Background
Peripheral Arterial Disease (PAD) is a significant manifestation of systemic atherosclerosis, influenced by risk factors such as smoking, dyslipidemia, and type 2 diabetes mellitus [1, 2]. According to a 2015 systematic review, the global prevalence of PAD was approximately 5.6%, affecting 236 million people worldwide [3]. The prevalence is notably higher in high-income countries (7.4%) compared to middle-income countries (5.1%), and there has been a 45% global increase in cases over the last 15 years [3].
In geriatric populations, specifically those over 70 years of age, an estimated 20% experience PAD, with a prevalence twice as high in men as in women [2, 4].
PAD is a critical marker for widespread vascular disorders, as the atherosclerotic process is not confined to the extremities but also affects coronary and cerebral vessels [5, 6]. The Ankle-Brachial Index (ABI) is a fundamental, non-invasive tool used to screen for and diagnose PAD. Its high validity and reliability, coupled with its simplicity and speed, make it an accurate and practical method for detecting peripheral arterial obstruction in clinical settings [5, 6].
Given that the ABI value reflects the severity of systemic atherosclerosis, it serves as a proxy for vascular health in the heart and brain.
While the shared pathophysiology of atherosclerosis suggests a strong link between PAD and cognitive impairment, direct evidence remains inconclusive. Johnson et al. reported older adults with a lower ABI to be associated with a lower level of cognitive function, suggesting a higher prevalence in the cohort [6]. Similarly, Waldstein et al. reported diminished cognitive function which are associated with higher prevalence of PAD and other manifestations of cardiovascular diseases [4].
This highlights a critical gap in the literature and a need for more robust scientific evidence to clarify the relationship between systemic vascular disease and its increased prevalence among geriatric patients with cognitive decline.
Therefore, this study aims to investigate the association between PAD and cognitive impairment within the specific geriatric population of Cipto Mangunkusumo Hospital, Indonesia’s national referral center. The novelty of this research lies in its focus on a unique patient demographic within a large-scale hospital setting, which has not been previously studied in this context.
Methods
Study design and participants
This cross-sectional study was conducted from January to March 2025 at the outpatient clinics (Internal Medicine, Geriatrics, Endocrinology, Cardiology) and inpatient wards of Cipto Mangunkusumo Hospital (RSCM) in Jakarta, Indonesia. The target population was geriatric patients in Indonesia, with the accessible population being individuals aged 60 years and older who visited or were admitted to these departments during the study period. A convenience sampling strategy was employed, enrolling all patients who met the eligibility criteria until the planned sample size was achieved.
All participants provided written informed consent before data collection.
The inclusion criterion was patients aged > 60 years. The exclusion criteria were: (1) a pre-existing diagnosis of a major psychiatric disorder or a clinical diagnosis of depression based on established diagnostic criteria; (2) severe cognitive impairment (MMSE score ≤ 7); (3) non-compressible arteries (ABI > 1.4); (4) a history of lower-limb amputation; (5) conditions that would preclude valid MMSE testing, such as aphasia, severe uncorrected visual or hearing impairment, or functional disability (e.g., right hemiparesis preventing writing or drawing).
This study received ethical approval from the Ethics Committee of the Faculty of Medicine, University of Indonesia, Jakarta, Indonesia (Registration Number: KET-1648/UN2.F1/ETIK/PPM.00.02/2024).
Measurements and definitions
Peripheral Artery Disease (PAD) Assessment PAD was assessed by measuring the Ankle-Brachial Index (ABI) using a 2.5 MHz Doppler device. The ABI was calculated by dividing the highest systolic blood pressure from the dorsalis pedis and posterior tibial arteries by the highest systolic pressure from both brachial arteries. PAD severity was categorized as: normal (ABI 0.9–1.4), mild (ABI 0.8–0.89), or moderate-to-severe (ABI < 0.8) [7, 8].
