Association between healthy eating index and sleep quality and duration in older adults: findings from the NeLSA cohort Study, Iran Authors: Haniyeh Khalafinejad, Fatemeh Sadat Hashemi Javaheri, Amin Sharifi, Seyed Mostafa Arabi, Mahnaz Amini, Reza Rezvani Published: 2026 ABSTRACT BackgroundPoor sleep and malnutrition are prevalent issues among older adults, contributing to negative outcomes such as depression, sarcopenia, and reduced quality of life. A bidirectional relationship has been proposed between diet and sleep. This study aimed to examine the association between the Healthy Eating Index (HEI) and both sleep quality and duration in older adults.MethodsThis cross-sectional analysis was conducted using baseline data from the Neyshabur Elderly Longitudinal Study (NeLSA, 2016–2022), including 2,026 adults aged ≥ 60. Dietary data were collected via a validated food frequency questionnaire, and HEI-2015 scores were calculated. Sleep quality was assessed using the Pittsburgh Sleep Quality Index (PSQI), with a score > 5 indicating poor sleep quality. Sleep duration was self-reported and categorized as adequate (≥ 7 h) or insufficient (< 7 h). Multivariable logistic regression was used to assess associations between HEI quartiles and sleep outcomes, adjusting for demographics, BMI, smoking, education, and chronic diseases.ResultsParticipants had a mean age of 69.1 ± 7.6 years; 55.5% were women. Women reported poorer sleep quality and shorter sleep duration than men (P < 0.001). There was no statistically significant association between HEI and sleep quality or duration in the overall population or by gender (P > 0.05). However, a borderline non-significant inverse association was observed between HEI and insufficient sleep duration in older women. (Q4 vs. Q1, OR: 0.709; 95% CI: 0.502–1.003; P = 0.052).ConclusionHigher HEI scores were not independently associated with sleep quality or duration among older adults. Although higher diet quality showed a non-significant inverse association between better sleep in women, the relationship may be influenced by other health-related factors. Further longitudinal and interventional studies are recommended. Introduction The global and national populations of older adults are steadily increasing, with projections indicating a doubling of those aged 60 and above in the coming decades [1, 2]. Aging brings physiological and psychological changes that predispose individuals to nutritional problems [3] and sleep disturbances [4, 5]. Malnutrition and poor sleep are common in older adults and are linked to outcomes such as depression, sarcopenia, and lower quality of life [6–8]. In epidemiological studies, it has been reported that approximately 30–45% of adults experience symptoms of insomnia, and this prevalence increases to nearly 50% among older adults [9, 10]. Common sleep disorders in the elderly include insomnia, circadian rhythm disturbances, excessive daytime sleepiness, sleep apnea, and restless legs syndrome [11]. Although some sleep changes are part of normal aging, primary and secondary sleep disorders are also highly prevalent in this age group [12]. Nutrition plays a key role in elderly health, especially as age-related changes and chronic diseases increase the risk of malnutrition [13]. Growing evidence suggests a bidirectional link between diet and sleep. For example, short sleep is associated with lower intake of healthy foods (e.g., fruits, vegetables, whole grains, lean proteins) and higher intake of unhealthy options (e.g., refined carbs, added sugars) [14–16]. Sleep deprivation may also alter appetite-regulating hormones like ghrelin and leptin, increasing caloric intake and cravings for high-fat, high-sugar foods [17, 18]. Thus, insufficient sleep may promote poor dietary habits, creating a vicious cycle of unhealthy behaviors [19]. Several studies have reported associations between higher Healthy Eating Index (HEI) scores and better sleep [20]. A national U.S. study found that individuals with higher HEI scores had a lower risk of sleep disorders [21]. Similarly, a cross-sectional study in China observed an inverse