Sex Differences in the Relationship between Arterial Stiffness and Incidence of Chronic Kidney Disease

Abstract

Aims: There is a lack of evidence regarding the sex-specific impact of arterial stiffness on the incidence of chronic kidney disease (CKD). This study assessed the relationship between arterial stiffness based on brachial-ankle pulse wave velocity (baPWV) and incident CKD in men and women.

Methods: Individuals who participated in health checkups and underwent concomitant baPWV measurement between 2006 and 2019 were included. They were free of CKD at baseline. The participants were categorized into 4 groups based on their baPWV values (cm/s) as follows: ＜1,200 cm/s for normal, ≥ 1,200 and ＜1,400 for high normal, ≥ 1,400 and ＜1,800 for borderline, and ≥ 1,800 cm/s. The primary outcome was CKD development (estimated glomerular filtration rate ＜60 mL/min/1.73 m²).

Results: A total of 130,100 participants were enrolled, with a mean age of 40.5±8.2 years old. During the mean of 5.6 years of follow-up, 906 (0.7%) participants developed incident CKD. The cumulative incidence of CKD was 0.3%, 0.5%, 1.4%, and 6.2% in the normal, high normal, borderline, and abnormal groups, respectively. In the multivariable-adjusted model including systolic blood pressure, compared with the normal baPWV group, abnormal baPWV group demonstrated a significantly increased risk of incident CKD in women. However, among men, any other baPWV groups were not associated with a significantly elevated risk of incident CKD.

Conclusions: Increased arterial stiffness, as measured by baPWV, was associated with an increased risk of incident CKD, with notable sex-specific differences. These findings underscore the utility of baPWV for identifying CKD risk in women and offer valuable insights into sex-specific differences in arterial stiffness and CKD development.

Introduction

Arterial stiffness not only accompanies aging and various pathological states¹⁾ but also contributes to target organ damage, including heart failure, impaired coronary perfusion, chronic kidney disease (CKD), and cerebrovascular disease²⁾. Patients with CKD have a strong risk of cardiovascular disease independent of traditional cardiovascular risk factors³⁾. Therefore, early detection and appropriate medical attention for individuals with CKD are crucial for improving the cardiovascular prognosis.

Brachial-ankle pulse wave velocity (baPWV), measured by wrapping pressure cuffs around the four extremities, is a straightforward marker for assessing the stiffness of medium- to large-sized arteries⁴⁾. Despite controversies regarding the association between the baPWV and a decreased renal function^{5, 6)}, increased arterial stiffness, as measured by baPWV, has been observed in the early stages of CKD^{7, 8)}. In addition, considering the substantial sex differences in the development of cardiovascular disease and arterial stiffness⁹⁾, there is a lack of evidence about the sex-specific impact of arterial stiffness on the incidence of CKD.

The present study therefore evaluated the sex differences in the relationship between arterial stiffness based on the baPWV and incident CKD in a large health checkup-based cohort.

Methods

Study Population

This cohort study, part of the Kangbuk Samsung Health Study, enrolled Korean adults ≥ 18 years old who underwent annual or biennial health examinations in Seoul or Suwon, South Korea. The majority (＞80%) of participants or their spouses were employees of companies and local governmental organizations. In South Korea, the Industrial Safety and Health Law mandates annual or biennial health screening examinations for all employees be provided free of charge. The remaining participants voluntarily underwent screening examinations at their own expense. The study included participants who underwent baPWV measurement as part of their health examination from 2006 to 2019 and had undergone at least 1 follow-up until December 2019 (n=139,245).

We excluded 9,145 participants who met ≥ 1 of the following exclusion criteria at baseline: 196 had missing data on the estimated glomerular filtration rate (eGFR) at baseline or follow-up; 258 had a missing baseline history of cardiovascular disease, cancer, or kidney disease; 6294 had a history of cardiovascular disease or malignancy; 2013 had a history of kidney disease based on a questionnaire; and 750 had a baseline eGFR ＜60 mL/min/1.73 m². After these exclusions, a total of 130,100 participants were included in the analysis (Fig.1).

Fig.1.

Flowchart of the study participants

Ethics Approval

This study was approved by the Institutional Review Board of Kangbuk Samsung Hospital (KBSMC 2021-10-005), and written informed consent was obtained from all participants.

Data Collection

Information on the medical and family history, medication use, lifestyle factors, and education level was obtained using a self-reported questionnaire. Anthropometric and serum biochemical parameters were measured by trained staff at baseline and follow-up visits. Current smokers were defined as those who currently smoked and reported having smoked ＞100 cigarettes in their lifetimes. Average alcohol consumption per day was calculated using the frequency and amount of alcohol consumed per day. The physical activity level was assessed using the validated Korean version of the International Physical Activity Questionnaire Short Form and categorized as inactive, active, or health-enhancing.

Blood pressure (BP) was measured by trained nurses using an automated oscillometric device (53000; Welch Allyn, New York, NY, USA) with the participants in a sitting position and the arm supported at the level of the heart. Three consecutive BP readings were obtained after the participants rested quietly in a sitting position for 5 min. The average of the second and third BP readings was used in the analysis.

