Anthropometric Status, Dietary Diversity, and Diet Quality in Relation to Cognitive Performance in Older Turkish Adults: A Cross-Sectional Study

1. Introduction

According to United Nations data, the global population is rapidly aging. Individuals aged 65 years and older accounted for approximately 10% of the world's population in 2022, and by 2100, they are expected to comprise nearly one-quarter of the global population [1]. In the context of this rapidly aging population, maintaining cognitive function and preventing age-related cognitive decline have become critical public health priorities. Recent years have highlighted those pharmacological interventions aimed at improving quality of life or preserving cognition in individuals with normal cognitive function, mild cognitive impairment (MCI), or dementia lack strong evidence, whereas the importance of non-pharmacological interventions and multidisciplinary care has been emphasized [2,3]. Non-pharmacological strategies such as dietary habits may influence biological aging and are associated with age-related conditions, including diabetes, cardiovascular disease, and dementia [4]. Identifying modifiable nutritional and anthropometric factors related to cognitive performance is important in clinical settings because both are routinely measured in outpatient care and can be targeted for intervention.

Healthy dietary patterns have been shown to support brain health and potentially reduce the risk of cognitive decline [5,6]. Observational studies have reported that higher intake of nutrients such as vitamins, minerals, carotenoids, fatty acids, and fiber may be associated with a reduced or increased risk of cognitive impairment and/or decline [7,8]. Findings from the Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability (FINGER) showed that higher adherence to healthy dietary patterns in older individuals was associated with better overall cognitive performance, and that improvements in diet quality may be linked to positive changes, particularly in executive functions [9]. Beyond overall diet quality, dietary diversity represents another dimension of diet, reflecting the variety of food groups consumed; the two constructs are not interchangeable, since a diet may be varied without being of high quality. The relationship between cognitive function and dietary diversity in older adults has not been fully clarified. A study conducted on older adults living in Japan found an association between high dietary diversity and low cognitive impairment, whereas a study conducted in Türkiye found no association between dietary diversity and cognitive impairment [10,11]. Jiang et al. reported that a higher Chinese Healthy Eating Index score was associated with better cognitive function in older adults, with psychological balance and depressive symptoms acting as chain mediators [12].

In addition to dietary factors, anthropometric status may be associated with cognitive performance in later life. Studies suggest an association between obesity, measured by body mass index (BMI) and especially waist–hip ratio (WHR) and waist circumference (WC), and lower cognitive performance in older adults [13,14]. However, it remains unclear whether general adiposity, (indexed by BMI) and central adiposity, (indexed by WC and WHR), provide comparable information on cognitive performance, since these indicators are frequently used interchangeably, despite capturing different aspects of body composition. Although research on the relationship between obesity and cognitive impairment in older adults has increased, conflicting results persist [15].

Although evidence regarding the associations of dietary patterns and anthropometric status with cognitive function is increasing, studies that evaluate these factors simultaneously remain limited. This distinction matters because adiposity is itself partly determined by long-term dietary intake; unless diet quality, dietary diversity, and anthropometric indicators are modelled jointly, it cannot be determined whether diet is associated with cognitive performance independently of body size, or whether an apparent dietary association is instead explained by adiposity. Evidence from Türkiye is particularly scarce. Previous studies conducted in Türkiye have examined nutritional status and cognitive function; however, evidence combining dietary diversity, overall diet quality, and multiple anthropometric indicators remains insufficient [16,17]. This setting is relevant because obesity is highly prevalent among older adults in Türkiye, and habitual dietary patterns differ from those of the Western cohorts in which most of the available evidence has been generated. Accordingly, this study aimed to examine the relationships of anthropometric status, dietary diversity, and dietary quality with cognitive performance in older adults attending a nutrition and dietetics outpatient clinic in Türkiye. We hypothesized that higher adiposity, lower diet quality, and lower dietary diversity would each be independently associated with poorer cognitive performance after accounting for age, sex, and years of education.