association between HEI and poor sleep prevalence [22]. An Iranian study on middle-aged workers also found that higher HEI scores correlated with better PSQI scores and longer sleep duration [23]. Given the limited research on Iranian older adults and the potential role of diet–sleep interactions in this group, the current study aimed to assess the association between HEI and sleep quality and duration in older adults using data from the Neyshabur Elderly Longitudinal Study (NeLSA). We hypothesized that individuals with higher HEI scores, reflecting healthier diets, would report better sleep quality and longer duration. Materials and methods Study design and participants This study was approved by the Ethics Committee of Mashhad University of Medical Sciences (Approval code: IR.MUMS.MEDICAL.REC.1401.137). Written informed consent was obtained from all participants. This cross-sectional study was conducted using data from the Neyshabur Elderly Longitudinal Study (NeLSA), an ongoing cohort initiated in 2016 in Neyshabur, northeastern Iran. The NeLSA includes adults aged 50 years and older residing in the community. For the present analysis, 2,026 older adults who met the inclusion criteria were selected. Eligibility criteria included: Age ≥ 60 years.Residency in the study area.Providing informed consent. Age ≥ 60 years. Age ≥ 60 years. Residency in the study area. Residency in the study area. Providing informed consent. Providing informed consent. Participants with acute illness or severe disabilities that prevented questionnaire completion were excluded. Data collection After obtaining written informed consent, trained staff collected demographic data (age, sex, marital status, education level, etc.), medical history, and current medication use through structured interviews and validated questionnaires. Socioeconomic status and lifestyle factors were also recorded. Dietary assessment Dietary intake was assessed using a validated semi-quantitative food frequency questionnaire (FFQ) containing 57 commonly consumed food items in Iran. This questionnaire is an adapted version of the PERSIAN Cohort FFQ for the elderly. The validity and reliability of the FFQ have been confirmed in previous Iranian studies [24, 25]. Participants reported the frequency and portion size of each food item over the past year. Dietary intake data (FFQ) were analyzed using ShaFA, a Microsoft Windows-based software [26], that includes a food composition database of 8,790 foods and 85 nutrients based on USDA-SR28. Subsequently, the HEI-2015 component and total scores were calculated in SPSS following the standard HEI-2015 protocol [27], including energy-density standardization per 1,000 kcal, ratio computation for fatty acids, and scoring of adequacy and moderation components. The HEI-2015 includes 13 components with scores ranging from 0 to 5 or 0 to 10. The total score ranges from 0 to 100 with higher scores indicating greater adherence to dietary guidelines, including higher intakes of fruits, vegetables, whole grains, lean proteins, and lower intakes of saturated fats, added sugars, and sodium. Sleep assessment Sleep quality was evaluated using the Pittsburgh Sleep Quality Index (PSQI), a widely used instrument for measuring sleep quality over the past month [28]. The validated Persian version of the PSQI was used [29]. The PSQI assesses seven domains: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep medications, and daytime dysfunction. Each component is scored from 0 to 3, with a total score ranging from 0 to 21. A total PSQI score > 5 was considered indicative of poor sleep quality. In addition, participants were asked about their average nocturnal sleep duration in quarter-hours units, which was categorized as adequate (≥ 7 h) or insufficient (< 7 h) based on guidelines from the National Sleep Foundation [30]. Other measurements Weight of all participants was measured in minimal clothing and without shoes, in a standing position with the head upright and eyes facing forward, using a mechanical (beam) scale with an accuracy of 0.1 kg. Height was measured in a standing position without shoes, with heels together against the wall, head upright, and eyes facing forward, using a stadiometer with an accuracy of 0.1 cm. