Association between the lifelines diet score and the odds of prediabetes and type 2 diabetes mellitus in Iranian adults: a cross-sectional study Authors: Sara Ebrahimi-Mousavi, Yasaman Aali, Nadia Homayounfar, Fatemeh Ehsanitabar, Sara Saffar Soflaei, Gordon A Ferns, Najmeh Seifi, Majid Ghayour-Mobarhan Published: 2026 ABSTRACT AbstractThe Lifelines Diet Score (LLDS) is a useful dietary index that assesses overall adherence to a healthy dietary pattern. To date, no studies have examined its association with type 2 diabetes mellitus (T2DM) in the adult Iranian population. This study aimed to investigate the association between adherence to the LLDS and the odds of prediabetes and T2DM in a large sample of Iranian adults. This cross-sectional study included 4,873 participants aged 35–65 years from the Mashhad Atherosclerotic Heart Disease and Stroke Cohort Study (MASHAD). Dietary intake was assessed using a validated 65-item food frequency questionnaire. The LLDS was calculated based on 12 food groups, and participants were categorized into tertiles based on their LLDS. Anthropometric, biochemical, and lifestyle data were collected. Multivariable logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for diabetes and prediabetes across LLDS tertiles after adjustment for age, sex, marital status, education, smoking, physical activity, and body mass index. Among 4,873 participants, 707 had diabetes and 434 had prediabetes. Higher LLDS was associated with healthier anthropometric and biochemical profiles. In the fully adjusted model (model 3), participants in the highest LLDS tertile had 33% lower odds of T2DM (OR = 0.67, 95% CI: 0.53–0.83, P-trend < 0.001) and 30% lower odds of prediabetes (OR = 0.70, 95% CI: 0.53–0.91, P-trend = 0.009) compared to the lowest tertile. Greater adherence to a high-quality diet, as reflected in the LLDS, was inversely associated with the odds of T2DM and prediabetes in Iranian adults. These findings suggest that dietary improvements may be associated with lower odds of glycemic disorders in this population.Keywords: Prediabetes, Type 2 diabetes mellitus, Lifelines Diet Score, Adults, Cross-sectionalSubject terms: Diseases, Endocrinology, Health care, Medical research, Risk factors Introduction Prediabetes is characterized by blood glucose levels above normal but below the diagnostic threshold for diabetes. It represents a critical intermediate metabolic state1. This condition significantly increases the risk of progression to overt type 2 diabetes mellitus (T2DM) and the development of cardiovascular complications1,2. According to International Diabetes Federation (IDF) estimates, the global prevalence of impaired glucose tolerance (IGT)—a key diagnostic criterion for prediabetes—was 9.1% (affecting approximately 464 million adults) in 2021. This figure is projected to rise to 10.0%, representing an estimated 638 million individuals, by 20451–3. T2DM continues to impose a significant global public health burden, with recent estimates indicating that over 800 million adults worldwide were living with diabetes in 2022, reflecting an increase of 630 million since 19904,5. The global age-standardized prevalence reached 6.1% in 2021 and is projected to exceed 10% in nearly half of all countries by 2050, with the most rapid increases occurring in low- and middle-income countries4,6. This increase is primarily driven by rising obesity rates, sedentary lifestyles, urbanization, and population aging. By 2050, the total number of affected individuals could surpass 1.3 billion worldwide6,7. Consistent with the global rise in diabetes prevalence, the prevalence in Iran increased from 7.08% in 1988 to 13.4% in 20238. Dietary patterns are among the most modifiable risk factors for T2DM and prediabetes9. Adherence to high-quality diets rich in plant-based foods, whole grains, and healthy fats, as captured by indices such as the Mediterranean Diet Score and the Dietary Approaches to Stop Hypertension (DASH) score, is consistently associated with improved glycemic control and