2. Materials and Methods

2.1. Study Design and Participants

This cross-sectional study conducted between April 2026 and June 2026 among individuals aged 65 years and older who applied to the Nutrition and Diet Outpatient Clinic of Kastamonu Physical Therapy and Rehabilitation Hospital. The sample size was calculated using G*Power 3.1.9.7. An a priori power analysis for multiple linear regression was performed with the following parameters: a medium effect size (f² = 0.15), a significance level (α = 0.05), statistical power (1 − β = 0.80), and six independent variables. The analysis indicated that at least 98 participants should be included in the study. This study used a non-probability sequential sampling method. Of the 881 older adults who attended the Nutrition and Diet Outpatient Clinic during the study period, 441 were excluded for not meeting the inclusion criteria. Of the remaining 440 eligible individuals, 317 declined to participate. A total of 123 participants were enrolled, of whom 6 were excluded because of missing or incomplete data. Thus, 117 older adults were included in the final analysis.

This study included individuals aged 65 years and older who voluntarily participated, were able to complete cognitive tests and nutritional assessments, had no acute health condition preventing measurements, and who applied to the Nutrition and Diet Outpatient Clinic of Kastamonu Physical Therapy and Rehabilitation Hospital. Individuals with physician-diagnosed dementia (including Alzheimer’s disease), psychiatric disorders, or a history of acute infection, surgery, major trauma, or hospitalization within the past two weeks were excluded.

The study was approved by the Non-Interventional Clinical Research Ethics Committee of Kastamonu University (Approval No: 2026-63, Date: 19 March 2026) and was conducted in accordance with the Declaration of Helsinki. Institutional permission was obtained from the Kastamonu Provincial Health Directorate for this research to be conducted at the Kastamonu Physical Therapy and Rehabilitation Hospital (Document number: E-44008972-770, Date: 25 February 2026). Additionally, this article was prepared in accordance with the STROBE checklist.

2.2. Data Collection

A face-to-face questionnaire has six sections. The first section included information on participants' sociodemographic characteristics (age, gender, marital status, education level, cohabiting partner). The second section includes anthropometric data such as body weight, height, waist and hip circumferences, along with body fat analysis results. The third section gathered information on health status, including the presence of chronic diseases, appetite, dietary habits, and smoking and alcohol use. The fourth section assessed participants' food intake using a 24-hour dietary recall. The fifth section included cognitive performance, assessed using the Montreal Cognitive Assessment (MoCA). The sixth section assessed physical activity level using the International Physical Activity Questionnaire Short Form (IPAQ-SF).

Participants' educational information was collected through self-reporting during face-to-face interviews. Educational level was determined by asking participants to indicate the most recent level of education they completed and received a diploma from, and was classified: primary school, secondary -high school, and university graduate. The number of years of education was recorded as the total duration of formal education completed.

2.3. Anthropometric Measurements

Height was measured using a portable stadiometer (SECA 213), with participants barefoot and wearing light clothing. Body weight and body composition were assessed using a Tanita BC-401 bioelectrical impedance analysis device (Tanita Corporation, Tokyo, Japan). Measurements were taken under standardized conditions between 08:00 and 10:00 a.m. following an overnight fast of at least 12 h. Participants were instructed to avoid vigorous physical activity and alcohol consumption for 24 h and to avoid caffeine intake for at least 12 h prior to the assessment. They were also asked to void their bladder immediately before the measurement and were assessed barefoot, wearing light clothing, with all metal accessories removed [18]. BMI was calculated as body weight (kg) divided by height squared (m²). Waist and hip circumferences were measured using a non-elastic tape measure; WC was taken at the midpoint between the last rib and the iliac crest, and hip circumference was measured at the widest part of the greater trochanter. WHR was calculated as waist circumference divided by hip circumference, with substantially increased risk defined as ≥0.90 for men and ≥0.85 for women [19].