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared (kg/m²). Smoking status was recorded as current, former, or never smoker. The presence of chronic diseases including diabetes, hypertension, cardiovascular disease, and respiratory illness was determined through self-report or medical documentation. Statistical analysis Participants were categorized into quartiles based on HEI scores. Sex-specific quartiles were: Women: Q1 ≤ 60.38, Q2 60.39–68.17, Q3 68.18–75.91, Q4 ≥ 75.91; Men: Q1 ≤ 58.28, Q2 58.29–65.14, Q3 65.15–73.34, Q4 ≥ 73.34; All participants: Q1 ≤ 59.54, Q2 59.55–66.95, Q3 66.96–74.97, Q4 ≥ 74.98. we examined the association between HEI quartiles and two sleep outcomes, short sleep duration (< 7 h) and poor sleep quality (PSQI > 5), using logistic regression. Sleep duration and sleep quality were each coded as binary dependent variables (1 = adverse category). HEI was entered as quartiles, with Q1 as the reference. Three models were fitted: Model 1, Crude model, including only HEI quartiles; Model 2, adjusted for age, BMI, smoking status, and education level; Model 3, further adjusted for chronic diseases (diabetes, hypertension, respiratory and heart conditions). Analyses were performed for the total sample and stratified by sex (men and women). Odds ratios (ORs) and 95% confidence intervals (CIs) were reported. All analyses were performed using SPSS, and a P-value < 0.05 was considered statistically significant. Results Participant characteristics A total of 2,026 older adults (1,125 women and 901 men) were included in the analysis. Table 1 shows dietary indices, age, and BMI in the overall population and stratified by sex. The median age of participants was 67 years (IQR: 62–72). The median BMI was 27.6 kg/m², which was significantly higher in women than men (P < 0.001). The median HEI score in the total population was 66.2 (out of 100), with no significant gender difference. Table 1Dietary indices, age, and BMI in the overall population and stratified by sexVariableGenderMeanMedianSDIQRMinMaxp-valueEnergy(Kcal)Male1698.731585.02630.16810.89803.324130.890.590Female1709.791585.28621.1784.54800.874129.08Total1704.871585.15625.01795.95800.874130.89Protein(gr)Male57.2853.9322.8727.1918.51176.410.720Female56.9552.3522.5325.715.74203.99Total57.0952.9422.6726.3415.74203.99Fat(gr)Male56.0450.5823.8527.5620.50187.570.240Female57.0351.1723.8828.7019.85187.81Total56.5950.8223.8628.3419.85187.81Carbohydrate(gr)Male258.31239.82109.07143.1476.83734.580.820Female259.24237.07109.65132.7770.73774.27Total258.83237.54109.37136.7970.73774.27HEIMale66.7866.0711.0414.3727.0095.820.970Female66.8066.2810.7515.5534.6394.54Total66.7966.1810.8715.0627.0095.82BMI(kg/m²)Male26.2426.24.255.415.744.6< 0.001Female28.9728.84.936.416.747.5Total27.7627.64.836.1015.747.5Age(years)Male70.0968.08.1111.060.099.0< 0.001Female68.3866.07.119.060.098.0Total69.14677.61106099The Mann-Whitney U test was used to compare variables between women and men P < 0.05 was considered statistically significant Table 1Dietary indices, age, and BMI in the overall population and stratified by sexVariableGenderMeanMedianSDIQRMinMaxp-valueEnergy(Kcal)Male1698.731585.02630.16810.89803.324130.890.590Female1709.791585.28621.1784.54800.874129.08Total1704.871585.15625.01795.95800.874130.89Protein(gr)Male57.2853.9322.8727.1918.51176.410.720Female56.9552.3522.5325.715.74203.99Total57.0952.9422.6726.3415.74203.99Fat(gr)Male56.0450.5823.8527.5620.50187.570.240Female57.0351.1723.8828.7019.85187.81Total56.5950.8223.8628.3419.85187.81Carbohydrate(gr)Male258.31239.82109.07143.1476.83734.580.820Female259.24237.07109.65132.7770.73774.27Total258.83237.54109.37136.7970.73774.27HEIMale66.7866.0711.0414.3727.0095.820.970Female66.8066.2810.7515.5534.6394.54Total66.7966.1810.8715.0627.0095.82BMI(kg/m²)Male26.2426.24.255.415.744.6< 0.001Female28.9728.84.936.416.747.5Total27.7627.64.836.1015.747.5Age(years)Male70.0968.08.1111.060.099.0< 0.001Female68.3866.07.119.060.098.0Total69.14677.61106099The