Laboratory Assessments

The following blood parameters were assessed after at least 10 h of fasting: plasma glucose, hemoglobin A1c, total cholesterol, and high-density lipoprotein cholesterol. Diabetes was diagnosed based on fasting plasma glucose levels of ≥ 126 mg/dL, glycated hemoglobin A1c levels of ≥ 6.5%, or the current use of antidiabetic drugs. Serum creatinine levels were analyzed using the Jaffe method. We calculated the eGFR using the Chronic Kidney Disease Epidemiology Collaboration equation as follows: eGFR (mL/min/1.73 m²)=141×min(Scr/κ, 1)^α×max(Scr/κ, 1)^−1.209×0.993^Age×1.018 [if female]×1.159 [if black], where Scr is serum creatinine, κ is 0.7 for women and 0.9 for men, and α is -0.329 for women and -0.411 for men¹⁰⁾.

baPWV Measurements

baPWV was measured in the supine position using a VP-1000 vascular profiler (Omron Healthcare, Kyoto, Japan). This device recorded pulse waveforms of the bilateral brachial and posterior tibial arteries using an oscillometric method. Cuffs were wrapped around the brachia and ankles, and electrocardiographic electrodes were placed on both wrists. Concurrently, a phonocardiogram was employed to record the acoustic signals of the heart, providing a reference for the onset of the cardiac cycle. The path lengths from the suprasternal notch to the brachium (Lb) and from the suprasternal notch to the ankle (La) were estimated based on the participant’s height. The baPWV was then calculated using the following formula: baPWV=(La−Lb) / Tba, where Tba is the time interval between the wavefront of the brachial waveform and that of the ankle waveform. Average values from the left and right sides were used in the analysis. Guideline proposed baPWV cutoff values of 1,400 and 1,800 cm/sec (＜1,400 for normal, ≥ 1,400 and ＜1,800 for borderline, ≥ 1,800 for abnormal)¹¹⁾. In this study, about 77% of the study population had a normal baPWV value. Thus, we divided them into two groups: ＜1,200 and ≥ 1,200 cm/sec value. Finally, the study population was categorized into 4 groups based on the baPWV as follows: ＜1,200 for normal, ≥ 1,200 and ＜1,400 for high normal, ≥ 1,400 and ＜1,800 for borderline, and ≥ 1,800 cm/sec for abnormal.

Study Outcome

The primary outcome was CKD development during follow-up, defined as an eGFR less than 60 mL/min/1.73 m^{2 12)}.

Statistical Analyses

Descriptive statistics were employed to summarize the characteristics of the participants, who were categorized into four groups based on their baseline baPWV. Incident CKD was defined as the primary endpoint.

Since follow-up began with all participants free of CKD at baseline, we could establish the first visit at which a participant showed CKD; however, we could not determine the precise time of outcome development (which occurred at some point between the first visit with CKD and the previous visit). To account for interval censoring, we used a parametric proportional hazards model with the ‘stpm’ command in Stata¹³⁾. We calculated adjusted hazard ratios and their 95% confidence intervals for incident CKD, comparing each increased baPWV group to the normal baPWV group (used as the reference). In these models, the baseline hazard function was parameterized using restricted cubic splines on the logarithmic scale of time with four degrees of freedom.

Statistical models were initially adjusted for the age and sex and subsequently adjusted for the year of the screening examination, center, fasting plasma glucose, body mass index, total cholesterol, high-density lipoprotein, medication for hypertension, medication for diabetes, medication for hyperlipidemia, alcohol consumption, smoking, physical activity, education level, family history of coronary heart disease (Model 1), baseline eGFR (Model 2), and systolic BP (Model 3).

Stratified analyses were conducted according to sex and age (＜40 vs. ≥ 40 years old). Likelihood ratio tests, comparing models with and without multiplicative interaction terms, were used to test the effects of interactions between the subgroups and baPWV groups on incident CKD. A Kaplan–Meier analysis and log-rank tests were performed to assess the cumulative incidence of CKD. Both univariate and multivariate parametric proportional models were used to evaluate the predictors of incident CKD within the strata of different sexes.

All statistical analyses were conducted using the STATA program, version 18.0 (Stata Corp LP, College Station, TX, USA). Statistical significance was set at P＜0.05.

Results

Baseline Characteristics

The mean age of study participants was 40.5±8.2 years old, and men were dominant (62.9%). Approximately 77% of study participants were included in the normal or high normal groups, while only 1,516 (1.2%) were in the abnormal baPWV group. Compared to the normal baPWV group (baPWV ＜1,200 cm/s), individuals in the elevated baPWV groups were older and had a higher prevalence of medications for hypertension, diabetes, or hyperlipidemia, and a lower prevalence of high education level in both men and women (Table 1). The prevalence of obesity increased with higher baPWV levels exclusively in women, whereas the prevalence of alcohol intake increased with higher baPWV levels in men (Table 1). Across both sexes, populations in the higher baPWV groups exhibited elevated BP, heart rate, glucose, glycated hemoglobin, and total cholesterol levels compared to those in the normal baPWV group (Table 1). When comparing the baseline characteristics by age and/or sex, baPWV levels increased with age and were higher in men than in women (Table 2). Men ＜40 years old had significantly higher baPWV values than women ≥ 40 years old (Table 3).

Table 1.Baseline characteristics based on the baPWV (cm/sec)