Cognitive Impairment Assessment Cognitive function was evaluated using the Mini-Mental State Examination (MMSE), which assesses domains including attention, executive function, short-term memory, and visuospatial skills. The MMSE was selected. While other tools such as the Montreal Cognitive Assessment (MoCA) are known for their sensitivity in vascular cases, the MMSE was chosen because it is the most widely used and studied cognitive screening tool worldwide, often serving as a benchmark for comparative evaluations.
A total score of < 24 was used to define the presence of cognitive impairment [9]. The Indonesian version of the MMSE used in this study was previously validated for the Indonesian population by Hogevorst et al. [10] Furthermore, a license for the use of the MMSE was formally obtained from the copyright holder, Psychological Assessment Resources (PAR), Inc.
Covariates
Data on covariates were collected through a combination of patient interviews and a thorough review of their medical records. Demographic information included sex (male or female) and age, which was categorized into two groups ≥ 65 and < 65 years. A detailed medical history was obtained, specifically noting physician-diagnosed conditions such as stroke, hypertension, type 2 DM, and dyslipidemia.
The use of relevant medications, including anti-diabetic, anti-hypertensive, and anti-dyslipidemia drugs, was also recorded.
In addition, a Comprehensive Geriatric Assessment (CGA) was performed to evaluate several key domains. Functional status was assessed using the Barthel Index of Activity Daily Living (BADL), with a score below 20 indicating dependency [11]. Nutritional status was evaluated with the Mini Nutritional Assessment (MNA), where a score below 23.5 was used to classify a patient as malnourished [12].
Frailty was determined using the 40-item Frailty Index (FI-40), categorizing patients as prefrail (score > 0.08 to < 0.25) or frail (score ≥ 0.25) [13, 14]. Finally, the overall comorbidity burden was measured using the Cumulative Illness Rating Scale-Geriatric (CIRS-G), with a score greater than 5 indicating a higher severity of illness [15].
Statistical analyses
Data were analyzed using SPSS version 26.0 Descriptive statistics were used to summarize participant characteristics. Bivariate analysis was conducted using the chi-square test to examine initial associations between potential risk factors and cognitive impairment.
For the multivariate analysis, a multiple logistic regression model was constructed to assess the association between ABI and cognitive impairment. To minimize bias and control for confounding, all clinically relevant covariates (including age, sex, education level, and comorbidities such as hypertension, DM, and stroke) were included in the final model regardless of their p-value in the bivariate analysis. The alpha level for statistical significance was set at 0.05.
Results
A total of 155 patients were initially screened for eligibility from online medical records. From this group, 25 individuals were excluded for the following reasons: inability to communicate in Indonesian (n = 3), presence of delirium (n = 10), severe pre-existing cognitive impairment (n = 7), and a history of lower-limb amputation (n = 5). Consequently, the final sample size for this study comprised 130 participants who underwent complete analysis.
The majority of participants were female (n = 68, 52.3%). Based on Ankle-Brachial Index (ABI) measurements, 69 participants (53.1%) were classified as having normal peripheral circulation, 49 (37.7%) were diagnosed with mild PAD, and 12 (9.2%) had moderate-to-severe PAD. A comprehensive summary of the baseline demographic characteristics is provided in Table 1.
A significant dose-response relationship was found between the severity of PAD and the prevalence of cognitive impairment (p = 0.007). As the severity of PAD increased, so did the proportion of participants with cognitive impairment, with the highest rates observed in the moderate-to-severe PAD category (Table 2).