Mann-Whitney U test was used to compare variables between women and men P < 0.05 was considered statistically significant Dietary indices, age, and BMI in the overall population and stratified by sex The Mann-Whitney U test was used to compare variables between women and men P < 0.05 was considered statistically significant As shown in Table 2, overall, 9.6% of participants reported current smoking, with a notably higher prevalence among men (19.9%) than women (1.4%) (P < 0.001). Regarding chronic diseases, 19.7% had diabetes (which was significantly higher in women), 36.0% had hypertension and 18.0% reported heart disease (which were significantly higher in men), and 5.9% had respiratory illness, with no significant gender difference. Table 2Disease status, education level, and smoking behavior in the overall population and stratified by sexVariableTotal (n = 2026)Male (n = 901)Female (n = 1125)p-valueDiseasesDiabetesYes399 (19.69%)150 (16.64%)249 (22.13%)0.004No1627 (80.31%)751 (83.36%)876 (77.87%)HypertensionYes729 (35.98%)367 (40.37%)362 (32.17%)< 0.001No1297 (64.02%)534 (59.27%)763 (67.83%)Heart DiseaseYes365 (18.01%)181 (20.08%)184 (16.35%)0.035No1661 (81.99%)720 (79.92%)941 (83.65%)Respiratory DiseaseYes120 (5.92%)55 (6.1%)65 (5.77%)0.750No1906 (94.08%)846 (93.9%)1060 (94.23)Educational LevelIlliterate719 (35.5%)208 (23.1%)511 (45.5%)< 0.001Under Diploma1017 (50.4%)504 (55.9%)513 (45.7%)< 0.001Diploma181 (8.9%)113 (12.6%)68 (6.1%)< 0.001University104 (5.2%)75 (8.4%)29 (2.7%)< 0.001Smoking statusCurrent smokingYes195 (9.62%)179 (19.9%)16 (1.4%)< 0.001No1831 (90.38%)722 (80.1%)1109 (98.6%)P-values were calculated using the Chi-square test to assess the association between gender and variables. P < 0.05 was considered statistically significant Table 2Disease status, education level, and smoking behavior in the overall population and stratified by sexVariableTotal (n = 2026)Male (n = 901)Female (n = 1125)p-valueDiseasesDiabetesYes399 (19.69%)150 (16.64%)249 (22.13%)0.004No1627 (80.31%)751 (83.36%)876 (77.87%)HypertensionYes729 (35.98%)367 (40.37%)362 (32.17%)< 0.001No1297 (64.02%)534 (59.27%)763 (67.83%)Heart DiseaseYes365 (18.01%)181 (20.08%)184 (16.35%)0.035No1661 (81.99%)720 (79.92%)941 (83.65%)Respiratory DiseaseYes120 (5.92%)55 (6.1%)65 (5.77%)0.750No1906 (94.08%)846 (93.9%)1060 (94.23)Educational LevelIlliterate719 (35.5%)208 (23.1%)511 (45.5%)< 0.001Under Diploma1017 (50.4%)504 (55.9%)513 (45.7%)< 0.001Diploma181 (8.9%)113 (12.6%)68 (6.1%)< 0.001University104 (5.2%)75 (8.4%)29 (2.7%)< 0.001Smoking statusCurrent smokingYes195 (9.62%)179 (19.9%)16 (1.4%)< 0.001No1831 (90.38%)722 (80.1%)1109 (98.6%)P-values were calculated using the Chi-square test to assess the association between gender and variables. P < 0.05 was considered statistically significant Disease status, education level, and smoking behavior in the overall population and stratified by sex P-values were calculated using the Chi-square test to assess the association between gender and variables. P < 0.05 was considered statistically significant Sleep quality and duration As shown in Table 3, the median self-reported sleep duration was 7 h per night (IQR: 6–8). Overall, 46% of participants reported insufficient sleep (< 7 h), which was significantly more common among women (P < 0.001). The median global PSQI score was 4 (IQR: 2.5–5.5), and 32% of the total population had poor sleep quality (PSQI > 5), again with a higher prevalence among women (P < 0.001). Specifically, women reported a 1-hour shorter median sleep duration (6 h vs. 7 h) and a higher median global PSQI score (5 vs. 4) compared with men. Moreover, scores across all seven PSQI components were significantly higher in women, reflecting greater disturbances across multiple dimensions of sleep. Table 3Components of sleep and global sleep quality score in the overall population and stratified by sexVariableGenderMeanMedianSDIQRMinMaxp-valueComponent1(Subjective Sleep Quality)Male0.951.00.530.00.03.0< 0.001Female1.081.00.580.00.03.0Total1.021.00.570.00.03.0Component2(Sleep Latency)Male0.951.00.961.00.03.0< 0.001Female1.321.01.052.00.03.0Total1.161.01.032.00.03.0Component3)Sleep