A Comparative Exploratory Study of Cognitive and Functional Outcomes by Sex and Age in Patients With Type 2 Diabetes Mellitus and Atrial Fibrillation

Abstract

Background/Aim

Atrial fibrillation (AF) and type 2 diabetes mellitus (T2DM) are associated with cognitive decline, but possible sex-related differences in cognitive and functional outcomes remain uncertain. This exploratory study evaluated sex-stratified associations of clinical factors, depressive symptoms, and functional independence with severe cognitive impairment in patients with T2DM, AF, and newly identified cognitive decline.

Patients and Methods

The study included 140 patients with T2DM, AF, and a Mini-Mental State Examination (MMSE) score <27, comprising 68 men and 72 women. Cognitive performance, depressive symptoms, and functional status were assessed using the MMSE, Montreal Cognitive Assessment, 15-item Geriatric Depression Scale (GDS-15), Activities of Daily Living (ADL), and Instrumental Activities of Daily Living scales. Logistic regression was used to evaluate factors associated with severe cognitive impairment, defined as MMSE <24, in the overall cohort and separately in male and female patients.

Results

Severe cognitive impairment was present in 32/68 men (47.06%) and 36/72 women (50.00%), with no significant difference between sexes. Increasing age and higher diastolic blood pressure were associated with higher odds of MMSE <24 in the overall cohort and in sex-stratified analyses. Higher GDS-15 scores were associated with increased odds of severe cognitive impairment in both sexes, whereas greater functional independence, reflected by higher ADL scores, was associated with lower odds of MMSE <24.

Conclusion

Age, higher diastolic blood pressure, and greater depressive symptom burden in patients with type 2 diabetes mellitus were associated with increased odds of severe cognitive impairment, whereas greater functional independence was associated with lower odds in both sexes. These exploratory findings support cognitive, cardiovascular, psychological, and functional assessment in patients with T2DM and AF.

Introduction

In recent years, the prevalence of dementia has markedly risen globally, with those in low and middle-income countries appearing to be at heightened risk. While the risk factors contributing to the accelerated onset of dementia in our society remain incompletely understood, ethnicity, geographical regions, cardiovascular pathologies, cardiovascular risk (CVR) factors for atherosclerosis, and type 2 diabetes mellitus (T2DM) appear to be significant determinants. Additionally, it has been shown that brain damage, insufficient mental stimulation, and alcohol consumption are primary risk factors for dementia (1-4). Reduced educational attainment, advanced age, diabetes mellitus, myocardial infarction, cardiac arrhythmias, particularly atrial fibrillation, congestive heart failure, epileptic seizures, global cerebral atrophy–especially of the medial temporal lobe–and white matter degeneration elevate the risk of dementia following a stroke.

There are two primary types of dementia: Alzheimer’s disease (AD), which constitutes 60-70% of dementia cases, and vascular dementia (VaD), representing 20%. Recent research indicates that AD appears to be more prevalent in women than in men (5-7). A recent notion suggests that a segment of the X chromosome, which encodes the cells’ internal clearance processes, may be accountable for the abnormal accumulation of amyloid and tau proteins in AD (7). VaD arises from multiple diseases, including stroke, multi-infarct conditions, and subcortical dementia. Sex significantly influences the prevalence of these vascular illnesses, with stroke and cerebral atrophy occurring more frequently in males, whereas medial temporal lobe atrophy, cerebrovascular white matter abnormalities, and a familial history of dementia are more prevalent in women (5). Dementia may arise from both AD and VaD (5). Gong et al. conducted a study with 29,850 individuals, 58% of whom were women, to investigate potential sex disparities in the prevalence of dementia, concluding that women may be more susceptible to developing dementia than men (1). Additional potential factors may include sex differences arising from socio-political, social, and cultural contexts. Global population analysis reveals that women exhibited lower educational attainment and a higher prevalence of depression, whilst men demonstrated a greater tendency towards smoking, alcohol use, and typically engaged in more vigorous physical activity than women (1, 8). Discrepancies in perspectives regarding the primary risk factors for cognitive decline (CD) and dementia may be attributed to the uneven distribution of these risk factors among different communities and sexes.

Other researchers have shown that detrimental social behaviors, such as smoking, substance abuse, and alcohol use, together with various diseases, including chronic coronary syndrome (CCS) and systemic hypertension (SH), may expedite the progression from CD to dementia in individuals with VaD (5, 9). Numerous studies have suggested that T2DM and atrial fibrillation (AF) are risk factors for the onset of CD and potentially dementia (10-13). Individuals with both AF and T2DM exhibited the greatest risk of developing severe cognitive decline or dementia, likely attributable to heightened susceptibility to vascular problems and related comorbidities in this population (10, 14). Certain researchers showed that the interplay between depression and T2DM was associated with an increased chance of dementia (5).

