 34)  Araghi M, Shipley MJ, Wilkinson IB, McEniery CM, Valencia-Hernández CA, Kivimaki M, Sabia S, Singh-Manoux A and Brunner EJ. Association of aortic stiffness with cognitive decline: Whitehall II longitudinal cohort study. Eur J Epidemiol, 2020; 35: 861-869
• 35)  Cukierman-Yaffe T, Gerstein HC, Colhoun HM, Diaz R, García-Pérez LE, Lakshmanan M, Bethel A, Xavier D, Probstfield J, Riddle MC, Rydén L, Atisso CM, Hall S, Rao-Melacini P, Basile J, Cushman WC, Franek E, Keltai M, Lanas F, Leiter LA, Lopez-Jaramillo P, Pirags V, Pogosova N, Raubenheimer PJ, Shaw JE, Sheu WH and Temelkova-Kurktschiev T. Effect of dulaglutide on cognitive impairment in type 2 diabetes: an exploratory analysis of the REWIND trial. Lancet Neurol, 2020; 19: 582-590
• 36)  Hughes TM, Wagenknecht LE, Craft S, Mintz A, Heiss G, Palta P, Wong D, Zhou Y, Knopman D, Mosley TH and Gottesman RF. Arterial stiffness and dementia pathology: Atherosclerosis Risk in Communities (ARIC)-PET Study. Neurology, 2018; 90: e1248-e1256
• 37)  Cooper LL, O’Donnell A, Beiser AS, Thibault EG, Sanchez JS, Benjamin EJ, Hamburg NM, Vasan RS, Larson MG, Johnson KA, Mitchell GF and Seshadri S. Association of Aortic Stiffness and Pressure Pulsatility With Global Amyloid-β and Regional Tau Burden Among Framingham Heart Study Participants Without Dementia. JAMA Neurol, 2022; 79: 710-719
• 38)  Adachi U, Toi S, Hosoya M, Hoshino T, Seki M, Yoshizawa H, Tsutsumi Y, Maruyama K and Kitagawa K. Association of Age-Related Spontaneous Internal Jugular Vein Reflux with Cognitive Impairment and Incident Dementia. J Alzheimers Dis, 2023; 96: 1221-1230