Duration(Male0.670.00.891.00.03.0< 0.001Female0.811.00.961.00.03.0Total0.750.00.931.00.03.0Component4)Sleep Efficiency(Male0.150.00.520.00.03.0< 0.001Female0.270.00.690.00.03.0Total0.210.00.620.00.03.0Component5(Sleep Disturbances)Male1.01.00.430.00.03.0< 0.001Female1.071.00.470.00.03.0Total1.041.00.450.00.03.0Component6(Use of Sleeping Medication)Male0.360.00.930.00.03.0< 0.001Female0.610.01.140.00.03.0Total0.500.01.050.00.03.0Component7(Daytime Dysfunction)Male0.060.00.30.00.03.0< 0.001Female0.110.00.420.00.03.0Total0.090.00.3740.00.03.0Sleep duration (hr)Male6.77.01.492.02.014.0< 0.001Female6.496.01.451.02.012.0Total6.597.001.471.02.014PSQI scoreMale4.134.02.53.00.017.0< 0.001Female5.285.03.04.00.020.0Total4.774.02.853.00.020.0The Mann-Whitney U test was used to compare variables between women and men. P < 0.05 was considered statistically significant Table 3Components of sleep and global sleep quality score in the overall population and stratified by sexVariableGenderMeanMedianSDIQRMinMaxp-valueComponent1(Subjective Sleep Quality)Male0.951.00.530.00.03.0< 0.001Female1.081.00.580.00.03.0Total1.021.00.570.00.03.0Component2(Sleep Latency)Male0.951.00.961.00.03.0< 0.001Female1.321.01.052.00.03.0Total1.161.01.032.00.03.0Component3)Sleep Duration(Male0.670.00.891.00.03.0< 0.001Female0.811.00.961.00.03.0Total0.750.00.931.00.03.0Component4)Sleep Efficiency(Male0.150.00.520.00.03.0< 0.001Female0.270.00.690.00.03.0Total0.210.00.620.00.03.0Component5(Sleep Disturbances)Male1.01.00.430.00.03.0< 0.001Female1.071.00.470.00.03.0Total1.041.00.450.00.03.0Component6(Use of Sleeping Medication)Male0.360.00.930.00.03.0< 0.001Female0.610.01.140.00.03.0Total0.500.01.050.00.03.0Component7(Daytime Dysfunction)Male0.060.00.30.00.03.0< 0.001Female0.110.00.420.00.03.0Total0.090.00.3740.00.03.0Sleep duration (hr)Male6.77.01.492.02.014.0< 0.001Female6.496.01.451.02.012.0Total6.597.001.471.02.014PSQI scoreMale4.134.02.53.00.017.0< 0.001Female5.285.03.04.00.020.0Total4.774.02.853.00.020.0The Mann-Whitney U test was used to compare variables between women and men. P < 0.05 was considered statistically significant Components of sleep and global sleep quality score in the overall population and stratified by sex Component1 (Subjective Sleep Quality) Component2 (Sleep Latency) Component3 )Sleep Duration( Component4 )Sleep Efficiency( Component5 (Sleep Disturbances) Component6 (Use of Sleeping Medication) Component7 (Daytime Dysfunction) The Mann-Whitney U test was used to compare variables between women and men. P < 0.05 was considered statistically significant Association between HEI and sleep duration Table 4 shows that in both crude and adjusted logistic regression models, no statistically significant association was detected between HEI quartiles and insufficient sleep duration in the total sample or within gender subgroups (P > 0.05). Among women, those in the highest HEI quartile had lower odds of insufficient sleep compared with the lowest quartile, though this association did not reach statistical significance (Model 2: OR = 0.709; 95% CI: 0.502–1.003; P = 0.052). Table 4Multivariable logistic regression assessing the association between sleep duration and HEIGroupModel 1Model 2Model 3QuartilesOR95% CIP-valueOR95% CIP-valueOR95% CIP-valueHEI AllQ1(reference)1——1——1——Q20.9400.725–1.2190.6400.9410.722–1.2250.6490.9490.714–1.2620.718Q30.8100.625–1.0510.1120.8210.631–1.0710.1440.8690.655–1.1530.331Q40.8440.651–1.0940.2000.8480.662–1.1030.2000.8740.657–1.1640.351HEI MenQ1(reference)1——1——1——Q21.0140.686–1.4980.9451.0420.700–1.5510.8411.1400.742–1.7510.357Q30.9150.612–1.3680.6660.9840.652–1.4850.9371.0790.709–1.6460.677Q41.0490.704–1.5630.8141.0640.704–1.6070.7661.0970.691–1.7430.659HEI WomenQ1(reference)1——1——1——Q20.9360.657–1.3330.7150.8840.617–1.2650.4990.8210.556–1.2120.199Q30.7500.532–1.0560.0990.7280.514–1.0310.0740.7430.512–1.0780.119Q40.7270.516–1.0240.0680.7090.502–1.0030.0520.7220.495–1.0540.091Multivariable logistic regression was conducted to assess the association between HEI quartiles and the likelihood of insufficient sleep duration (< 7 h)Model 1 was