Experts in CD identified age, informant-reported cognitive symptoms, apathy, psychosis, CVR, CVD, T2DM, scores from the Montreal Cognitive Assessment (MoCA) and/or Mini-Mental State Examination (MMSE), findings from fluoro-deoxyglucose-positron emission tomography, hippocampal atrophy, parkinsonism, and global atrophy as the most significant predictors for the progression from CD to dementia (15, 16).

In 2021, Chen et al. and in 2023, Wood et al. were among the initial researchers to investigate the potential influence of sex differences on the relationship between atrial fibrillation and cognitive decline (17, 18). Their findings indicated that women with AF exhibited an elevated risk of CD and a more rapid progression of CD to dementia in comparison to males with a comparable risk profile (17, 18). Certain researchers proposed that association between depression and T2DM correlates with an elevated risk of dementia (19). Other investigators noted that T2DM correlates with more severe CD in females compared to males, and that depressive disorder is more frequently detected in patients with T2DM (11, 20). Other research indicated that while a greater number of men are diagnosed with T2DM, this occurs at a lower body mass index and at an earlier age compared to women. Conversely, women experience a greater CVR burden and an increased incidence of cardiovascular problems at the time of diagnosis with T2DM. A further issue stems from the reality that women experience more unpleasant pharmaceutical effects and are less frequently administered appropriate cardiovascular risk reduction therapy compared to males, perhaps due to psychological factors (21). Frailty appears to elevate the incidence of atrial fibrillation in women more significantly than in males (22). Depression is a common comorbidity in patients with diabetes, and its coexistence with diabetes has been associated with a greater disease burden, adverse clinical outcomes and an increased risk of diabetes-related complications (23-25).

This study sought to evaluate sex-related patterns in the severity of cognitive impairment, as measured using the MMSE, in patients with T2DM, AF, and recently identified cognitive decline (MMSE <27). The second objective was to evaluate the associations of activities of daily living and depressive symptoms with severe cognitive impairment (MMSE <24) separately in women and men.

Patients and Methods

Study cohort. This was an exploratory, cross-sectional, single-center observational study. Our study group was comprised of patients selected from a total of 3,250 individuals hospitalized for cardiovascular evaluation to the Internal Medicine Department at the Municipal Emergency Hospital Timisoara between June 1, 2021, and March 2025.

Inclusion criteria: 1) age 50 years or older; 2) an MMSE score of less than 27 during the neurologic evaluation conducted during hospitalization; 3) no prior diagnosis of cognitive decline or severe cognitive impairment, nor treatment for these conditions before hospitalization; 4) history of CVD; and 5) chronic, stable T2DM and AF.

Exclusion criteria: 1) uncontrolled T2DM or using insulin therapy; 2) acute decompensated cardiac problems, including acute coronary syndrome and severe heart failure; 3) uncontrolled hypertension; 4) other acute or unstable medical conditions; 5) inability or unwillingness to provide informed consent.

Upon reviewing the inclusion and exclusion criteria, 3,110 individuals were excluded, leaving just 140 patients with T2DM, AF, and MMSE scores below 27 in the final study cohort. The patient cohort was categorized by sex as follows: group A comprised 68 male patients, whereas group B consisted of 72 female patients. The participant selection process is shown in Figure 1.

Figure 1.

STROBE flow diagram of patient selection.

Patient evaluation. Clinical and laboratory evaluations. All 140 patients underwent a comprehensive clinical examination, accompanied by neurologic and cardiologic assessments, which included an extensive medical history and physical evaluations measuring heart rate (HR), systolic blood pressure (SBP), and diastolic blood pressure (DBP). Laboratory data were obtained from the hospital’s electronic database. The triglyceride-glucose index (TyG) was computed using the formula ln [BBG (mg/dl) × TG (mg/dl)/2], where TG denotes fasting triglyceride levels. The TyG index was calculated as described previously (26). The TG/HDL ratio was calculated as well. AF was managed in accordance with the relevant guideline recommendations (27). The CHA₂DS₂-VA and HASBLED scores were computed for every patient. The CHA₂DS₂-VA score was calculated based on congestive heart failure/left ventricular dysfunction (1 point), hypertension (1 point), age ≥75 years (2 points), diabetes mellitus (1 point), previous stroke, transient ischemic attack or systemic thromboembolism (2 points), vascular disease (1 point), and age 65-74 years (1 point). Female sex was not included as a component of the score; thus, the score ranged from 0 to 8 points. T2DM was diagnosed based on medical history, baseline blood glucose (BBG) levels, and glycated hemoglobin.

All patients underwent a resting electrocardiogram (EKG). To diagnose peripheral artery disease, we computed the ankle-brachial index (ABI) and intima-media thickness (IMT) (28). Transthoracic echocardiogram (TTE) was performed to assess the size and function of the left ventricle (LV) and left atrium (LA) (29). The left ventricular ejection fraction (LVEF) was utilized to assess left ventricular systolic performance.