Table 1Characteristics of participants (n=130)Characteristics of participantsAll participants (n= 130)%Age, Interquartil rangeAge group (y), n (%)67.53 (60–88) 60–6992 (68,14) 70–7935 (25,92) > 803 (0,02)Sex, n (%) Male62 (47,70) Female68 (52,30)Marital Status, n (%) Married124 (94,62) Divorce7 (5,38)Educational Level, n (%) Did not complete Elementary School2 (1,56) Elementary School16(12,30) Junior High School50(38,46) Senior High School47(36,15) Bachelor Degree15 (11,53) ABI value, mean0,84 ± 0,11PAD, n (%) Normal-Mild (0,8-0,9)117 (90) Moderate-Severe (<0,4-<0,8)13 (10) MMSE, mean23,13 ± 2,65Cognitive Impairment, n (%) Cognitive Impairment (MMSE < 24)61 (46,90) Normal (MMSE >/24)69 (53,10) Previous CAD, n (%)17 (13,10) Previous stroke, n (%)11 (8,50) Hypertension, n (%)72 (55,40) Type 2 DM, n (%)42 (32,30) Dislipidemia, n (%)24 (18,50)Functional status, n (%) Independent (BADL 20)42 (32,3) Mild-Moderate Dependent (BADL 9–19)78 (60,0) Severe Dependent (BADL <9)10 (7,40)Nutritional status (MNA), n (%) Normal (MNA > 23.5)47(36,15) Malnutrition risk (MNA 17 - 23,5)60 (46,15) Malnutrition (MNA < 17)23 (17,03) (CIRS-G), median (min- maks)4 (1–15) CIRS-G score < 569 (53,07) CIRS-G score > 561 (46,93)Frailty Status (FI40), n (%) Robust (FI40 < 0.08)32(23,70) Pre-Frail (FI40 >0.08 - < 0.25)53(40,76) Frail (FI40> 0.25)45(34,61)Previous medication, n (%) Anti diabetic drugs 1. Biguanides21(61) 2. Thiazolidinediones0 (0) 3.
Sulphonylureas3 (9) 4. Alfa glucosidase inhibitors3 (9) 5. DPP-4 inhibitors4 (12) 6.
SGLT-2 inhibitors0 (0) 7. Insulin3 (9) Anti hypertensive drugs 1. ACE-inhibitor/ARB42 (35) 2.
Beta blocker13 (12) 3. CCB40 (34) 4. Diuretic10 (8) 5.
Other anti hypertensive drugs (Tiazid, alfablocker, MRA, ARNI)12 (11) Anti dislipidemia drugs 1. Simvastatin 10 mg0 (0) 2. Simvastatin 20–40 mg4(20) 3.
Atorvastatin 10–20 mg11(50) 4. Atorvastatin 40–80 mg6 (30)Abbreviation description: ABI ankle brachial index, PAD peripheral artery disease, MMSE mini mental state examination, CAD coronary artery disease, DM diabetes mellitus, BADL barthel activity daily living, MNA mini nutritional assessment, CIRS-G cumulative illness rating scale geriatric, FI-40 frailty index-40, DPP-4 dipeptidyl peptidase-4, SGLT-2 sodium glucose cotransporter-2, ACE angiotensin converting enzyme, ARB angiotensin II receptor blockers, CCB calcium channel blocker, MRA mineralcorticoid receptor antagonist, ARNI angiotensin receptor neprilysin inhibitor
Table 2The association between ankle brachial index (ABI) and cognitive impairmentVariableCognitive ImpairmentP valueYes, n = 61 (%)No, n = 69 (%)PAD n (%) Normal24 (34.8)45 (65.2)0.007 Mild28 (57.1)21 (42.9) Moderate-Severe9 (75.0)3 (25.0)
In the analysis of potential confounders, a history of stroke (p < 0.001) and hypertension (p < 0.001) were both found to have a significant bivariate association with cognitive impairment (Table 3). Subsequently, a multivariate logistic regression analysis was performed to determine if PAD was an independent predictor of cognitive impairment after controlling for these and other variables. In this model, PAD was categorized as either absent or present based on the ABI.