a lower risk of T2DM across various populations10–14. While these established scores share common principles—such as emphasizing fruit and vegetable intake and distinguishing between healthy and unhealthy fats—they also have specific limitations. For instance, the Mediterranean diet explicitly discourages the consumption of artificially sweetened beverages and limits dairy, which may not fully align with all modern dietary contexts or global food practices7. These limitations have motivated the development of more contemporary dietary assessment tools such as the Lifelines Diet Score (LLDS). The LLDS classifies food groups based on current evidence regarding their health effects and provides a comprehensive measure of diet quality15. Emerging cross-sectional and cohort studies from 2022 onward have associated higher LLDS with lower odds of metabolically unhealthy obesity, hypertension, and polycystic ovary syndrome7,16,17, though evidence specifically linking LLDS to T2DM and prediabetes remains limited in diverse populations, especially in the Iranian adult population. In a Kurdish population of postmenopausal women, Fateh et al. demonstrated that higher adherence to the LLDS was associated with a 39% lower likelihood of diabetes and a 38% lower likelihood of hypertension7. Considering the increasing burden of glucose metabolism disorders and diabetes in Iran, along with the significant potential of dietary modifications for risk reduction, evaluating contemporary dietary assessment tools in this population is imperative. This need is particularly relevant given the distinct dietary patterns and high cardiometabolic risk profile prevalent in the Middle Eastern region18. To our knowledge, no prior study has investigated the association between the LLDS and diabetes among Iranian adults. Therefore, the present study was conducted to examine the relationship between LLDS and the odds of T2DM and prediabetes in a large adult cohort in Iran. Methods Study design and population Participants in the present study were part of the Mashhad Atherosclerotic Heart Disease and Stroke Cohort Study (MASHAD). This prospective cohort was launched in 2010 to investigate cardiovascular disease risk factors and initially enrolled 9,704 men and women, aged 35–65 years, who were randomly selected via cluster sampling from the urban population of Mashhad, Eastern Iran. Participation was contingent upon providing informed consent19. At the baseline examination, trained health personnel collected comprehensive data on demographic characteristics, anthropometric measures, biochemical profiles, and lifestyle factors from all participants. For the current cross-sectional analysis, individuals with missing diabetes data (N = 140), missing dietary data, or implausible daily energy intake (< 800 or > 4200 kcal/day)20 were excluded (N = 4,648). Pregnant or lactating women were also excluded (N = 43). After applying these exclusion criteria, a total of 4,873 participants were included in the final analytical sample. A detailed flowchart of the participant selection process, including specific exclusion criteria and the number of individuals excluded at each stage, is presented in Fig. 1. Flow diagram of study participants. The MASHAD study protocol was approved by the Ethics Committee for Human Research of the Mashhad University of Medical Sciences (Ethics Number: IR.MUMS.MEDICAL.REC.1395.458) and all methods were carried out in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants before their inclusion in the study. Anthropometric measurements and physical activity level (PAL) Body weight was measured to the nearest 0.1 kg using calibrated digital scales21, with participants in light clothing and without shoes. Height was recorded to the nearest 0.1 cm using a wall-mounted stadiometer, with the head positioned in the Frankfort plane. Waist circumference was measured with a non-stretch tape to the nearest millimeter19. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared (kg/m²). PAL was evaluated using an adapted version of the Scottish Heart Health Study (SHHS)/MONICA questionnaire22. This instrument captures the duration of physical activity during occupational, leisure-time, and sedentary periods. PAL was derived as the ratio of total energy expenditure7 to basal metabolic rate (BMR). Total energy expenditure7 was estimated by assigning sex-specific physical activity intensity values to reported activity levels (low, moderate, high). BMR was calculated using the FAO/WHO/UNU predictive equations (23). Blood pressure measurement Blood pressure was measured on the left arm using a mercury sphygmomanometer (accuracy ± 1 mmHg). After participants had rested in a seated position for at least 15 min, two consecutive readings were taken. If the two readings differed by more than 25 mmHg for systolic blood pressure (SBP) or 15 mmHg for diastolic blood pressure (DBP), a third measurement was performed, and the average of the two closest readings was recorded. The mean of the final readings was used for analysis19,24. Laboratory evaluation Blood samples were collected from all participants after a minimum 14-hour fasting. Following collection, samples were centrifuged at ambient temperature to isolate serum and plasma components. Serum aliquots of 500 µL were archived at -80 °C in the study biobank for subsequent investigation. Biochemical assays were performed on serum. Concentrations of fasting plasma glucose (FPG), total cholesterol, triglycerides (TG), and high-density lipoprotein cholesterol (HDL-C) were determined through enzymatic methods employing commercial assay kits (Pars Azmoon, Tehran, Iran) and an automated analyzer (Autoanalyzer BT-3000, Biotecnica, Italy). For individuals with TG levels under 400 mg/dL, low-density lipoprotein cholesterol (LDL-C) was estimated via the Friedewald equation. Full methodological details for these laboratory procedures are available in prior publications19. T2DM and prediabetes definition Prediabetes was defined according to WHO criteria as impaired fasting glucose (IFG). IFG was characterized by a fasting plasma glucose level of 110–125 mg/dL1,25. The diagnosis of T2DM was based on an FPG level of 126 mg/dL or higher, or a documented history of treatment with insulin or oral hypoglycemic medications19. Dietary intake assessment Food and nutrient intake were measured using a semiquantitative food frequency questionnaire (FFQ) of 65 food items that had been previously validated in the Iranian population26. This questionnaire was completed at baseline by a trained and experienced dietitian based on participants’ reported consumption patterns during the past year. For each item, data were collected on serving size, utilizing household measures, and consumption frequency. Frequency options included daily, weekly, monthly, rarely, or never. The completed questionnaires were analyzed using Diet Plan 6 dietary analysis software (Forestfield Software Ltd., Horsham, West Sussex, UK). This software facilitated the quantification of Total daily energy intake (in kilocalories) and Daily intake of macronutrients (protein, fat, and carbohydrates). This process yielded detailed, quantified data on the dietary patterns, total energy consumption, and nutrient intake for all study participants in the first phase of the MASHAD study19. Lifelines Diet Score calculation The LLDS was calculated following the established methodology by Vinke and colleagues15. This scoring system categorizes food groups according to their hypothesized health impact into beneficial, adverse, neutral, or indeterminate groups. Food groups classified as neutral, such as eggs, or of indeterminate effect—including potatoes, refined grains, unprocessed poultry, cheese, savory snacks, confectionery, soups, sweetened dairy, and artificially-sweetened beverages—were omitted from the final score. The LLDS is derived from 12 food groups: nine beneficial groups (vegetables, fruits, whole grains, legumes and nuts, fish, vegetable oils and soft margarines, unsweetened dairy, coffee, and tea) and three adverse