A senior neurology specialist evaluated the presence and severity of cognitive decline using multiple scales commonly utilized in healthcare settings, which are available in Romanian at no cost: MMSE, MoCA, Activities of Daily Living (ADL), Instrumental Activities of Daily Living (IADL), and Geriatric Depression Scale (GDS-15).

All patients presented with chronic, stable illnesses and received appropriate treatment in accordance with medical guidelines. Patients received treatment for AF with direct oral anticoagulants (DOAC) and for T2DM with oral hypoglycemic agents. The research was executed in accordance with the principles of the Declaration of Helsinki and received approval from the Institutional Review Board of Municipal Emergency Hospital Timisoara, Romania (number. E1518 / 17 March 2021).

Neurological assessment of CD. Mini-Mental State Examination (MMSE). The MMSE scale is frequently utilized for screening cognitive diseases, evaluating the course and severity of cognitive impairment, and monitoring these patients. The MMSE scale assesses cognitive functions, including direction, attention, computation, recall, construction activities, and language manipulation. The maximum score is 30 points; however, it must be assessed in conjunction with the patient’s age and educational attainment (30, 31). A score ranging from 24 to 26 indicates mild cognitive impairment, while a score below 24 was classified in the present study as severe cognitive impairment (30, 31).

Montreal Cognitive Assessment (MoCA). The MoCA scale assesses many cognitive skills, including orientation, attention, language, working memory, visual-spatial abilities, concentration, short-term memory, executive functions, recall tasks, a two-item verbal abstraction problem with phonemic fluency, and the reproduction of a three-dimensional cube. The use of this assessment typically requires approximately 10 min, with a maximum score of 30 points. Minor cognitive impairment is indicated by scores below 27, whereas lower scores indicate greater cognitive impairment (32, 33).

Activities of daily living. The ADL score is frequently employed to assess patients’ daily self-care capabilities. Five fundamental domains are evaluated: eating, clothing, transferring and mobility, preserving personal hygiene, and continence. The ADL scale measures daily functional status, primarily in people with disabilities who are elderly (34-36). The highest score is 10 points, and the administration duration is 10 min. The minimum score is 0 points (indicating low functionality), while the maximum score is 10 points (indicating great functioning) (35).

Instrumental activities of daily living. The IADL scale is frequently utilized to assess a patient’s ability to perform daily care activities. It consists of a series of inquiries examining eight activity domains, including cooking, shopping, and financial management. It is communicated to the patient, either verbally or in written format, and its duration is approximately 10 min. A total score of 0 signifies low functionality, whereas a score of 8 denotes good functionality (37, 38).

Geriatric Depression Scale. The GDS scale detects depressed symptoms and is applicable to all age demographics. Our study utilized the abbreviated version consisting of 15 questions (GDS-15). The GDS-15 scale is more readily applicable to individuals without cognitive decline, as well as to those with cognitive impairment, for whom a multi-question evaluation may be challenging (38-42). Its administration is more challenging in patients with CD and should be interpreted in respect to age and educational attainment. This scale exhibits a specificity of 89% and a sensitivity of 92% (39-43). Patients are required to respond to 15 questions, and the assessment lasts roughly 15 min. A score ranging from 0 to 4 points is deemed normal, a score from 5 to 8 points indicates mild depression, a score from 9 to 11 points characterizes moderate depression, and a score from 12 to 15 points implies severe depression (39-43).

Statistical evaluation. Statistical analysis was conducted with IBM SPSS software for Windows version 20.0. (IBM Corporation with global corporate headquarters located in Armonk, NY, USA). Results are presented as counts and percentages for categorical data, and as mean value±standard deviation for continuous data. We utilized the unpaired t-test, with a significance threshold of 0.05, to assess differences between groups A and B regarding biological parameters, heart rate, SBP, DBP, intima-media thickness, ankle-brachial index, left ventricular function parameters, and neuropsychological assessments. The chi-square test was utilized to evaluate disparities across categorical variables by calculating the associated p-values. Correlations among the data were determined using Spearman’s correlation coefficient, and the associated 95% confidence intervals (95%CI) were generated employing Bonnett and Wright’s approach (44).

To evaluate factors associated with severe cognitive impairment (MMSE <24), logistic regression analyses were performed in the overall study population and separately in male and female patients. Model 1 included age together with blood pressure parameters (SBP and DBP), whereas model 2 evaluated ADL and GDS-15. Variables were selected using the forward Wald method, and odds ratios (ORs) with corresponding 95% confidence intervals (95% CIs) were calculated. Nagelkerke R² was used to evaluate model fit. Sex-stratified analyses were used to evaluate the associations separately in male and female patients; formal sex-by-predictor interaction testing was not performed.