Characteristics	Overall	Normal (＜1,200)	High normal (≥ 1,200 and＜1,400)	Borderline (≥ 1,400 and＜1,800)	Abnormal (≥ 1,800)	P for trend
Women
Number (%)	48,229 (100.0)	24,716 (51.2)	18,071 (37.5)	5,039 (10.4)	403 (0.8)
Age (years) a	39.4 (8.3)	36.5 (6.2)	40.5 (7.4)	48.4 (9.6)	62.2 (9.2)	＜0.001
Obesity (%) b	14.2	10.2	14.8	29.6	40.9	＜0.001
Current smoker (%)	2.0	1.9	2.0	2.0	1.2	0.026
Alcohol intake (%) c	5.2	5.1	5.5	4.7	2.7	＜0.001
Regular exercise (%) d	12.2	10.8	12.9	16.4	18.6	0.001
High education level (%) e	60.6	66.2	57.7	43.4	23.3	＜0.001
Medication for hypertension (%)	3.3	0.4	2.8	16.1	38.5	＜0.001
Medication for diabetes (%)	0.9	0.3	0.7	3.8	12.2	＜0.001
Medication for hyperlipidemia (%)	1.9	0.6	1.9	7.5	14.7	＜0.001
Family history of CHD (%)	6.9	6.2	7.4	9.0	8.9	＜0.001
eGFR (CKD–EPI)	105.1 (13.8)	107.9 (13.0)	103.9 (13.5)	97.0 (14.2)	86.3 (12.7)	0.001
Body mass index (kg/m2)	21.9 (3.1)	21.5 (2.8)	22.1 (3.1)	23.6 (3.6)	24.4 (3.2)	＜0.001
Systolic BP (mmHg) a	106.0 (12.5)	101.5 (9.9)	107.9 (11.3)	119.4 (13.7)	132.4 (15.1)	＜0.001
Diastolic BP (mmHg) a	67.0 (8.7)	64 (7.0)	68.4 (8.1)	75.7 (9.6)	79.3 (11.5)	＜0.001
Heart rate (per min)	66.0 (9.0)	64.8 (8.2)	66.8 (9.2)	69.1 (10.4)	71.2 (11.5)	＜0.001
Fasting plasma glucose (mg/dl) a	91.4 (11.7)	89.2 (8.4)	92.0 (11.4)	98.4 (18.6)	108.6 (27.2)	＜0.001
Glycated hemoglobin (mg/dl) a	5.6 (0.4)	5.5 (0.3)	5.6 (8.4)	5.8 (0.7)	6.1 (0.9)	＜0.001
Total cholesterol (mg/dl) a	189.0 (32.9)	183.6 (30.7)	191.8 (32.6)	204.3 (36.7)	207.6 (42.4)	＜0.001
HDL–C (mg/dl) a	64.3 (14.8)	65.5 (14.5)	63.9 (14.8)	60.8 (25.2)	57.6 (16.0)	＜0.001
Men
Number (%)	81,871 (100.0)	14,338 (17.5)	43,294 (52.9)	23,126 (28.2)	1,113 (1.4)
Age (years) a	41.1 (8.2)	38.6 (6.9)	40.3 (7.3)	43.4 (8.9)	55.9 (10.9)	＜0.001
Obesity (%)b	41.7	43.2	40.6	43.0	41.7	0.832
Current smoker (%)	36.5	36.0	36.5	37.2	28.7	＜0.001
Alcohol intake (%)c	35.7	31.2	35.1	39.6	39.7	＜0.001
Regular exercise (%)d	14.9	15.0	14.8	14.7	21.2	＜0.001
High education level (%)e	79.2	83.7	80.4	75.0	57.5	＜0.001
Medication for hypertension (%)	7.5	3.8	6.0	11.5	29.1	＜0.001
Medication for diabetes (%)	2.7	1.0	2.0	4.4	13.3	＜0.001
Medication for hyperlipidemia (%)	3.5	2.6	3.2	4.7	8.0	0.193
Family history of CHD (%)	6.2	5.4	6.0	7.1	4.6	0.001
eGFR (CKD–EPI)	95.8 (13.3)	97.7 (13.0)	96.4 (13.1)	94.1 (13.5)	85.4 (13.6)	0.308
Body mass index (kg/m2)	24.6 (2.9)	24.7 (3.0)	24.5 (2.9)	24.8 (2.9)	24.7 (2.9)	0.007
Systolic BP (mmHg) a	116.6 (11.9)	110.6 (9.9)	115.3 (10.6)	121.9 (12.2)	134.4 (16.1)	＜0.001
Diastolic BP (mmHg) a	75.4 (9.0)	70.7 (7.5)	74.5 (8.1)	79.5 (9.3)	86.1 (12.6)	＜0.001
Heart rate (per min)	65.1 (9.3)	62.5 (8.5)	64.5 (8.9)	67.5 (9.7)	72.2 (12.1)	＜0.001
Fasting plasma glucose (mg/dl) a	97.4 (16.4)	93.9 (11.1)	96.3 (14.3)	100.9 (20.6)	112.1 (29.2)	＜0.001
Glycated hemoglobin (mg/dl) a	5.6 (0.6)	5.5 (0.4)	5.6 (0.5)	5.7 (0.7)	6.1 (1.0)	＜0.001
Total cholesterol (mg/dl) a	200.1 (34.4)	194.5 (32.8)	199.4 (33.9)	204.5 (35.4)	204.0 (38.1)	＜0.001
HDL–C (mg/dl) a	52.3 (12.5)	52.8 (12.5)	52.5 (12.5)	51.8 (12.5)	52.3 (14.0)	＜0.001

Data are expressed as the ^a mean (standard deviation), or percentage.

^b ≥ 25 kg/m²; ^c ≥ 20 g/day; ^d ≥ 3 times/week; ^e ≥ college graduate.

baPWV, brachial-ankle pulse wave velocity; BP, blood pressure; CHD, coronary heart disease; CKD, chronic kidney disease Epidemiology Collaboration; eGFR, estimated glomerular filtration rate; HDL–C, high-density lipoprotein cholesterol.