unadjustedModel 2 was adjusted for age, body mass index (BMI), smoking status, and education levelModel 3 was further adjusted for the variables in Model 2 plus history of cardiovascular or respiratory disease, diabetes, and hypertension. The first quartile of HEI was considered the reference group Table 4Multivariable logistic regression assessing the association between sleep duration and HEIGroupModel 1Model 2Model 3QuartilesOR95% CIP-valueOR95% CIP-valueOR95% CIP-valueHEI AllQ1(reference)1——1——1——Q20.9400.725–1.2190.6400.9410.722–1.2250.6490.9490.714–1.2620.718Q30.8100.625–1.0510.1120.8210.631–1.0710.1440.8690.655–1.1530.331Q40.8440.651–1.0940.2000.8480.662–1.1030.2000.8740.657–1.1640.351HEI MenQ1(reference)1——1——1——Q21.0140.686–1.4980.9451.0420.700–1.5510.8411.1400.742–1.7510.357Q30.9150.612–1.3680.6660.9840.652–1.4850.9371.0790.709–1.6460.677Q41.0490.704–1.5630.8141.0640.704–1.6070.7661.0970.691–1.7430.659HEI WomenQ1(reference)1——1——1——Q20.9360.657–1.3330.7150.8840.617–1.2650.4990.8210.556–1.2120.199Q30.7500.532–1.0560.0990.7280.514–1.0310.0740.7430.512–1.0780.119Q40.7270.516–1.0240.0680.7090.502–1.0030.0520.7220.495–1.0540.091Multivariable logistic regression was conducted to assess the association between HEI quartiles and the likelihood of insufficient sleep duration (< 7 h)Model 1 was unadjustedModel 2 was adjusted for age, body mass index (BMI), smoking status, and education levelModel 3 was further adjusted for the variables in Model 2 plus history of cardiovascular or respiratory disease, diabetes, and hypertension. The first quartile of HEI was considered the reference group Multivariable logistic regression assessing the association between sleep duration and HEI Q1 (reference) Q1 (reference) Q1 (reference) Multivariable logistic regression was conducted to assess the association between HEI quartiles and the likelihood of insufficient sleep duration (< 7 h) Model 1 was unadjusted Model 2 was adjusted for age, body mass index (BMI), smoking status, and education level Model 3 was further adjusted for the variables in Model 2 plus history of cardiovascular or respiratory disease, diabetes, and hypertension. The first quartile of HEI was considered the reference group Association between HEI and sleep quality As shown in Table 5, across all crude and adjusted logistic regression models, no statistically significant association was detected between HEI quartiles and poor sleep quality in the total sample or within sex-specific analyses (P > 0.05). Table 5Multivariable logistic regression assessing the association between sleep quality and HEIGroupModel 1Model 2Model 3QuartilesOR95% CIP-valueOR95% CIP-valueOR95% CIP-valueHEI AllQ1(reference)1——1——1——Q21.0470.803–1.3640.7361.0020.765–1.3120.9890.9220.691–1.2310.582Q31.0530.808–1.3720.7011.0380.793–1.3580.7870.9730.729–1.2990.854Q41.0630.816–1.3840.6521.0350.791–1.3530.8041.0180.765–1.3540.905HEI MenQ1(reference)1——1——1——Q20.8430.542–1.3100.4480.9030.574–1.4200.6580.8770.534–1.4380.602Q30.8970.575–1.3990.6310.9090.574–1.4410.6851.0210.625–1.6670.935Q40.9710.624–1.5120.8981.0290.653–1.6210.9051.1180.689–1.8140.652HEI WomenQ1(reference)1——1——1——Q21.2080.861–1.6950.2741.2210.864–1.7260.2581.3240.910–1.9270.143Q31.1260.806–1.5730.4881.1190.796–1.5730.5191.1690.812–1.6820.402Q40.9620.687–1.3470.8230.9740.691–1.3660.8671.0930.755–1.5830.639Multivariable logistic regression analysis was performed to examine the association between HEI quartiles and the likelihood of poor sleep quality (defined as PSQI > 5)Model 1 was crude (unadjusted)Model 2 was adjusted for age, body mass index (BMI), smoking status, and education levelModel 3 was further adjusted for the covariates in Model 2 along with history of cardiovascular or respiratory diseases, diabetes, and hypertension. The first quartile of HEI was considered the reference group Table 5Multivariable logistic regression assessing the association between sleep quality and HEIGroupModel 1Model 2Model 3QuartilesOR95% CIP-valueOR95% CIP-valueOR95% CIP-valueHEI AllQ1(reference)1——1——1——Q21.0470.803–1.3640.7361.0020.765–1.3120.9890.9220.691–1.2310.582Q31.0530.808–1.3720.7011.0380.793–1.3580.7870.9730.729–1.2990.854Q41.0630.816–1.3840.6521.0350.791–1.3530.8041.0180.765–1.3540.905HEI