Ethical considerations. The present study was conducted within the same ongoing research program and under the same ethics approval (E1518/17.03.2021) as two previous publications from our group. However, the 140 participants included in the present analysis constitute a distinct patient sample. Cross-checking of participant records confirmed that none of the participants included in the current study had been included in either the previously published Biomedicines 2024 cohort or the Journal of Clinical Medicine 2025 cohort (overlap with each previous cohort: n=0). Thus, although the studies share the same institution, research framework, and some clinical assessment methods, the participant-level data analyzed in the present study have not previously been reported.

Results

Study population. In our cohort of 140 patients, the mean age was 73.89±8.40 years (ranging from 50 to 89 years), with men being younger (mean age=72.49±7.61 years) than women (mean age=75.22±8.93 years), as illustrated in Table I. In group A, 64 patients (94.10%) were hypertensive, 22 (32.40%) exhibited paroxysmal AF, 17 (25.00%) had persistent AF, and 29 (42.60%) presented with chronic AF. Additionally, 25 patients (36.80%) were smokers, and 44 (64.70%) had hyperlipidemia (Table I). In group B, 67 (93.10%) women exhibited SH, 27 (37.50%) had paroxysmal AF, 15 (20.80%) had persistent AF, and 30 (41.70%) had chronic AF. Additionally, 27 (37.50%) were smokers, and 54 (75.00%) had hyperlipidemia (Table I).

Table I. Demographic characteristics, cardiovascular risk factors, and comorbidities of male (group A) and female (group B) patients with type 2 diabetes mellitus (T2DM), atrial fibrillation (AF), and cognitive decline [Mini-Mental State Examination (MMSE) score <27].

AF: Atrial fibrillation; CCS: chronic coronary syndrome; CHF: chronic heart failure; CKD: chronic kidney disease; NYHA: New York Heart Association; P: patient, PAD: peripheral artery disease; T2DM: diabetes mellitus type 2; p: statistical significance according to the chi-square test and unpaired t-test (*).

Comparative analysis of biologic parameters in groups A and B. Our findings indicated statistically significant elevations in total cholesterol and low-density lipoprotein (LDL) cholesterol levels in group A compared to group B (p<0.05). Additionally, elevated levels of triglycerides (TG), high-density lipoprotein (HDL) cholesterol, and diminished values of the TyG index and TG/HDL ratio were seen in Group A compared to group B; these differences were not statistically significant (Table II). Although NT-proBNP levels were high in both cohorts, women exhibited lower values than men; however, this difference was not statistically significant. No statistically significant changes were seen between groups A and B when assessing other biological markers (Table II).

Table II. Laboratory parameters of male (group A) and female (group B) patients with type 2 diabetes mellitus (T2DM), atrial fibrillation (AF), and cognitive decline [Mini-Mental State Examination (MMSE) score <27].

eGFR: Estimated glomerular filtration rate; HDL cholesterol: high-density lipoprotein; LDL cholesterol: low-density lipoprotein; NT-proBNP: N-terminal pro-B-type natriuretic peptide; TG: Triglycerides; TyG: triglyceride glucose index; p: statistical significance according to the unpaired t-test. Bold values indicate statistical significance.

Hemodynamic, metabolic and subclinical parameters in both groups of patients. There were no statistically significant differences between men and women in SBP, DBP, HR, LVEF, BMI, WC, left or right IMT, left or right ABI, or HASBLED score. Women had significantly higher CHA₂DS₂-VA scores than men (3.93±1.52 vs. 3.23±0.88, p=0.001) (Table III).

Table III. Hemodynamic parameters, IMT, ABI, CHA₂DS₂-VA, and HASBLED scores in both groups of patients [68 male (Group A) and 72 female (Group B) patients with T2DM/AF and CD (MMSE<27)].

Legend: ABI: Ankle Brachial Index; BMI: body mass index; CHA₂DS₂-VA: Score for Atrial Fibrillation Stroke Risk; DBP: diastolic blood pressure; HASBLED: Score for Hemorrhagic Risk Evaluation in Atrial Fibrillation patients; HR: heart rate; IMT: intima-media thickness; LVEF: left ventricular ejection fraction; SBP: systolic blood pressure; WC: waist circumference; p: statistical significance according to the unpaired t-test. Bold values denote statistical significance.

Neuropsychological tests in both groups of patients. There were no statistically significant differences between men and women in MMSE, MoCA, ADL, or GDS-15 scores. IADL scores were significantly higher in women than in men (6.47±1.79 vs. 5.65±1.76, p=0.007) (Table IV).

Table IV. Neuropsychological tests in both groups of patients [68 male (Group A) and 72 female (Group B) patients with T2DM/AF and CD (MMSE<27)].