2.4. Dietary Intake

Dietary intake was assessed using the 24-hour dietary recall method include one weekday. Participants were asked to recall and report in detail all foods and beverages consumed during the previous 24 hours. Nutrition Photo Catalog Measurements and Amounts were used to ensure accurate reporting of portion size [20]. The percentages of energy and nutrient intake for older adults were analyzed using the BEBIS (version 9.0) software [21]. Dietary diversity and dietary quality scores were calculated using data obtained from participants' 24-hour retrospective food intake records.

2.5. Dietary Diversity

The evaluation of dietary diversity was based on the 2013 report of the Food and Agriculture Organization of the United Nations [22]. According to this report, the Dietary Diversity Score (DDS) is defined as the number of food groups consumed within 24 h. The diet was classified according to the nine food groups recommended by the FAO: (1) starchy staples; (2) dark green leafy vegetables; (3) vitamin A-rich fruits and vegetables; (4) other fruits and vegetables; (5) organ meats; (6) meat, poultry, and fish; (7) eggs; (8) legumes, nuts, and seeds; and (9) milk and milk products. Tea, sugar, and sweets were not included in the DDS calculations. Participants were categorized according to their DDS as follows: poorly diversified diet (DDS 0–3), moderately diversified diet (DDS 4–5), and diversified diet (DDS 6–9).

2.7. Diet Quality

Diet quality was assessed using the Healthy Eating Index-2020 (HEI-2020), which evaluates adherence to the Dietary Guidelines for Americans, 2020–2025. HEI-2020 scores were calculated from dietary intake records using the simple scoring algorithm. Intakes of the food group components were expressed per 1,000 kcal of energy intake (energy-density approach), whereas fatty acids were evaluated using the ratio of unsaturated to saturated fatty acids and added sugars and saturated fats were expressed as percentages of total energy intake. The HEI-2020 includes 13 components, grouped into two categories: adequacy (total fruits, whole fruits, total vegetables, greens and beans, whole grains, dairy, total protein foods, seafood and plant proteins, and fatty acids) and moderation (refined grains, sodium, added sugars, and saturated fats). Each component is scored based on intake levels: adequacy components receive higher scores for greater consumption, while moderation components receive higher scores for lower consumption. The total HEI-2020 score ranges from 0 to 100, with higher scores reflecting better diet quality [23].

2.8. Montreal Cognitive Assessment

The MoCA, developed to evaluate mild cognitive impairment, assesses various cognitive domains including executive functions, visuospatial skills, memory, language, attention and concentration, abstract thinking, calculation, and orientation [24]. The MoCA was adapted into Turkish and validated by Selekler et al. [25]. The maximum total score is 30. Education-specific cut-off scores proposed by Kaya et al. were used to classify cognitive performance. Participants with MoCA scores below 18 among those with ≤5 years of education, below 21 among those with 6–11 years of education, and below 23 among those with ≥12 years of education were classified as screening positive for possible MCI. Scores at or above the respective cut-off values were classified as normal cognitive performance [26]. The MoCA was administered face-to-face by trained researchers using the validated Turkish version. Because education-specific cut-off points were used, the standard one-point adjustment for participants with 12 or fewer years of education was not applied.

2.9. Physical Activity

The IPAQ-SF was designed to assess physical activity and sedentary behavior in adults, and is available in both short and long versions. Its international validity and reliability were established by Craig et al. [27], and the Turkish adaptation was validated by Öztürk [28]. In this questionnaire, physical activities are considered if performed for at least 10 min at a time. Participants were asked about vigorous and moderate physical activity, walking, and daily sitting time over the past seven days. The durations of vigorous, moderate, and walking activities were converted to metabolic equivalents (METs; 1 MET = 3.5 mL/kg/min) using the following calculations; the total physical activity score (MET-min/week) was then obtained:

Walking, moderate, and vigorous physical activity MET-min/week scores were calculated by multiplying the MET coefficient of the respective activity (3.3; 4.0 and 8.0) by the activity duration and the number of days per week, respectively. The total physical activity score was obtained by summing the walking, moderate, and vigorous activity scores. Based on the TPAS, physical activity levels were classified as follows: TPAS < 600 MET-min/week, low physical activity; TPAS = 600–3000 MET-min/week, moderate physical activity and TPAS > 3000 MET-min/week, high physical activity level.