Table 3The relationship between confounding variables and cognitive impairmentVariableCognitive ImpairmentPR (CI 95%)P valueYes, n = 61 (%)No, n = 69 (%)Age n (%) ≥ 65 yo40 (48.78)42 (51.22)1.11 (0.75–1.64)0.58 < 65 yo21 (43.75)27 (56.25)Sex n (%) Female34 (50.0)34 (50.0)1.14 (0.79–1.66)0.46 Male27 (43.55)35 (56.45)Previous CAD n (%) Yes9 (52.94)8 (47.06)1.15 (0.70–1.88)0.57 No52 (46.02)61 (53.98)Previous Stroke n (%) Yes9 (81.82)2 (18.18)1.87 (1.32–2.64)< 0.01 No52 (43.70)67 (56.30)Hypertension n (%) Yes42 (58.33)30 (41.67)1.78 (1.17–2.70)< 0.01 No19 (32.76)39 (67.24)Type 2 DM n (%) Yes24 (57.14)18 (42.86)1.35 (0.94–1.94)0.09 No37 (42.95)51 (42.05)Dyslipidemia n (%) Yes13 (54.17)11 (45.83)1.19 (0.78–1.82)0.40 No48 (45.28)58 (54.72)Bold p-values = <0.01
The final logistic regression model identified three variables as independent predictors of cognitive impairment. A history of stroke increased the prevalence of cognitive impairment by 4.3 times (p = 0.009), and a history of hypertension increased the odds by 2.5 times (p = 0.016). Crucially, after adjusting for confounders, PAD remained a significant and independent factor.
Participants with PAD had a 2.69 times greater risk of having cognitive impairment compared to those without PAD (Prevalence Ratio [PR] 2.69; 95% CI 1.28–5.68; p = 0.009), as detailed in Table 4.
Table 4Multivariate logistic regressionVariablePR (95% CI)P valueBSEStep 6 PAD2.69 (1.27–5.67)0.0090.9910.38 Hystory of Stroke4.32 (0.84–22.2)0.0801.4630.836 Hystory of Hypertension2.53 (1.19–5.38)0.0160.9280.385
Discussion
In this study, it was found that the majority of patients were ≥ 65 years old (63.10%), with 36.9% being < 65 years old. The majority of the subjects in this study were female (52.3%). This percentage is similar to the gender distribution of the geriatric population reported by the Central Statistics Agency in 2024, which is 52.2% female and 47.8% male [16].
The most frequent comorbidity in this study was hypertension (55.4%), followed by type 2 diabetes mellitus (32.3%) and dyslipidemia (18.5%), respectively. These results are different from the study of de Lima et al., who studied the prevalence of comorbidities in geriatrics living in the community [17–20]. de Lima et al. found that the most common comorbidities were similar to this study, but different in terms of order and prevalence, namely hypertension in 56.9% of patients, dyslipidemia in 18.6% of patients, and type 2 DM in 16.7% [21]. A study of comorbidities in geriatrics in Indonesia conducted by Purwata et al., in 10hospitals spread across Indonesia, found that the most common comorbidities were hypertension (41.31%) and type 2 DM (14.23%) [22].
This study found that patients with PAD had a higher risk of cognitive impairment than patients without (Prevalence Ratio [PR] 2.69; 95% CI 1.28–5.68; p = 0.009),. The potential mechanisms contributing to the genesis of cognitive impairment in patients with PAD are not well understood and may involve decreased cerebral perfusion. It is possible that atherosclerotic lesions of the larger cerebral arteries and functional disturbances in cerebral microcirculation are other causative factors, which may trigger cognitive impairment [23–25].
According to Stillfried et al. in PROSCIS-B, a prospective cohort study with homogeneous research subjects in 325 patients which assessed the relationship between low ABI index, especially ABI < 0.9 is closely related to decreased cognitive function in patients who have experienced a first ischemic stroke attack [26]. A higher prevalence of cognitive impairment is observed, The mean age was 72.9 years, with 59.3% of them female, where there was a 17% prevalence of cognitive impairment in low ABI subjects (< 0.9) and 12% prevalence of cognitive impairment in high ABI subjects (> 1.4) [27].
A study conducted by Wang A et al., obtained results - low ABI was associated with cognitive impairment (OR 1.98; CI: 1.15–3.41) after adjustment for risk factors, decreased ABI values were significantly correlated with decreased MMSE scores (β = 0.70, CI 95%: 0.18–1.21, p= 0.0074) [28]. Low ABI (< 0.9) reflects systemic atherosclerosis leading to decreased cerebral perfusion, especially in the hippocampus and prefrontal cortex, which play a role in memory and executive function, as well as endothelial dysfunction and cerebral microangiopathy leading to subcortical ischemic lesions and vascular dementia [28].