Legend: ADL: Activities of Daily Living Score; GDS-15: Geriatric Depression Scale 15 questions; IADL: Instrumental Activities of Daily Living Score; MMSE: Mini-Mental State Examination Scale; MoCA: Montreal Cognitive Assessment Scale; p: statistical significance according to the unpaired t-test. Bold values denote statistical significance.

Of the 68 male patients, 36 (52.94%) had MMSE scores of 24-26 and 32 (47.06%) had MMSE <24. Of the 72 female patients, 36 (50.00%) had MMSE scores of 24-26 and 36 (50.00%) had MMSE <24.

There was no statistically significant difference in the prevalence of MMSE <24 between men and women (47.06% vs. 50.00%; χ² test, p=0.728).

Correlation between neuropsychological tests and hemodynamic and subclinical atherosclerosis parameters in all patients included in the study and, in male patients and in female patients. CHA₂DS₂-VA correlated weakly negatively (p<0.05), with MoCA (r=–0.288) and ADL (r=–0.271) and weakly positively with GDS-15 (r=0.361); also, age had a moderate negative correlation (p<0.05) with MMSE (r=–0.404) and moderate positive correlation with GDS-15 (r=0.497), and increased SBP values were associated with weak to moderate negative correlations (p<0.05) with cognitive parameters, MoCA (r=–0.305) and MMSE (r=–0.231), with functional parameters, IADL (r=–0.371) and ADL (r=–0.270), and weak positive correlations with depression, GDS-15 (r=0.351). Increased DBP values were associated with weak negative correlations (p<0.05) with MoCA (r=–0.382), ADL (r=–0.323), moderate with MMSE (r=–0.486), and weak positive correlations with GDS-15 (r=0.319). LVEF showed weak negative correlations (p<0.05) with GDS-15 (r=–0.174), weak positive correlations with MoCA (r=0.250), MMSE (r=0.267), IADL (r=0.291) and ADL (r=0.293) and NT-proBNP showed weak negative correlations (p<0.05) with MoCA (r=–0.216), MMSE (r=–0.271), IADL (r=–0.181), and ADL (r=–0.242, p=0.004), left IMT correlated (p<0.05) negatively with MoCA (r=–0.329), MMSE (r=–0.248), IADL (r=–0.313) and ADL (r=–0.276) and positively with GDS-15 (r=0.414), and right IMT (p<0.05) with MoCA (r=–0.253), MMSE (r=–0.347), IADL (r=–0.257) ADL (r=–0.329) and GDS-15 (r=0.322) .

In male patients (group A), age demonstrated a moderate positive statistically significant correlation with GDS-15 (r=0.436, p<0.001) and a weak negative correlation with MMSE (r=–0.256, p=0.035). ABI showed a moderate negative association with GDS-15 (r=–0.491, p<0.001), while weak positive statistically significant correlations were observed between ABI and MMSE, MoCA, ADL, and IADL. IMT values were moderately and negatively associated with MoCA (r=–0.458, p<0.001) and IADL (r=–0.434, p<0.001).

CHA₂DS₂-VA values demonstrated modest statistically significant correlations with GDS-15 (r=0.263, p=0.030) and ADL (r=–0.334, p=0.005). NT-proBNP levels showed weak negative correlations with MMSE (r=–0.290, p=0.016), ADL (r=–0.262, p=0.031), and IADL (r=–0.245, p=0.044). In contrast, LVEF values were weakly and positively correlated with MMSE (r=0.305, p=0.011), MoCA (r=0.270, p=0.026), and IADL (r=0.281, p=0.020).

A strong negative association was identified between GDS-15 and ADL (r=–0.666, p<0.001), while moderate negative correlations were found between GDS-15 and MMSE (r=–0.505, p<0.001), MoCA (r=–0.445, p<0.001), and IADL (r=–0.572, p<0.001). IADL values showed strong positive statistically significant associations with MMSE (r=0.651, p<0.001), MoCA (r=0.665, p<0.001), and ADL (r=0.656, p<0.001). Furthermore, MMSE demonstrated a strong positive correlation with MoCA (r=0.776, p<0.001) and ADL (r=0.671, p<0.001), while ADL was moderately positively associated with MoCA (r=0.484, p<0.001) (Table V).

Table V. Correlation analysis between MMSE, MoCA, ADL, IADL, GDS-15, age, SBP, DBP, left/right IMT, NT-proBNP, LVEF, and CHA₂DS₂-VA in male patients with T2DM/AF and cognitive decline (group A).

Values represent correlation coefficients (r); r: Spearman’s correlation coefficient; no asterisk indicates non-significant results (p>0.05); *p≤0.05; **p≤0.01; ***p≤0.001. ADL: Activities of Daily Living score; CHA₂DS₂-VA: atrial fibrillation stroke risk score; DBP: diastolic blood pressure; IADL: Instrumental Activities of Daily Living score; IMT: intima-media thickness; GDS-15: 15-item Geriatric Depression Scale; LVEF: left ventricular ejection fraction; MoCA: Montreal Cognitive Assessment; NT-proBNP: N-terminal pro-B-type natriuretic peptide; SBP: systolic blood pressure.