2.10. Statistical Analysis

The data obtained in the study were analyzed using IBM SPSS Statistics 25.0 (IBM Corp., Armonk, NY, USA). Descriptive statistics, including the arithmetic mean, standard deviation, median, minimum, maximum, percentages, and frequency distributions, were used to analyze the data. In this study, cognitive performance score was considered as the dependent variable, while age, gender, years of education, anthropometric measurements, diet diversity score, and diet quality score were evaluated as independent variables. The normality of the variables was assessed using the Kolmogorov–Smirnov test, and Q–Q plots; histograms were also used for visual inspection. Homogeneity of variance was tested for group comparisons using the Levene test. For comparisons between two groups, the Student’s t-test was applied to normally distributed variables, while the Mann–Whitney U test was applied to non-normally distributed variables. Categorical variables were compared using Pearson's chi-square test. When expected cell frequencies were insufficient, Fisher's exact test was applied when expected cell frequencies were insufficient. When an overall association was significant for variables with more than two categories, post-hoc comparisons were performed using adjusted standardised residuals with Bonferroni correction. Spearman’s correlation analysis was used to examine the relationships and strengths among two or more numerical variables. Hierarchical regression analysis was performed to identify factors associated with cognitive performance among older adults. Variables were entered into the hierarchical regression in three blocks: first, sociodemographic factors established as determinants of cognitive performance (age, sex, and years of education); second, the anthropometric indicator; and third, the dietary indicators (DDS and HEI-2020), so that the incremental contribution of each domain could be evaluated. Before interpretation, the assumptions of linear regression were examined: linearity and homoscedasticity, assessed by inspection of residual plots; normality of residuals, assessed by normal probability plots; independence of residuals, assessed by the Durbin-Watson statistic, and multicollinearity by tolerance and variance inflation factors. Physical activity level was assessed, but since no statistically significant difference was found between the groups, it was not considered a potential confounding variable and was not included in the regression models. All analyses were performed on complete cases; participants with missing data were excluded before analysis. A significance level of p < 0.05 was used as the threshold for statistical significance in all analyses.

3. Results

3.1. Participant Characteristics

A total of 117 older adults, of whom 36 were men and 81 were women were initially assessed. General characteristics of the participants are presented in Table 1. Accordingly, the percentage of married older adults was higher in men than in women (91.7% vs. 65.4%, respectively; p<0.05). However, the percentage of older adults living alone was higher in women than in men (25.9% vs. 2.8%, respectively; p<0.05). Furthermore, the prevalence of hypertension was found to be higher in women than in men (70.4% vs. 47.2%; p<0.05). The percentage of older adults with obesity was higher in the women group than in the men group (65.4% vs. 25.0%; p<0.01). There were no statistically significant differences between the groups in median age, education level, chronic disease prevalence, smoking status, appetite, meal frequency, physical activity level, and MoCA classification (p > 0.05).

3.2. Anthropometric and Dietary Characteristics According to Cognitive Performance

General characteristics of participants according to the MoCA classification are presented in Table 2. Older adults with possible MCI differed significantly from those with normal cognitive performance with respect to education level, appetite status, and the number of snacks. Primary school education was more common in the possible MCI group (93.5% vs. 74.5%, p = 0.016). Snack frequency differed significantly between groups (p < 0.001); not snacking was more common in the group with normal cognitive performance, whereas consuming two snacks was more common in the possible MCI group. No significant differences were observed for age, sex, marital status, living arrangement, chronic disease status, specific chronic diseases, smoking, appetite status number of main meals, and physical activity level.