Table 2.Baseline characteristics according to age or sex

Characteristics	Women	Men	＜40 years old	≥ 40 years old
Number (%)	48,229 (37.1)	81,871 (62.9)	63,432 (48.8)	66,668 (51.2)
Age (years) a	39.4 (8.3)	41.1 (8.2)	33.9 (3.5)	46.7 (6.4)
Men (%)	0.0	100.0	36,581(57.7)	45,290 (67.9)
Obesity (%) c	14.2	41.7	27.5	35.3
Current smoker (%)	2.0	36.5	22.4	25.5
Alcohol intake (%) d	5.2	35.7	22.4	29.4
Regular exercise (%) e	12.2	14.9	10.7	17
High education level (%) f	60.6	79.2	75.4	68.7
Medication for hypertension (%)	3.3	7.5	1.3	10.3
Medication for diabetes (%)	0.9	2.7	0.5	3.5
Medication for hyperlipidemia (%)	1.9	3.5	0.9	4.9
Family history of CHD (%)	6.9	6.2	5.5	7.4
eGFR (CKD-EPI)	105.1 (13.8)	95.8 (13.3)	104.9 (13.3)	93.9 (13.0)
Body mass index (kg/m2)	21.9 (3.1)	24.6 (2.9)	23.2 (3.4)	24.1 (3)
Systolic BP (mmHg) a	106 (12.5)	116.6 (11.9)	111 (12.8)	114.2 (13.3)
Diastolic BP (mmHg) a	67 (8.7)	75.4 (9)	70.4 (9.4)	74 (9.8)
Heart rate (per min)	66 (9)	65.1 (9.3)	65.7 (9.2)	65.2 (9.2)
Fasting plasma glucose (mg/dl) a	91.4 (11.7)	97.4 (16.4)	92.3 (11.7)	97.9 (17.4)
Glycated hemoglobin (mg/dl) a	5.6 (0.4)	5.6 (0.6)	5.5 (0.4)	5.7 (0.6)
Total cholesterol (mg/dl) a	189 (32.9)	200.1 (34.4)	191.0 (33.3)	200.7 (34.5)
HDL–C (mg/dl) a	64.3 (14.8)	52.3 (12.5)	58.2 (14.7)	55.5 (14.4)
baPWV (cm/sec)b	1,200 (1,110–1,300)	1,320 (1,230–1,420)	1,240 (1,140–1,340)	1,320 (1,220–1,430)

Data are expressed as the ^a mean (standard deviation), ^b median (interquartile range), or percentage.

^c ≥ 25 kg/m²; ^d ≥ 20 g/day; ^e ≥ 3 times per week; ^f ≥ college graduate

baPWV, brachial-ankle pulse wave velocity; BP, blood pressure; CHD, coronary heart disease; CKD, chronic kidney disease Epidemiology Collaboration; eGFR, estimated glomerular filtration rate; HDL–C, high-density lipoprotein cholesterol.

Table 3.Baseline characteristics according to age and sex

Characteristics	Women ＜40 years old	Women ≥ 40 years old	Men ＜40 years old	Men ≥ 40 years old
Number (%)	26,851 (20.6)	21,378 (16.4)	36,581 (28.1)	45,290 (34.8)
Age (years) a	33.7 (3.6)	46.6 (6.7)	34.1 (3.5)	46.7 (6.3)
Obesity (%) c	9.6	20.0	40.7	42.5
Current smoker (%)	2.1	1.7	37.2	36.0
Alcohol intake (%) d	6.1	4.0	32.0	38.7
Regular exercise (%) e	8.5	17.1	12.3	16.9
High education level (%) f	66.6	52.4	82.1	76.7
Medication for hypertension (%)	0.3	7.0	2.0	11.9
Medication for diabetes (%)	0.2	1.8	0.6	4.3
Medication for hyperlipidemia (%)	0.3	3.9	1.3	5.4
Family history of CHD (%)	5.8	8.4	5.2	7.0
eGFR (CKD-EPI)	110.3 (12.4)	98.6 (12.8)	100.9 (12.5)	91.7 (12.5)
Body mass index (kg/m2)	21.3 (2.9)	22.7 (3.1)	24.6 (3.1)	24.7 (2.8)
Systolic BP (mmHg) a	103.6 (10.9)	109.1 (13.6)	116.5 (11.3)	116.7 (12.4)
Diastolic BP (mmHg) a	65.3 (7.7)	69.1 (9.3)	74.2 (8.7)	76.3 (9.2)
Heart rate (per min)	66.2 (8.8)	65.8 (9.2)	65.4 (9.4)	64.9 (9.2)
Fasting plasma glucose (mg/dl) a	89.3 (9.1)	93.9 (13.9)	94.4 (12.8)	99.8 (18.5)
Glycated hemoglobin (mg/dl) a	5.5 (0.3)	5.7 (0.5)	5.5 (0.4)	5.7 (0.6)
Total cholesterol (mg/dl) a	182.7 (30.6)	197.0 (34.0)	197.2 (33.8)	202.4 (34.6)
HDL–C (mg/dl) a	65.5 (14.5)	62.9 (15.0)	52.8 (12.4)	52.0 (12.6)
baPWV (cm/sec) b	1,150 (1,070–1,240)	1,260 (1,160–1,380)	1,300 (1,210–1,390)	1,340 (1,250–1,450)

Data are expressed as the ^a mean (standard deviation), ^b median (interquartile range), or percentage.

^c ≥ 25 kg/m²; ^d ≥ 20 g/day; ^e ≥ 3 times per week; ^f ≥ college graduate

baPWV, brachial-ankle pulse wave velocity; BP, blood pressure; CHD, coronary heart disease; CKD, chronic kidney disease Epidemiology Collaboration; eGFR, estimated glomerular filtration rate; HDL–C, high-density lipoprotein cholesterol.

Incidence of CKD

During a mean of 5.6±3.0 years of follow-up, 906 (0.7%) participants experienced incident CKD. The cumulative incidence of CKD was 102 (0.3%), 321 (0.5%), 389 (1.4%), and 94 (6.2%) in the normal, high normal, borderline, and abnormal groups, respectively. Stratified by sex, the incidences were 82 (0.5%), 272 (0.6%), 339 (1.4%), and 69 (6.1%) in men, and 20 (0.08%), 49 (0.2%), 50 (0.9%), and 25 (6.2%) in women for the normal, high normal, borderline, and abnormal groups, respectively. Incidence rates (/10³ person-years) significantly increased from the normal to abnormal groups in both men and women (Table 4). Fig.2 depicts the cumulative incidence of CKD according to baPWV groups within different sex groups.