In female patients (group B), age showed moderate negative statistically significant correlations with MMSE (r=–0.462, p<0.001) and ADL (r=–0.567, p<0.001), as well as a moderate positive correlation with GDS-15 (r=0.573, p<0.001).

In female patients (group B), SBP values showed moderate negative statistically significant correlations with MoCA (r=–0.487, p<0.001) and IADL (r=–0.420, p<0.001). DBP values demonstrated a moderate negative correlation with MMSE (r=–0.459, p<0.001) and a weak positive correlation with GDS-15 (r=0.388, p=0.001). IMT values were moderately and positively correlated with GDS-15 (r=0.501, p<0.001).

Regarding CHA₂DS₂-VA, a moderate negative statistically significant correlation was observed with MoCA (r=–0.421, p<0.001), while a moderate positive correlation was found with GDS-15 (r=0.490, p<0.001). NT-proBNP values were weakly and negatively associated with MMSE (r=–0.258, p=0.029). In addition, LVEF values showed weak but statistically significant positive correlations with MoCA (r=0.264, p=0.025), ADL (r=0.326, p=0.005), and IADL (r=0.239, p=0.043) (Table VI).

Table VI. Correlation analysis between MMSE, MoCA, ADL, IADL, GDS-15, age, SBP, DBP, left/right IMT, NT-proBNP, LVEF, and CHA₂DS₂-VA in female patients (group B; n=72) with T2DM/AF and cognitive decline (MMSE <27).

Values represent correlation coefficients (r); r: Spearman’s correlation coefficient; no asterisk indicates non-significant results (p>0.05); *p≤0.05; **p≤0.01; ***p≤0.001; ADL: Activities of Daily Living; CHA₂DS₂-VA: atrial fibrillation-related stroke risk score; DBP: diastolic blood pressure; IADL: Instrumental Activities of Daily Living; IMT: intima-media thickness; GDS-15: 15-item Geriatric Depression Scale; LVEF: left ventricular ejection fraction; MoCA: Montreal Cognitive Assessment; NT-proBNP: N-terminal pro-B-type natriuretic peptide; SBP: systolic blood pressure.

Evaluation of factors associated with severe cognitive impairment using logistic regression analysis. Logistic regression analyses were conducted using age, SBP, DBP, ADL, and GDS-15 in the entire study population and separately in male and female patients. Model 1 included age together with SBP and DBP, whereas model 2 evaluated ADL and GDS-15.

In the total cohort of 140 patients, increasing age (OR=1.111; 95% CI=1.055-1.170; p<0.001) and higher DBP (OR=1.099; 95% CI=1.060-1.140; p<0.001) were significantly associated with MMSE <24. In male patients (n=68), age (OR=1.131; 95% CI=1.031-1.241; p=0.009) and DBP (OR=1.099; 95% CI=1.046-1.155; p<0.001) were significantly associated with MMSE <24. In female patients (n=72), age (OR=1.113; 95% CI=1.041-1.191; p=0.002) and DBP (OR=1.109; 95% CI=1.047-1.174; p<0.001) were also significantly associated with MMSE <24. The Nagelkerke R² values were 0.416 for the overall cohort, 0.434 for male patients, and 0.417 for female patients (Table VII).

Table VII. Multivariable logistic regression analysis of age and diastolic blood pressure as factors associated with severe cognitive impairment (MMSE <24) in the total T2DM/AF cohort (n=140) and separately in male (n=68) and female (n=72) patients.

n: Number of patients; OR: odds ratio; 95%CI: 95% confidence interval; T2DM: type-2 diabetes mellitus; AF: atrial fibrillation. The reported percentages were determined based on the total study population with T2DM/AF and CVRFs (n=140), as well as according to sex-specific subgroups, including male patients (n=68) and female patients (n=72).

In model 2, higher GDS-15 scores were associated with increased odds of MMSE <24 in the overall cohort (OR=1.469; 95% CI=1.178-1.832; p=0.001), in male patients (OR=1.674; 95% CI=1.108-2.528; p=0.014), and in female patients (OR=1.351; 95% CI=1.021-1.787; p=0.035). Higher ADL scores were associated with lower odds of MMSE <24 in the overall cohort (OR=0.606; 95% CI=0.450-0.817; p=0.001), in male patients (OR=0.594; 95% CI=0.353-0.998; p=0.049), and in female patients (OR=0.625; 95% CI=0.427-0.915; p=0.016). The Nagelkerke R² values were 0.381 for the overall cohort, 0.516 for male patients, and 0.265 for female patients (Table VIII).