3.3. Anthropometric and Dietary Characteristics According to Cognitive Performance

Anthropometric measurements and dietary characteristics according to the MoCA classification are presented in Table 3. Participants who had possible MCI had a significantly higher mean BMI than those with normal cognitive performance (32.67 ± 5.78 vs. 29.54 ± 3.39 kg/m², p < 0.001). Energy intake per kilogram of body weight was significantly lower in the possible MCI group than in the normal cognitive performance group (19.26 (14.85–25.46) vs. 22.64 (18.87–28.80) kcal/kg/day, p = 0.029). No significant differences were observed between the groups for body fat percentage, WC, WHR, DDS, HEI-2020 score, total energy intake, total protein intake, protein intake per kilogram of body weight, or the percentage of energy derived from protein (all p > 0.05). When BMI was dichotomized as without obesity (<30.0 kg/m²) and obesity (≥30.0 kg/m²), the prevalence of obesity was significantly higher in the possible MCI group than in the normal cognitive performance group (62.9% vs. 41.8%; Fisher’s exact test, p = 0.027).

3.4. Correlations with MoCA Scores

Spearman correlations among MoCA score, anthropometric measurements, nutrient intakes, and diet quality indicators are presented in Table 4. Correlation analysis showed that the MoCA score was significantly negatively correlated with BMI (rs = −0.303, p = 0.001), indicating that higher BMI was associated with lower cognitive performance. No statistically significant correlations were observed between the MoCA score and body fat percentage, WC, WHR, DDS, HEI-2020 score, total or body weight-adjusted energy intake, total or body weight-adjusted protein intake, or the percentage of energy derived from protein (all p > 0.05).

3.5. Factors Associated with MoCA Scores

A hierarchical multiple linear regression analysis examining factors associated with MoCA scores is presented in Table 5. Model 1, which included age, sex, and years of education, was statistically significant and explained 17.8% of the variance in MoCA scores (R² = 0.178, adjusted R² = 0.156; F(3,113) = 8.140, p < 0.001). Adding of BMI to Model 2 significantly increased the explained variance by 3.6% (ΔR² = 0.036, ΔF = 5.179, p = 0.025). The inclusion of the DDS and HEI-2020 in Model 3 increased the explained variance by a further 2.2%; however, this improvement was not statistically significant (ΔR² = 0.022, ΔF = 1.591, p = 0.208). The final model was statistically significant and explained 23.6% of the variance in MoCA scores (R² = 0.236, adjusted R² = 0.194; F(6,110) = 5.668, p < 0.001). In the final model, more years of education were associated with higher MoCA scores (B = 0.795, β = 0.375, 95% CI 0.436 to 1.153, p < 0.001), whereas higher BMI was associated with lower MoCA scores (B = −0.282, β = −0.225, 95% CI −0.510 to −0.054, p = 0.016). Age, sex, dietary diversity, and HEI-2020 were not significantly associated with MoCA scores.

3.6. Sensitivity Analyses

Sensitivity analyses were conducted by separately replacing BMI with WC and WHR (Table 6). In the primary model, the addition of BMI significantly increased the explained variance (ΔR² = 0.036, p = 0.025), and BMI remained negatively associated with the MoCA score in the final model. In contrast, neither WC nor WHR significantly increased the explained variance (ΔR² = 0.005, p = 0.407; ΔR² = 0.003, p = 0.514, respectively). The dietary block did not significantly improve any of the models. Neither dietary diversity nor HEI-2020 was significantly associated with MoCA scores.

4. Discussion

This study investigated the relationships between anthropometric status, dietary diversity, and nutritional quality, and cognitive performance in older Turkish adults. 53.0% of participants had a positive screening result for possible mild cognitive impairment (MCI) according to the MoCA. One of the main findings of the study was that higher body mass index (BMI) was independently associated with lower MoCA scores after controlling for age, gender, years of education, and nutritional indicators. In contrast, waist circumference and waist-to-hip ratio, used in sensitivity analyses, did not show a significant relationship with cognitive performance. Furthermore, years of education was found to be the strongest independent predictor of MoCA scores, with longer years of education being associated with better cognitive performance. However, no significant relationship was found between cognitive performance and DDS and HEI-2020 scores, which reflect dietary diversity and nutritional quality.