MenQ1(reference)1——1——1——Q20.8430.542–1.3100.4480.9030.574–1.4200.6580.8770.534–1.4380.602Q30.8970.575–1.3990.6310.9090.574–1.4410.6851.0210.625–1.6670.935Q40.9710.624–1.5120.8981.0290.653–1.6210.9051.1180.689–1.8140.652HEI WomenQ1(reference)1——1——1——Q21.2080.861–1.6950.2741.2210.864–1.7260.2581.3240.910–1.9270.143Q31.1260.806–1.5730.4881.1190.796–1.5730.5191.1690.812–1.6820.402Q40.9620.687–1.3470.8230.9740.691–1.3660.8671.0930.755–1.5830.639Multivariable logistic regression analysis was performed to examine the association between HEI quartiles and the likelihood of poor sleep quality (defined as PSQI > 5)Model 1 was crude (unadjusted)Model 2 was adjusted for age, body mass index (BMI), smoking status, and education levelModel 3 was further adjusted for the covariates in Model 2 along with history of cardiovascular or respiratory diseases, diabetes, and hypertension. The first quartile of HEI was considered the reference group Multivariable logistic regression assessing the association between sleep quality and HEI Q1 (reference) Q1 (reference) Q1 (reference) Multivariable logistic regression analysis was performed to examine the association between HEI quartiles and the likelihood of poor sleep quality (defined as PSQI > 5) Model 1 was crude (unadjusted) Model 2 was adjusted for age, body mass index (BMI), smoking status, and education level Model 3 was further adjusted for the covariates in Model 2 along with history of cardiovascular or respiratory diseases, diabetes, and hypertension. The first quartile of HEI was considered the reference group Discussion In this cross-sectional study of 2,026 Iranian older adults, we examined the relationship between overall diet quality, as measured by the Healthy Eating Index (HEI-2015), and sleep outcomes including sleep quality and duration. Contrary to the initial hypothesis, no statistically significant associations were found between HEI scores and either sleep quality or duration. In other words, older adults with higher HEI scores, reflecting healthier dietary patterns, did not necessarily report better sleep. A borderline inverse association was observed between HEI and insufficient sleep duration in older women. However, this association did not reach statistical significance. This gender-specific pattern may indicate a possible tendency rather than a robust relationship, and the trend was attenuated after adjustment for chronic disease burden. Comparison with previous studies Our findings are consistent with several previous studies that reported no significant associations between diet quality and sleep outcomes. For instance, Yazdani et al. [31], in a cross-sectional study of 739 Iranian adults, found no significant association between HEI and sleep disorders (OR: 1.16; 95% CI: 0.77–1.74). Similarly, Oliveira et al. [32] reported no significant relationship between HEI-2015 and sleep duration, measured objectively via actigraphy, in the Swiss CoLaus study. Conversely, some studies have reported significant associations. Nie et al. [33], analyzing U.S. national data, found that individuals with optimal HEI scores had significantly lower odds of sleep disorders. Huang et al. [34] reported that adults adhering more closely to the Chinese Healthy Eating Index had better sleep quality. Similarly, Behbahani et al. (Iran, 2022), in a cross-sectional study, reported that higher HEI scores were associated with better sleep quality and even longer sleep duration [35]. These differences may be due to variations in dietary assessment methods, sleep evaluation tools, or the characteristics of the study populations. Our study was conducted among older adults in a specific region of Iran, a group whose sleep and dietary behaviors may be influenced by age-related physiological changes, the presence of chronic diseases, medication use, reduced physical activity, and culturally specific lifestyle patterns. In contrast, many previous studies have included broader or younger populations with different cultural, social, and