The result of the present study is also consistent with previous studies conducted by Chen H et al. in a prospective longitudinal study in China on 408 geriatric hypertensive patients (222 age 60–75 years, 186 age 75–89 years) followed for 2 years analyzed the association of low ABI (≤ 1.1) with cognitive impairment as well as decline in cognitive function over 2 years, with MMSE cut off < 24 categorized as cognitive impairment [29]. Low ABI scores were associated with cognitive impairment in the ≥ 75 years age group (OR = 0.01, p = 0.01), the decline in MMSE score was greater in the low ABI group compared to high ABI (0.76 vs. 0.29, p= 0.01) [29].
Vascular risk factors, such as hypertension, type 2 diabetes melitus, and atherosclerosis, contribute to the pathophysiology of cognitive impairment in vascular dementia [30–34]. These risk factors lead to endothelial dysfunction, vascular remodeling, and reduced cerebral blood flow, which impair the delivery of nutrients and oxygen to brain regions and promote cerebral hypoperfusion [29, 30]. Chronic cerebral hypoperfusion contributes to white matter lesions, microinfarcts, and small vessel disease associated with cognitive impairment.
In response to cerebral ischemia, inflammatory processes are activated within the brain [33, 34]. Activated microglia and astrocytes release pro-inflammatory cytokines, such as interleukin-1β (IL-1β), tumor necrosis factor-alpha (TNF-α), and interleukin-6 (IL-6) [25, 26]. Neuroinflammation exacerbates neuronal damage, disrupts synaptic connections, and contributes to neurodegeneration, ultimately leading to cognitive impairment [33, 34].
While low ABI values indicate the severity of PAD in patients, which directly illustrates the process of systemic atherosclerosis in macrovascular and brain hypoperfusion in general [27–30]. Chronic hypoperfusion, chronic inflammation, along with elevated CRP and IL-6, microvascular dysfunction also result in white matter lesions (WMLs) which directly lead to decreased memory and executive brain function [35–38]. While the potential mechanism of diabetes mellitus causes cognitive impairment due to metabolic dysfunction, due to the accumulation of oxidative stress and increased advanced glycation end products (AGEs), resulting in chronic inflammation.
Further mechanisms can inhibit insulin signal transduction in the brain, which affects synapse neuroplasticity. Insulin resistance in DM also triggers the acceleration of amyloid plaque accumulation as the basic pathophysiology of Alzheimer’s dementia [39–42].
Future direction and limitations
Despite the significant findings, this study has limitations. The cross-sectional design precludes the determination of causality and only establishes an association. Furthermore, while we controlled for major medical conditions, residual confounding from unmeasured factors such as socioeconomic status, education level, and lifestyle could not be fully ruled out.
A critical limitation, as noted by the reviewers, is the small number of participants in the moderate-to-severe PAD subgroup (n = 12). This small sample size means that the observed link between PAD severity and cognitive impairment in the initial analysis could be overestimated or may not hold true in a larger, more balanced dataset. We acknowledge that this approach, while statistically necessary for our data, may have masked a potential dose-response relationship, where the risk of cognitive impairment might increase proportionally with the severity of PAD.
Therefore, future research with a prospective cohort design and a larger sample size is essential to confirm causality and to more accurately explore the dose-dependent relationship between ABI levels and cognitive decline in geriatric patients.
Conclusion
This study concludes that there is a significant association between PAD and cognitive impairment in elderly Indonesian patients, even after controlling for other major vascular risk factors. The Ankle-Brachial Index may serve as a simple, non-invasive screening tool to help identify geriatric patients at higher risk of cognitive impairment. We suggest that screening for both PAD and cognitive impairment should be an essential part of the comprehensive geriatric assessment, especially for individuals with a high vascular risk profile.
Future prospective cohort studies are needed to confirm these findings and to clarify whether managing PAD can help prevent or delay the onset of cognitive decline.