Table 4.Development of chronic kidney disease according to baPWV (cm/sec) (N = 130,100)

baPWV	PY	Incident cases	Incidence rates (/103 PY)	Age and sex-adjusted HR (95% CI)	Multivariable-adjusted HRa (95% CI)
					Model 1	Model 2	Model 3
Total
Normal (＜1,200)	224445.3	102	0.5	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
High normal (≥ 1,200 and ＜1,400)	353530.3	321	0.9	1.09 (0.87-1.37)	0.99 (0.79-1.25)	0.98 (0.78-1.23)	0.93 (0.74-1.16)
Borderline (≥ 1,400 and ＜1,800)	150313.9	389	2.6	1.73 (1.37-2.19)	1.39 (1.10-1.77)	1.35 (1.07-1.71)	1.17 (0.91-1.49)
Abnormal (≥ 1,800)	6493.7	94	14.5	2.58 (1.87-3.57)	1.94 (1.39-2.70)	1.91 (1.37-2.66)	1.45 (1.02-2.08)
P for trend				＜0.001	＜0.001	＜0.001	0.009
Age ＜40 years
Normal (＜1,200)	153852.9	18	0.1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
High normal (≥ 1,200 and ＜1,400)	184017.5	55	0.3	1.71 (1.00-2.92)	1.53 (0.89-2.61)	1.49 (0.87-2.54)	1.40 (0.82-2.39)
Borderline (≥ 1,400 and ＜1,800)	56517.9	35	0.6	3.03 (1.71-5.39)	2.42 (1.36-4.30)	2.15 (1.21-3.82)	1.89 (1.06-3.37)
Abnormal (≥ 1,800)	496.0	4	8.1	38.3 (12.93-113.63)	23.55 (7.90-70.22)	11.63 (3.88-34.89)	8.10 (2.65-24.73)
P for trend				＜0.001	0.001	0.002	0.008
Age ≥ 40 years
Normal (＜1,200)	70592.4	84	1.2	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
High normal (≥ 1,200 and ＜1,400)	169512.8	266	1.6	0.93 (0.73-1.19)	0.86 (0.67-1.11)	0.87 (0.68-1.12)	0.82 (0.64-1.05)
Borderline (≥ 1,400 and ＜1,800)	93796.0	354	3.8	1.48 (1.15-1.90)	1.22 (0.95-1.57)	1.20 (0.93-1.55)	1.04 (0.80-1.35)
Abnormal (≥ 1,800)	5997.7	90	15.0	2.35 (1.67-3.29)	1.76 (1.27-2.48)	1.70 (1.20-2.41)	1.30 (0.89-1.88)
P for trend				＜0.001	＜0.001	＜0.001	0.038
Men
Normal (＜1,200)	78544.1	82	1.0	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
High normal (≥ 1,200 and ＜1,400)	248927.6	272	1.1	0.85 (0.67-1.09)	0.82 (0.64–1.05)	0.83 (0.65–1.07)	0.79 (0.62-1.02)
Borderline (≥ 1,400 and ＜1,800)	125606.3	339	2.7	1.36 (1.06-1.75)	1.19 (0.93–1.54)	1.19 (0.93–1.53)	1.04 (0.80-1.36)
Abnormal (≥ 1,800)	4980.9	69	13.9	1.82 (1.28-2.59)	1.54 (1.07–2.21)	1.59 (1.11–2.28)	1.21 (0.82-1.79)
P for trend				＜0.001	＜0.001	＜0.001	0.073
Women
Normal (＜1,200)	145901.2	20	0.1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
High normal (≥ 1,200 and ＜1,400)	104602.7	49	0.5	2.19 (1.30-3.70)	1.90 (1.13–3.21)	1.82 (1.08–3.08)	1.69 (1.00-2.85)
Borderline (≥ 1,400 and ＜1,800)	24707.6	50	2.0	3.55 (2.08-6.04)	2.26 (1.32–3.88)	1.96 (1.14–3.37)	1.63 (0.94-2.82)
Abnormal (≥ 1,800)	1512.8	25	16.5	8.49 (4.58-18.07)	4.21 (2.24–7.91)	3.51 (1.87–6.58)	2.64 (1.39-5.03)
P for trend				＜0.001	＜0.001	＜0.001	0.009

Note: The risk of CKD showed significant interactions between age (＜40 versus 40 ≥ years) and baPWV (p = 0.011, Model 2), as well as between sex and baPWV(p = 0.043, Model 2).

^a Estimated from parametric proportional hazards models.

Model 1 was adjusted for age, sex, center, year of screening examination, fasting plasma glucose, body mass index, total cholesterol, HDL cholesterol, medication for hypertension, medication for diabetes, medication for hyperlipidemia, alcohol consumption, smoking, physical activity, education level, and family history of CHD; model 2 was adjusted for variables included in model 1 and the baseline eGFR (CKD-EPI); and model 3 was adjusted for variables included in model 2 and systolic blood pressure.

baPWV, brachial-ankle pulse wave velocity; CHD, coronary heart disease; CI, confidence interval; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; eGFR, estimated glomerular filtration rate; HDL cholesterol, high-density lipoprotein cholesterol; HR, hazard ratio; PY, person-years.

Fig.2.

Cumulative incidence of CKD according to the baPWV within different sex strata

Risk of Incident CKD in Groups by baPWV Status

In the total study population, compared to the normal baPWV group, the abnormal group (＞1,800 cm/sec) had a significantly higher risk of incident CKD in the multivariable-adjusted hazard model (Models 1, 2, and 3 in Table 4). In the analysis of age-specific strata (＜40 and ≥ 40 years old), the borderline and abnormal baPWV groups had a significantly higher risk of incident CKD than the normal group in the population ＜40 years old (Table 4). However, only the abnormal baPWV group had a significantly higher risk of incident CKD among those ≥ 40 years old. An analysis according to sex-specific strata showed that only the abnormal baPWV group had a significantly increased risk of incident CKD in men, whereas populations in the high normal, borderline, and abnormal baPWV groups had a significantly higher risk of incident CKD in women than those in the normal group (Models 1 and 2 in Table 4). After adjusting for the systolic BP, a significant correlation between the abnormal baPWV group and incident CKD was no longer observed in men, whereas a significant relationship between the baPWV and incident CKD persisted in women (Model 3 in Table 4).