Table VIII. Multiple logistic regression analysis of ADL and GDS-15 as factors associated with severe cognitive impairment (MMSE <24) in the total T2DM/AF cohort (n=140) and separately in male (n=68) and female (n=72) patients.

n: Number of patients; OR: odds ratio; 95%CI: 95% confidence interval; T2DM: type-2 diabetes mellitus; AF: atrial fibrillation; GDS-15: 15-item Geriatric Depression Scale; ADL: activities of daily living. Percentages were calculated based on the total cohort of patients with T2DM/AF (140 patients), the subgroup of male patients with T2DM/AF (68 patients), and the subgroup of female patients with T2DM/AF (72 patients) included in the study.

The sex-stratified estimates were not formally compared using sex-by-predictor interaction tests and should therefore not be interpreted as demonstrating statistically significant differences in effect magnitude between men and women.

Discussion

Several investigations regarding CD and dementia have indicated a higher prevalence of these conditions in women compared to men; however, it is important to note that most of these studies were performed on the general population, rather than on specifically selected patient cohorts, as in our study. This exploratory study examined sex-stratified cognitive and functional outcomes and factors associated with severe cognitive impairment in patients with both T2DM and AF. The study also examined associations of depressive symptoms and functional independence with severe cognitive impairment separately in male and female patients.

Chen et al. reported a greater risk of dementia among women than men with AF (17). Age-specific dementia diagnosis rates increase substantially with age in both sexes. Olfson et al. reported that dementia diagnosis rates were slightly higher in men at age 66 but higher in women at older ages (45). In a study of US Medicare beneficiaries, the age-standardized incidence of dementia was slightly higher in men than in women in 2015 (3.5% vs. 3.4%) and remained higher in 2021 (2.9% vs. 2.6%) (46). In the present study, increasing age was associated with higher odds of MMSE <24 in the overall cohort and in both sex-stratified analyses.

Cardiovascular risk has been associated with cognitive decline and dementia. Vishwanath et al. found that higher ASCVDRS and SCORE2-OP categories were associated with increased dementia risk in both men and women; in the highest risk category, the increase was slightly greater in women for ASCVDRS and slightly greater in men for SCORE2-OP (47). Nowell et al. reported that cardiovascular risk and obesity had their strongest association with lower grey-matter volume at ages 55-64 years in men and 65-74 years in women (48). Endothelial dysfunction has also been implicated in both Alzheimer’s disease and vascular dementia (49). In patients with diabetes, poor glycemic control and cardiovascular complications have been associated with cognitive impairment and progression to dementia (50). Hypertension is an established modifiable risk factor for cognitive impairment and dementia. In a large cluster-randomized trial, intensive blood-pressure reduction reduced the risk of all-cause dementia in individuals with uncontrolled hypertension (51). Shields et al. found that midlife hypertension was associated with poorer attention in men and with poorer attention, associative memory, and working memory in women (52). In older women, elevated systolic blood pressure and pulse pressure were associated with increased risks of mild cognitive impairment and cognitive loss (53). In older men, reverse dipping of systolic blood pressure was associated with an increased risk of dementia (54). In SPRINT-MIND participants who achieved systolic blood pressure below 130 mmHg, elevated on-treatment DBP was associated with impaired cerebral perfusion, although DBP was not independently associated with dementia or mild cognitive impairment after adjustment for covariates (55). In the present study, higher DBP was associated with increased odds of MMSE <24 in both sexes. All participants had T2DM, AF, and cognitive impairment defined by an MMSE score <27 and were receiving antihypertensive treatment.

Depressive symptoms are common in patients with cognitive impairment, but their relationship with cognitive decline may vary according to underlying neuropathology, disease stage, and sex. Tall et al. reported that the prevalence of depression in mild cognitive impairment and mild dementia differed according to neuropathology, disease stage, and sex (56). In a Korean cohort, Jahng and Kang reported a stronger association between depression and dementia in women than in men (57). Kim et al. found that recent depression predicted incident Alzheimer’s disease in women but not in men in sex-stratified analyses, although the formal sex-by-depression interaction was not statistically significant (58). Among patients with mild cognitive impairment, Wang et al. reported that worsening depressive symptoms were associated with faster cognitive decline and progression to dementia in men, whereas these associations were not observed in women (59). In the present study, greater depressive symptom burden was associated with increased odds of MMSE <24 in both male and female patients. The study population consisted of patients with T2DM, AF, and cognitive impairment, and depressive symptoms were assessed using the GDS-15, which may partly account for differences from previous studies.