Older adults participating in the present study (median age 71 years), 53.0% screened positive for possible MCI based on the MoCA. This rate was reported as 70.9% in the study conducted by Ye et al. (mean age 71.18 ± 4.97 years) [29] and as 35.6% in the study conducted by Wu et al. (mean age 68.9 years) [30]. In the study by Ye et al., [29] the use of a higher MoCA cut-off point (<26) may have contributed to the higher prevalence of possible MCI. Similar to the present study, Wu et al. [30] used education-specific MoCA cut-off points. However, the hospital-based design of the present study, the lower educational level of our sample, and its higher mean age may have contributed to the higher prevalence of possible MCI. Differences in countries, cultures, health status, and sampling methods may produce different results when cognitive performance is assessed. In this study, years of education showed the strongest independent association with MoCA scores, with a longer duration of education being associated with better cognitive performance. In addition, a low educational level was more prevalent in the possible MCI group. Our findings are consistent with previous studies reporting positive correlations between MoCA scores and duration of education [29,31]. However, the association between education and higher MoCA scores may not be attributable entirely to better cognitive health; the sensitivity of the MoCA to educational level may also partially explain this association [32].

The negative association between BMI and MoCA scores remained significant after controlling for age, sex, years of education, and dietary indicators in our study. Furthermore, participants with obesity had significantly lower MoCA scores than those without obesity. In contrast, neither WC nor WHR was associated with MoCA scores in the sensitivity analyses. Although Phirom et al. [15] founded that BMI, WC, and WHR may be associated with poor cognitive performance, they reported that the available evidence was of low certainty and highly uncertain. Another study conducted among older adults observed a positive linear association between WC and cognitive impairment but found no significant association with BMI [33]. Another study based on data from the China Health and Retirement Longitudinal Study reported a dose-dependent association between BMI and cognitive impairment. Being underweight was identified as a risk factor for the development of cognitive impairment, whereas being overweight or obese could reduce the likelihood of cognitive impairment, particularly among older women [34]. In contrast, Alvarez et al. reported that higher WC and BMI were associated with cognitive impairment assessed using the MoCA [35]. These differences across studies suggest that the relationship between anthropometric indicators and cognitive performance in older adults is complex and may vary according to the indicator used, sample characteristics, and comorbid health conditions. Potential mechanisms through which obesity may contribute to reduced cognitive performance include comorbidities, genetic factors, and inflammatory processes. Although some studies have suggested that this association may arise through comorbidities such as hypertension, cardiovascular diseases, and diabetes, the evidence supporting this hypothesis has not been definitively confirmed [36]. However, some alternative biological and methodological explanations should also be considered in interpreting the relationship between BMI and cognitive performance. First, BMI may not fully reflect body composition in older adults; conditions such as sarcopenic obesity, where decreased muscle mass and increased fat mass occur together, may be associated with cognitive health beyond measurements based solely on body weight. Recent systematic reviews and meta-analyses have shown that sarcopenic obesity is associated with an increased risk of cognitive dysfunction and dementia [37,38]. Second, lifetime changes in body weight and the effects of obesity in middle and late life may differ on cognitive health. Longitudinal studies report that the relationship between adiposity and cognitive decline may vary depending on age, the direction of weight change, and the life stage assessed [39,40]. Furthermore, due to the cross-sectional design of our study, the possibility of reverse causality cannot be ruled out; Early cognitive decline may have contributed to the observed relationship by influencing physical activity levels, dietary habits, and body weight. Indeed, previous studies highlight the potentially bidirectional relationship between obesity and cognitive function [41,42]. Finally, the possibility of confounding can no longer be entirely ruled out, as depressive symptoms, frailty status, and other unmeasured health factors may be associated with both BMI and cognitive performance. In particular, the relationship between obesity and frailty, as well as the close association of frailty with cognitive impairment, may complicate the interpretation of the BMI-cognition relationship observed in older adults [43,44]. Similarly, in the present study, the prevalence of comorbidities did not differ between the MoCA groups. Considering the rapidly increasing prevalence of obesity worldwide and in Türkiye and the growing prevalence of cognitive performance problems in parallel with population ageing, interventions aimed at preventing obesity are becoming increasingly important.