environmental backgrounds. Additionally, cultural practices related to bedtime routines and meal timing, such as late-night eating or variations in meal frequency, have been shown to affect sleep quality [36, 37]. Potential explanations Several factors may explain the null findings in our study. First, the elderly population has unique physiological and psychological attributes. Age-related changes in circadian rhythm, medication use, and multimorbidity may independently affect sleep quality, potentially diminishing the effect of dietary factors. Second, our findings may reflect the limitations of using a summary index like HEI. While HEI provides a broad estimate of dietary quality, it may obscure the effects of specific food groups or nutrients that influence sleep, such as vitamin B complex, vitamin D, magnesium, and omega-3 fatty acids. Even if these nutrients are present in a generally healthy diet, their individual contributions might not be captured adequately by HEI scoring. Third, other lifestyle factors not captured by HEI, such as meal timing, caffeine intake before bedtime, or late-night eating, can affect sleep independent of overall diet quality. Gender-Specific considerations We observed a non-significant inverse association between HEI and insufficient sleep duration in older women. Previous studies assessing overall diet quality have reported gender-specific associations with sleep outcomes. For example, a study assessing diet quality using a Diet Quality Score (DQS) found that higher diet quality was associated with lower odds of short sleep duration among women, whereas no significant association was observed in men [38]. This may point toward gender-specific mechanisms in the diet-sleep relationship, possibly mediated by hormonal, behavioral, or metabolic differences. Importantly, this borderline association observed among women should be interpreted as exploratory and hypothesis-generating rather than confirmatory, particularly given its lack of statistical significance. Further gender-stratified analyses in future studies are warranted. Strengths and limitations This study has several strengths. It focused on the elderly population, a growing yet understudied group in Iran, and used validated tools for data collection, including a semi-quantitative FFQ and the Healthy Eating Index (HEI) for dietary assessment, and the Pittsburgh Sleep Quality Index (PSQI) for sleep evaluation. The relatively large and community-based sample, along with standardized data collection procedures, further supports the robustness of the findings. However, certain limitations should be considered. The cross-sectional design precludes causal inference, and the possibility of reverse causation cannot be excluded, whereby poor sleep quality or insufficient sleep may influence dietary choices rather than the reverse. Reliance on self-reported measures may also introduce recall bias. Additionally, unmeasured confounding factors, such as mental health status or socioeconomic variables, might have influenced the observed associations. Finally, generalizability to other age groups or regions may be limited. Conclusion This study found no significant association between overall diet quality, as measured by the Healthy Eating Index (HEI-2015), and sleep quality or duration among community-dwelling older adults in Iran. Although a trend toward better sleep was observed among women with higher HEI scores, this relationship did not reach statistical significance after controlling for potential confounders. Future directions and recommendations To clarify causal links between diet, and sleep, future studies should employ longitudinal and interventional designs. Integrating subjective tools like PSQI with objective measures such as actigraphy can improve sleep assessment. The use of digital dietary records, mobile apps, and nutritional biomarkers may enhance the accuracy of diet and activity measurement, particularly for indices like HEI. Moreover, exploring the combined effects of psychosocial factors, including stress and mental health, with lifestyle behaviors on sleep is recommended.