Predictors of Incident CKD according to Sex-Specific Strata

Results from the multivariable-adjusted parametric proportional hazards models indicated that high normal, borderline, and abnormal baPWV groups, medication for hypertension, and medication for diabetes were significant predictors of incident CKD in women (Table 5). In a multivariable-adjusted analysis in men, the abnormal baPWV group, age, body mass index, medication for hypertension, medication for diabetes, current and ex-smoking, and a high education level were significant predictors of incident CKD (Table 5). The baseline eGFR was shown to be a protective predictor of incident CKD in both men and women.

Table 5.Sex-specific univariate and multivariate HRs^a (95% CI) for incident chronic kidney disease

Characteristics		Univariate analyses		Multivariate analyses
		Women	Men	Women	Men
baPWV (cm/sec)	Normal (＜1,200)	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
	High normal (≥ 1,200 and ＜1,400)	3.43 (2.04–5.77)	1.03 (0.81–1.32)	1.92 (1.12–3.28)	0.83 (0.65–1.07)
	Borderline (≥ 1,400 and ＜1,800)	15.48 (9.21–26.01)	2.58 (2.03–3.28)	2.41 (1.29–4.49)	1.18 (0.92–1.52)
	Abnormal (≥ 1,800)	138.29 (76.54–249.87)	13.8 (10.02–19.01)	4.61 (2.06–10.30)	1.56 (1.08–2.26)
Age		1.15 (1.14–1.17)	1.12 (1.12–1.13)	1.02 (0.99–1.05)	1.05 (1.04–1.06)
Fasting plasma glucose		1.02 (1.02–1.03)	1.01 (1.01–1.02)	1.01 (1.00–1.02)	1.01 (1.00–1.01)
Body mass index		1.18 (1.14–1.23)	1.10 (1.08–1.13)	1.05 (0.99–1.11)	1.05 (1.02–1.09)
Total cholesterol		1.01 (1.01–1.01)	1.00 (1.00–1.00)	1.00 (1.00–1.01)	1.00 (1.00–1.00)
HDL cholesterol		0.97 (0.96–0.99)	0.97 (0.97–0.98)	1.00 (0.98–1.01)	0.99 (0.98–0.99)
eGFR (CKD–EPI)		0.87 (0.86–0.89)	0.85 (0.85–0.86)	0.88 (0.87–0.90)	0.86 (0.86–0.87)
Medication for hyperlipidemia	No	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
	Yes	10.6 (5.94–18.95)	4.27 (3.34–5.46)	0.93 (0.47–1.83)	1.19 (0.91–1.54)
Medication for hypertension	No	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
	Yes	17.8 (12.53–25.28)	5.84 (4.99–6.83)	1.72 (1.12–2.66)	1.62 (1.36–1.94)
Medication for diabetes	No	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
	Yes	24.03 (14.97–38.57)	5.93 (4.77–7.39)	2.41 (1.27–4.58)	1.56 (1.19–2.03)
Regular exercise	＜ 3 days/week	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
	≥ 3 days/week	2.03 (1.36–3.02)	1.71 (1.44–2.02)	0.91 (0.60–1.38)	1.05 (0.88–1.25)
Smoking status	Never	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
	Former	0.87 (0.48–1.58)	1.41 (1.13–1.75)	1.05 (0.58–1.93)	1.26 (1.01–1.58)
	Current	1.28 (0.41–4.02)	0.99 (0.78–1.26)	2.24 (0.69–7.23)	1.40 (1.09–1.79)
Alcohol intake	＜20 g/day	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
	≥ 20 g/day	0.63 (0.20–2.00)	0.95 (0.81–1.11)	0.94 (0.29–3.02)	1.01 (0.86–1.20)
Education	＜ college graduate	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
	≥ college graduate	0.31 (0.21–0.47)	0.81 (0.67–0.98)	0.68 (0.44–1.05)	1.31 (1.08–1.60)
Family history of CHD	No	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)
	Yes	0.77 (0.38–1.57)	1.16 (0.89–1.53)	0.79 (0.38–1.64)	1.17 (0.89–1.53)

^a Estimated from parametric proportional hazards models.

baPWV, brachial-ankle pulse wave velocity; CHD, coronary heart disease; CI, confidence interval; CKD–EPI, Chronic Kidney Disease Epidemiology Collaboration; eGFR, estimated glomerular filtration rate; HDL cholesterol, high–density lipoprotein cholesterol; HR, hazard ratio; PY, person-years.

Discussion and Conclusion

Our study findings show that increased arterial stiffness, as measured by baPWV, is associated with a higher risk of incident CKD with notable sex-specific differences. In women, even moderate increases in baPWV were significantly associated with an elevated risk of CKD, whereas in men, only markedly elevated baPWV was associated with an elevated CKD risk. However, after adjusting for systolic BP, the abnormal baPWV group remained at an elevated risk of CKD compared to the normal baPWV group in women, but this association was not observed in men.

Several studies have reported an association between arterial stiffness based on the baPWV and a decreased renal function, but these results were not consistent^5-8). In cross-sectional studies, one showed a significant relationship between the degree of GFR loss and baPWV values⁸⁾. However, this study was limited to young and middle-aged men⁸⁾. The other two studies did not show a significant association between the baPWV and a decreased GFR; however, this study showed a significant relationship between the baPWV and albuminuria or serum cystatin c levels^{6, 7)}. One longitudinal study also did not show a relationship between the baPWV and a decline in the kidney function⁵⁾. However, this study included moderate CKD patients (≥ 30 mL/min/1.73 m²), defined a rapid decline in the kidney function (＞3 mL/min/1.73 m² per year) as the primary endpoint, and had a relatively short follow-up period (median 3.2 years)⁵⁾. These factors might have influenced the negative results of this study. Unlike previous studies, our study was conducted in a large population (n=130,100 and mean age 40.5 years old), excluded subjects with abnormal baseline eGFR values (＜60 mL/min/1.73 m²), and had a relatively long follow-up duration (mean 5.6 years). Furthermore, the populations in this cohort were generally young and healthy; thus, our study may be less biased toward cardiovascular comorbidities and/or medications that may affect elderly individuals. Given the above, our study is likely to demonstrate the inherent impact of baPWV on the occurrence of CKD more effectively than other studies.