Sex differences have also been described in frailty and functional decline. de Oliveira et al. reported sex-specific patterns in factors associated with increasing frailty over time (60). Ward et al. discussed the sex-frailty paradox, noting that women generally exhibit higher levels of frailty than men despite longer survival and highlighting sex as a potentially relevant modifier of the relationship between frailty and dementia (61). Guan et al. found that greater frailty was associated with greater mild behavioral impairment severity, with a stronger association in men (62). Ryu et al. found that IADL decline during progression from amnestic mild cognitive impairment to Alzheimer’s disease dementia was item- and sex-specific; “conducting home repair” had a relatively greater impact in men, whereas “using transportation” had a relatively greater impact in women (63). In the present study, lower ADL scores were associated with increased odds of MMSE <24 in both male and female patients. Because formal sex-by-predictor interaction testing was not performed, the sex-stratified estimates should not be interpreted as demonstrating a difference in effect magnitude between women and men.

Study limitations. A primary drawback of our study is the very small sample size, which may have diminished the statistical ability to identify significant differences between groups in sex-based analyses. The single-center approach may restrict the generalizability of our findings to broader and more heterogeneous groups. The absence of longitudinal follow-up limits our capacity to examine the temporal progression of neuropsychological parameters and to ascertain whether sex-related differences may arise or fluctuate over time. Moreover, we did not examine the potential interaction between sex and other pertinent factors, such as transient or asymptomatic episodes of atrial fibrillation, nor did we evaluate whether treatment for T2DM and CVD would yield varying outcomes in men and women. The hormonal status, particularly in postmenopausal women, was not considered, which may constitute an additional variable for future research. Educational attainment was not systematically recorded and could therefore not be included as a covariate in the analyses. Because MMSE performance is influenced by education, the absence of adjustment for educational attainment represents an important potential source of residual confounding. This limitation is particularly relevant to the interpretation of sex-stratified cognitive findings, and the observed associations should therefore not be interpreted as demonstrating intrinsic sex-related differences in cognitive performance.

Prospective research. Future extensive, randomized, multicenter studies should prioritize sex-specific analyses to enhance the understanding of potential disparities in cognitive performance, clinical features, and treatment responses between males and females. Longitudinal studies that include extensive cognitive evaluations in patients with T2DM and AF would enhance understanding of cognitive decline progression and aid in the identification of sex-specific prevention and treatment methods.

Future longitudinal studies should investigate if the progression of CD varies by sex and if these variations are affected by factors such as age, hormonal status, presence or escalation of CVR, duration of T2DM, glycemic management, and type of AF. It would be beneficial to investigate whether therapeutic measures, such as anticoagulant therapy, rate or rhythm control methods, and lifestyle modifications, exert varying cognitive effects in men and women.

Integrating sex-specific analysis may enhance the formulation of tailored preventive and treatment strategies to mitigate CD in individuals with T2DM and AF.

Conclusion

Evaluating cognitive function in individuals with T2DM and AF may facilitate the early identification of cognitive impairment. In this exploratory cross-sectional study, increasing age, higher DBP, and greater depressive symptom burden were associated with increased odds of MMSE <24 in both men and women with T2DM and AF. Greater functional independence, reflected by higher ADL scores, was associated with lower odds of MMSE <24 in both sexes. Because formal sex-by-predictor interaction testing was not performed, the sex-stratified estimates should not be interpreted as demonstrating statistically significant differences in effect magnitude between men and women. Further longitudinal studies are needed to determine whether cognitive trajectories differ by sex in this population.

Conflicts of Interest

The Authors declare no conflicts of interest in relation to this study.

Authors’ Contributions

Conceptualization, M.M., D.F.L., C.T., and A.G.M.; methodology, M.M., C.T., and A.G.M.; software, M.M., C.T., N.B., and A.G.M.; validation, M.M., D.F.L., C.T., and A.G.M.; formal analysis, M.M., C.T., and A.G.M.; investigation, M.M., D.F.L., C.T., N.B., and A.G.M.; resources, M.M., D.F.L., C.T., and A.G.M.; data curation, M.M., C.T., and A.G.M.; writing–original draft preparation, M.M., D.F.L., C.T., and A.G.M.; writing–review and editing, C.A.M., M.M., C.T., and A.G.M.; visualization, M.M., D.F.L., C.T., and A.G.M.; supervision, C.A.M., M.M., C.T., and A.G.M.; project administration, M.M., C.T., and A.G.M.; funding acquisition, not applicable. All Authors have read and agreed to the published version of the manuscript.

Acknowledgements

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of the Municipal Emergency Hospital Timisoara, Romania (No. E1518, approval date: 17 March 2021). Informed consent was obtained from all subjects involved in the study. Written informed consent has been obtained from the patient(s) to publish this article.

Funding

This research received no external funding. Internal funding: We would like to acknowledge the “Victor Babeș” University of Medicine and Pharmacy, Timișoara, for their support in covering the costs of publication for this research paper.

Data Availability

The data supporting the findings of this study are available from the first or corresponding author upon reasonable request, subject to applicable ethical and institutional requirements. The statistical analysis syntax/code used for the analyses can also be made available by the first or corresponding author upon reasonable request.

Artificial Intelligence (AI) Disclosure

No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.