In this study, no significant association was found between the DDS, an indicator of dietary diversity, and MoCA scores. Previous studies have reported that low dietary diversity is associated with poorer cognitive performance [45,46]. Increasing dietary diversity, which refers to the variety or number of different food groups consumed over a given period, may contribute to the maintenance of brain health by supporting adequate nutrient intake and may potentially play a role in reducing the risk of neurodegenerative diseases. However, contrary to some findings in the literature, dietary diversity was not associated with cognitive performance in the present study. This discrepancy may be attributable to differences in the methods used to calculate dietary diversity, the cognitive assessment instruments employed, and sample characteristics. Furthermore, assessing dietary intake using a single-day dietary record may not have fully reflected the participants’ usual dietary diversity. The relatively high DDS scores and limited variability among the participants may also have reduced the ability to detect a possible association.

In the present study, the HEI-2020 score, an indicator of diet quality, was not significantly associated with the MoCA score. Using data from the Microbiome in Aging Gut and Brain (MiaGB) Consortium Cohort, Arikawa et al. examined the associations of the inflammatory potential of the diet and overall diet quality with cognitive impairment [47]. The researchers reported no significant differences in Dietary Inflammatory Index or HEI-2020 scores between participants with and without cognitive impairment. Similarly, another study examining the associations of the Mediterranean–DASH Intervention for Neurodegenerative Delay (MIND) diet and HEI-2020 with cognitive performance found that higher MIND diet scores were associated with better global cognitive performance and memory, whereas HEI-2020 scores were not associated with global cognition, memory, or executive function [48]. In contrast, Wessinger et al. reported that higher HEI-2020 scores were associated with better executive function and working memory among participants who were not carriers of the APOE-ε4 allele, whereas no significant association was observed among APOE-ε4 carriers [49]. Unlike our study, another research conducted on middle-aged and older adults evaluated the relationship between cognitive function and healthy lifestyle components (diet quality, physical activity, and smoking). This research reported that higher levels of physical activity were associated with higher scores in overall cognitive performance and some cognitive domains, and that individuals with healthy lifestyle characteristics performed better cognitively. Although this study considered lifestyle factors as a whole and also evaluated physical activity, the results support the importance of nutrition and other lifestyle behaviors in maintaining cognitive health. Similarly, our study demonstrates that dietary diversity and quality are associated with cognitive performance in older Turkish adults, revealing that nutrition plays a significant role in maintaining cognitive health [50]. These findings suggest that the association between overall diet quality and specific cognitive domains may vary according to genetic susceptibility. The fact that the HEI-2020 was developed to assess adherence to general healthy eating recommendations rather than as an index specifically designed for cognitive health may be one possible explanation for the absence of a significant association with cognitive performance in the present study.