The kidney and arterial wall exert significant influences on each other, and this interaction may lead to a potentially vicious cycle of arterial stiffening and the development or progression of CKD and its complications²⁾. While the kidney can influence blood pressure regulation, it is also vulnerable to damage from hypertension over time¹⁴⁾. The kidney has the highest flow rate and lowest vascular resistance among large organs. This characteristic makes the kidney highly susceptible to trauma from pulsatile pressure and blood flow, which can damage the glomeruli and lead to albuminuria and a reduced GFR¹⁵⁾.

To our knowledge, the present study is the only investigation demonstrating the association between the baPWV and CKD development, along with sex-specific differences. In terms of the sex-specific difference in the cardiovascular impact of arterial stiffness, a previous study indicated that the mortality impact of arterial stiffness was two to three times higher in women than in men¹⁶⁾. In this study, the influence of arterial stiffness on cardiovascular mortality was notably amplified in women over 55 years old compared with their younger counterparts, while such an increase was not observed in men¹⁶⁾. The sex-specific disparity in the impact of arterial stiffness on cardiovascular outcomes is attributed to the loss of estrogenic action in the large arterial wall, resulting in increased stiffness and reduced elasticity^{17, 18)}. This phenomenon is linked to sex differences in the regulation of vascular tone, where sex hormones induce endothelium-dependent vascular relaxation and inhibit vascular smooth muscle contraction¹⁹⁾. These hormonal influences may contribute to sex differences in vascular tone. Furthermore, women may experience elevated arterial stiffness due to various clinical conditions, such as pre-eclampsia, polycystic ovary syndrome, endometriosis, and various autoimmune disorders⁹⁾. However, the women who participated in the present study were relatively young, and therefore the suggested mechanism related to hormonal influence may not be applicable in our findings. In our study, except for baPWV, medications for hypertension and diabetes were identified as significant risk factors for CKD development in women. However, in men, various factors such as the age, body mass index, smoking, medication for hyperlipidemia, and education, in addition to medications for hypertension and diabetes, were found to influence the occurrence of CKD. This difference suggests that the effect of the baPWV on CKD development may be attenuated in men. As arterial stiffness is pathophysiologically linked to aging, our results demonstrated significantly increased baPWV values with older age in men. Consequently, if women have higher baPWV levels for any reason, they would face an elevated risk of cardiovascular outcomes, including CKD development, compared with men.

Limitations

Several limitations associated with the present study warrant mention. First, this was a retrospective analysis, which has inherent limitations. Second, the study population comprised relatively healthy young and middle-aged Koreans. Therefore, generalizing our results to other age groups or patients of other races/ethnicities is limited. In addition, this study included subjects with a normal baseline eGFR; therefore, our results may not be applicable to patients with a decreased renal function. Third, we defined CKD based on an eGFR calculation. While guidelines state that CKD is defined as abnormalities of kidney structure or function present for over three months²⁰⁾, we did not use markers of kidney damage, including albuminuria or structural abnormalities detected by imaging, as definitions of CKD. Furthermore, we used a single abnormality of an eGFR ＜60 mL/min/1.73 m² as the primary study endpoint but could not determine whether these abnormalities persisted over 3 months. However, while our study definition did not meet all the criteria of the guideline-suggested CKD definition, our results clearly showed a decline in the renal function according to the baseline baPWV. Fourth, the serum creatinine levels in this cohort were measured using the Jaffe method. This method is somewhat non-specific, as substances such as glucose, proteins, and ketones can interfere with the reaction, potentially leading to overestimation of creatinine levels. In contrast, the enzymatic method is more specific and accurate, utilizing enzyme reactions to measure creatinine with minimal interference²¹⁾. Consequently, our results may have been influenced by the choice of method for serum creatinine measurement. Finally, the carotid-femoral PWV has been accepted as the current reference method for arterial stiffness; however, arterial stiffness was measured using the baPWV in this study. While the measurement of the carotid-femoral PWV involves more effort and the need for probes to detect the arterial pulse waveform, the method of measuring the baPWV is simple which involves wrapping pressure cuffs around all four extremities. Furthermore, the baPWV correlates better with the aortic PWV than other PWV measurements²²⁾. A recent meta-analysis established the baPWV as an independent predictor of the risk of developing cardiovascular disease in 14,673 Japanese subjects without preexisting cardiovascular disease²³⁾. Currently, the baPWV is widely used in Asia, and evidence of its value is growing with time. Thus, the baPWV adequately reflects arterial stiffness to the same extent as the carotid-femoral PWV.

Conclusions

Increased arterial stiffness, as measured by the baPWV, is associated with an increased risk of incident CKD with notable sex-specific differences. In women, even moderate increases in the baPWV were significantly associated with an elevated risk of CKD, whereas in men, only a markedly elevated baPWV was associated with an elevated CKD risk. After adjusting for systolic BP, the abnormal baPWV group remained at an elevated risk of CKD compared to the normal baPWV group in women, but this association was not observed in men. These findings underscore the utility of the baPWV for identifying CKD risk in women and offer valuable insights into sex-specific differences in arterial stiffness and CKD development.

Acknowledgements

We thank the health-screening group at Kangbuk Samsung Hospital, Seoul, Korea.

This study was written as part of Konkuk University’s research support program for its faculty on sabbatical leave in 2023.

Financial Support

No funding was received for this study.

Conflict of Interest

The authors declare that there are no conflicts of interest.

Author Contributions

Chang Hee Kwon: Conceptualization, methodology, writing – original draft, writing – review, and editing.

Jeonggyu Kang: Conceptualization, methodology, formal analysis, writing, review, and editing.

Ki-Chul Sung: Conceptualization, methodology, writing (review and editing), and visualization.

Abbreviations

baPWV=brachial-ankle pulse wave velocity

BP=blood pressure

CKD=chronic kidney disease

eGFR=estimated glomerular filtration rate

References

• 1)  Kaess BM, Rong J, Larson MG, Hamburg NM, Vita JA, Levy D, Benjamin EJ, Vasan RS and Mitchell GF: Aortic stiffness, blood pressure progression, and incident hypertension. JAMA, 2012; 308: 875-881
• 2)  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
• 3)  Vanholder R, Massy Z, Argiles A, Spasovski G, Verbeke F, Lameire N and European Uremic Toxin Work G: Chronic kidney disease as cause of cardiovascular morbidity and mortality. Nephrol Dial Transplant, 2005; 20: 1048-1056
• 4)  Tomiyama H and Shiina K: State of the Art Review: Brachial-Ankle PWV. J Atheroscler Thromb, 2020; 27: 621-636
• 5)  Kim CS, Kim HY, Kang YU, Choi JS, Bae EH, Ma SK and Kim SW: Association of pulse wave velocity and pulse pressure with decline in kidney function. J Clin Hypertens (Greenwich), 2014; 16: 372-377
• 6)  Kim ED, Tanaka H, Ballew SH, Sang Y, Heiss G, Coresh J and Matsushita K: Associations Between Kidney Disease Measures and Regional Pulse Wave Velocity in a Large Community-Based Cohort: The Atherosclerosis Risk in Communities (ARIC) Study. Am J Kidney Dis, 2018; 72: 682-690
• 7)  Song SH, Kwak IS, Kim YJ, Lee HS, Rhee H, Lee DW, Kang YH and Kim SJ: Serum cystatin C is related to pulse wave velocity even in subjects with normal serum creatinine. Hypertens Res, 2008; 31: 1895-1902
• 8)  Yoshida M, Tomiyama H, Yamada J, Koji Y, Shiina K, Nagata M and Yamashina A: Relationships among renal function loss within the normal to mildly impaired range, arterial stiffness, inflammation, and oxidative stress. Clin J Am Soc Nephrol, 2007; 2: 1118-1124
• 9)  DuPont JJ, Kenney RM, Patel AR and Jaffe IZ: Sex differences in mechanisms of arterial stiffness. Br J Pharmacol, 2019; 176: 4208-4225
• 10)  Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF, 3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J and Ckd EPI: A new equation to estimate glomerular filtration rate. Ann Intern Med, 2009; 150: 604-612
• 11)  Tanaka A, Tomiyama H, Maruhashi T, Matsuzawa Y, Miyoshi T, Kabutoya T, Kario K, Sugiyama S, Munakata M, Ito H, Ueda S, Vlachopoulos C, Higashi Y, Inoue T, Node K and Physiological Diagnosis Criteria for Vascular Failure C: Physiological Diagnostic Criteria for Vascular Failure. Hypertension, 2018; 72: 1060-1071
• 12)  Stevens PE, Levin A and Kidney Disease: Improving Global Outcomes Chronic Kidney Disease Guideline Development Work Group M: Evaluation and management of chronic kidney disease: synopsis of the kidney disease: improving global outcomes 2012 clinical practice guideline. Ann Intern Med, 2013; 158: 825-830
• 13)  Royston P and Parmar MK: Flexible parametric proportional‐hazards and proportional‐odds models for censored survival data, with application to prognostic modelling and estimation of treatment effects. Statistics in medicine, 2002; 21: 2175-2197
• 14)  Klahr S: The kidney in hypertension--villain and victim. N Engl J Med, 1989; 320: 731-733
• 15)  Hashimoto J and Ito S: Central pulse pressure and aortic stiffness determine renal hemodynamics: pathophysiological implication for microalbuminuria in hypertension. Hypertension, 2011; 58: 839-846
• 16)  Regnault V, Thomas F, Safar ME, Osborne-Pellegrin M, Khalil RA, Pannier B and Lacolley P: Sex difference in cardiovascular risk: role of pulse pressure amplification. J Am Coll Cardiol, 2012; 59: 1771-1777
• 17)  Mullick AE, Walsh BA, Reiser KM and Rutledge JC: Chronic estradiol treatment attenuates stiffening, glycoxidation, and permeability in rat carotid arteries. Am J Physiol Heart Circ Physiol, 2001; 281: H2204-2210
• 18)  Stefanadis C, Tsiamis E, Dernellis J and Toutouzas P: Effect of estrogen on aortic function in postmenopausal women. Am J Physiol, 1999; 276: H658-662
• 19)  Thompson J and Khalil RA: Gender differences in the regulation of vascular tone. Clin Exp Pharmacol Physiol, 2003; 30: 1-15
• 20)  Chapter 1: Definition and classification of CKD. Kidney Int Suppl (2011), 2013; 3: 19-62
• 21)  Panteghini M and Division IS: Enzymatic assays for creatinine: time for action. Clin Chem Lab Med, 2008; 46: 567-572
• 22)  Tsuchikura S, Shoji T, Kimoto E, Shinohara K, Hatsuda S, Koyama H, Emoto M and Nishizawa Y: Brachial-ankle pulse wave velocity as an index of central arterial stiffness. J Atheroscler Thromb, 2010; 17: 658-665
• 23)  Ohkuma T, Ninomiya T, Tomiyama H, Kario K, Hoshide S, Kita Y, Inoguchi T, Maeda Y, Kohara K, Tabara Y, Nakamura M, Ohkubo T, Watada H, Munakata M, Ohishi M, Ito N, Nakamura M, Shoji T, Vlachopoulos C, Yamashina A and Collaborative Group for JB: Brachial-Ankle Pulse Wave Velocity and the Risk Prediction of Cardiovascular Disease: An Individual Participant Data Meta-Analysis. Hypertension, 2017; 69: 1045-1052