In comparisons of dietary characteristics between the groups, energy intake per kilogram of body weight was higher among participants with normal cognitive performance (22.64 kcal/kg/day) than among those in the possible MCI group (19.26 kcal/kg/day). However, no significant differences were found between the groups in terms of total energy or protein intake. Similarly, Doorduijn et al. [51] reported no significant differences in total energy, protein, carbohydrate, or fat intake between older adults with MCI and controls. Park et al., [52] however, showed that after controlling for confounding factors, older adults with energy intakes below the recommended level were more likely to have cognitive impairment than those whose energy intakes met the recommended level. Although these findings indicate the potential importance of adequate energy intake in maintaining cognitive health, further studies are needed to clarify the relationship between energy intake and cognitive decline in older adults and the potential effects of energy restriction. In the present study, good appetite was more prevalent in the normal cognitive performance group than in the possible MCI group, whereas normal appetite was more prevalent in the possible MCI group. This finding suggests that good appetite may be associated with normal cognitive performance. Similarly, a study conducted among individuals with diabetes aged over 60 years found that participants with cognitive impairment were more likely to report a reduced appetite [53]. In contrast, no significant association between poor appetite and cognitive function was observed among community-dwelling individuals aged 55 years and older [54]. Although our study did not evaluate physical activity, previous research highlights the importance of the interaction between cognitive health and lifestyle factors. A study of community-bound older adults in China showed that the protective effect of physical exercise on cognitive impairment varied depending on certain dietary habits, such as the consumption of fruits, dairy products, and legumes. While differing from our study in terms of research design and evaluated variables, these findings support the idea that nutritional status may be related to cognitive health in older individuals. Similarly, our study reveals that dietary diversity and nutritional quality are related to cognitive performance, pointing to the importance of nutrition in maintaining cognitive health in old age [55]. Differences across studies may be attributable to the age and health characteristics of the samples, the methods used to assess appetite, and the confounding factors controlled for in the analyses.

This study has several limitations. First, its cross-sectional design precludes causal inference and determination of the direction of the observed associations; reverse causation is possible, in that early cognitive decline may itself alter dietary behaviour and body weight. Second, the relatively small sample size of 117 participants and recruitment from a single centre may limit the generalisability of the findings. Moreover, 317 of the 440 eligible individuals declined to participate, and this high refusal rate may have introduced selection bias, the direction of which cannot be determined from the available data. Individuals who agreed to participate in the study were likely to be more concerned about their health, have better access to healthcare, or have relatively preserved cognitive function. In this case, the prevalence of cognitive impairment may have been underestimated. On the other hand, if individuals seeking healthcare for cognitive complaints were more willing to participate in the study, the prevalence may have been found to be higher. Similarly, measurement errors due to single-day food intake records may have led to misclassification of participants' usual dietary habits, weakening the true associations between DDS and HEI-2020 and cognitive performance. Third, dietary intake was assessed using a single-day dietary record, which may not reflect usual intake and is susceptible to recall bias; this is a particular limitation for the HEI-2020, which was designed to capture habitual dietary patterns, and the resulting measurement error may have attenuated any true association with cognitive performance. Fourth, physical activity was self-reported, and the MoCA is a screening instrument rather than a diagnostic tool. Finally, potential confounders such as depressive symptoms, medication use, and overall comorbidity burden were not included in the regression models.

The sample characteristics of the study should be taken into account when evaluating the generalizability of the findings. Since the participants consisted of older adults who attended the geriatric outpatient clinic of a single university hospital, the results may not fully represent all elderly individuals living in the community. The magnitude of the relationships may vary, especially in rural areas or community-based samples with different characteristics in terms of health care-seeking behaviors, education level, socioeconomic status, and chronic disease burden. However, it is thought that the study provides important preliminary information on the relationships between cognitive performance, obesity and nutrition among the rapidly growing elderly population in Türkiye. The findings need to be confirmed with data obtained from different geographical regions and community-based samples.

Nevertheless, this study also has several strengths. Education-specific MoCA cut-off points appropriate for the Turkish population were used. Diet quality was assessed comprehensively using both the DDS and the HEI-2020. Furthermore, hierarchical regression analyses and sensitivity analyses using alternative anthropometric indicators strengthened the robustness of the findings. Larger longitudinal studies are needed to clarify the temporal and potentially causal relationships among the factors examined.

5. Conclusions

In conclusion, obesity was more prevalent among older adults with possible MCI than among those with normal cognitive performance, and a higher BMI was independently associated with lower MoCA scores. Dietary diversity and HEI-2020 scores were not associated with MoCA scores. The association with BMI was not replicated when WC or WHR was used, indicating that the anthropometric findings should be interpreted cautiously. The duration of education showed a strong positive association with cognitive performance. Longitudinal studies using repeated dietary assessments and more comprehensive measures of body composition are needed to clarify these relationships