groups (red and processed meat, butter and hard margarines, and sugar-sweetened beverages). To account for differences in total energy intake, consumption for each food group was standardized as grams per 1000 kilocalories (g/1000 kcal). For each participant, their standardized intake for each of the 12 groups was ranked into sex-specific population quintiles and assigned a corresponding score from 1 (lowest intake) to 5 (highest intake). For beneficial food groups, a higher quintile (i.e., higher intake) received a higher score, while for adverse food groups, a higher quintile (i.e., higher intake) received a lower score. The scores for all 12 components were summed to yield a total LLDS ranging from 12 to 60. For analytical purposes, the total LLDS was subsequently divided into tertiles (T1, T2, T3) based on the following distribution of scores in our study population: T1: 14–27; T2: 28–32; T3: 33–45. The actual LLDS scores in our sample ranged from 14 to 45, indicating that extreme scores were uncommon in this population. Statistical analysis Data for continuous variables are summarized as means with standard deviations (SD), and categorical data are reported as counts and percentages. Normality of distribution for quantitative measures was evaluated with the Kolmogorov-Smirnov test. Group differences across LLDS tertiles were assessed using appropriate statistical tests: one-way analysis of variance (ANOVA) or the Kruskal-Wallis test for continuous measures, and the chi-square test for categorical variables. To assess the relationship between LLDS and the odds of T2DM and prediabetes, we utilized multivariable logistic regression analyses. The LLDS was analyzed in tertiles, using the lowest tertile (T1) as the reference. Odds ratios (ORs) with 95% confidence intervals (CIs) were estimated using a series of progressively adjusted models: an unadjusted model, a model controlling for age and sex (model 1), a model that further included marital status, smoking status, educational level, and physical activity level (model 2), and a final model further adjusted for BMI (model 3). A test for linear trend was conducted by treating the LLDS tertiles as a continuous ordinal variable (scores 1–3) in the regression models. This method assumes a linear relationship on the log-odds scale and equal increments in odds between adjacent tertiles, which is a common and acceptable assumption in trend analyses. All statistical computations were executed using IBM SPSS Statistics version 25.0. Statistical significance was defined as a two-tailed P-value below 0.05. Results Participant characteristics across LLDS tertiles In total, 4,873 participants with a mean age of 48.68 years and a mean BMI of 27.93 kg/m2 were analyzed; 9.0% had prediabetes and 14.5% had T2DM. Participants were divided into tertiles based on their LLDS: T1 (14–27, n = 1,514), T2 (28–32, n = 1,860), and T3 (33–45, n = 1,499). Table 1 summarizes the baseline characteristics of the study population according to LLDS tertiles. There was a significant mean difference in age, BMI, weight, WC, FPG, SBP, cholesterol, and TG across tertiles of LLDS. Other results are reported in Table 1. Table 1Basic characteristics of study participants across categories of LLDS.Tertiles of LLDST1(n = 1514)T2(n = 1860)T3(n = 1499)P-value(14–27)(28–32)(33–45)(34–47)Age (year)47.51 ± 8.0848.68 ± 8.2149.89 ± 8.10 < 0.001 BMI (kg/m2)28.57 ± 4.5727.94 ± 4.7627.28 ± 4.54 < 0.001 Weight (kg)73.02 ± 12.3972.17 ± 13.0770.35 ± 12.87 < 0.001 WC (cm)95.58 ± 12.0495.46 ± 12.3594.01 ± 12.15 < 0.001 Physical activity1.59 ± 0.281.56 ± 0.271.59 ± 0.26 0.001 FBG (mg/dL)94.52 ± 40.3992.63 ± 38.5089.81 ± 30.17 0.001 SBP (mm Hg)121.41 ± 17.67123.10 ± 18.97123.01 ± 19.03 0.017 DBP (mm Hg)79.27 ± 11.3779.97 ± 11.2279.64 ± 11.280.194Cholesterol (mg/dL)190.07 ± 36.22192.40 ± 38.47193.57 ± 40.23 0.038 HDL (mg/dL)41.93 ± 9.5741.72 ± 9.3442.12 ± 9.050.460LDL (mg/dL)119.56 ± 32.44119.26 ± 34.50118.53 ± 35.670.696TG (mg/dL)151.55 ± 97.54145.91 ± 90.58133.71 ± 78.60 < 0.001 Female%, (N)57.5 (817)57.8 (1076)62.8 (941) 0.004 Married %, (N)92.9 (1406)92.9 (1728)92.8 (1391)0.542Current smokers%, (N)22.3 (337)24.9 (463)16.1 (242) < 0.001 Education (university graduated) %, (N)5.6 (90)6(114)9 (141) < 0.001 Data are presented as mean (SD) or percent (N).BMI, body mass index; WC, waist circumference; FBG, fasting blood glucose; TG, triglyceride; HDL, high-density lipoprotein; SBP, systolic blood pressure; DBP, diastolic blood pressure. Basic characteristics of study participants across categories of LLDS. Data are presented as mean (SD) or percent (N). BMI, body mass index; WC, waist circumference; FBG, fasting blood glucose; TG, triglyceride; HDL, high-density lipoprotein; SBP, systolic blood pressure; DBP, diastolic blood pressure. Dietary intakes of study participants across LLDS tertiles Dietary intakes varied markedly across LLDS tertiles, as detailed in Table 2. There was a significant mean difference in energy intake, protein, carbohydrate, fat, fiber, and PUFA across the LLDS tertiles (all P < 0.001). Other results are reported in Table 2. Table 2Dietary intakes of study participants across categories of LLDS*.Tertiles of LLDST1T2T3P-value†LLDS ranges(14–27)(28–32)(33–45)Number of participants151418601499 Energy (kcal/d)2022.22 ± 673.691882.94 ± 663.691739.55 ± 608.14 < 0.001 Protein (g/d)76.65 ± 35.1571.24 ± 32.7068.48 ± 30.42 < 0.001 Carbohydrate (g/d)265.01 ± 102.42248.96 ± 95.95228.62 ± 90.08 < 0.001 Fat (g/d)79.68 ± 35.5673.45 ± 35.9267.28 ± 32.89 < 0.001 Fiber (g/d)16.88 ± 9.6717.86 ± 10.5218.70 ± 11.85 < 0.001 PUFA (g/d)23.70 ± 15.4425.51 ± 16.3527.91 ± 16.12 < 0.001 TFA (g/d)1.06 ± 0.780.99 ± 0.790.85 ± 0.65 < 0.001 SFA (g/d)20.55 ± 10.2019.38 ± 9.8717.90 ± 9.02 < 0.001 MUFA (g/d)17.91 ± 9.3819.84 ± 10.3921.58 ± 10.53 < 0.001 LLDS components Fruits (g/d)139.78 ± 167.63233.20 ± 203.73313.78 ± 223.21 < 0.001 Vegetables (g/d)116.93 ± 118.07170.50 ± 150.34240.72 ± 164.33 < 0.001 Whole grains (g/d)99.61 ± 113.76152.95 ± 130.60177.67 ± 124.20 < 0.001 Legumes and nuts (g/d)53.40 ± 23.4155.82 ± 26.4860.29 ± 29.61 < 0.001 Fish (g/d)4.41 ± 6.246.48 ± 9.9711.03 ± 13.25 < 0.001 Coffee and tea (g/d)13.82 ± 86.1722.16 ± 137.0717.81 ± 66.390.068 Unsweetened dairy (g/d)187.17 ± 125.22256.50 ± 154.48341.67 ± 187.86 < 0.001 Red and processed meats (g/d)52.83 ± 37.9448.81 ± 37.7944.25 ± 38.50 < 0.001 Butter, hard margarines (g/d)7.20 ± 8.575.84 ± 8.603.34 ± 8.44 < 0.001 SSB ((ml/d)109.10 ± 125.25103.10 ± 146.7482.46 ± 115.77 < 0.001 All values are presented as mean (SD).*P-values were obtained using one-way ANOVA for normally distributed variables and the Kruskal–Wallis test for non-normally distributed variables (fruits, vegetables, fish, Coffee and tea, and butter hard margarines).LLDS, Lifeline Diet Score; PUFA, Polyunsaturated fatty acid; TFA, Trans fatty acid; SFA, Saturated fatty acid; MUFA, Monounsaturated fatty acid; SSB, Sugar-sweetened beverages. Dietary intakes of study participants across categories of LLDS*. All values are presented as mean (SD). *P-values were obtained using one-way ANOVA for normally distributed variables and the Kruskal–Wallis test for non-normally distributed variables (fruits, vegetables, fish, Coffee and tea, and butter hard margarines). LLDS, Lifeline Diet Score; PUFA, Polyunsaturated fatty acid; TFA, Trans fatty acid; SFA, Saturated fatty acid; MUFA, Monounsaturated fatty acid; SSB, Sugar-sweetened beverages. Association of LLDS with the odds of T2DM and prediabetes Table 3 presented the ORs for T2DM and prediabetes across LLDS tertiles, regarding the first tertile as reference. Among the 4873 participants, 707 had diabetes (T1 = 262; T2 = 274; T3 = 171). In the crude model, participants in T3 had 40% lower odds of T2DM compared to T1 (OR = 0.60; 95% CI: 0.48–0.74). A significant inverse trend was observed, indicating that higher adherence to the LLDS was associated with progressively lower odds of T2DM (P-trend < 0.001). After adjustment for age and sex in model 1, these associations persisted (OR = 0.63, 95% CI: 0.51–0.78, P-trend < 0.001). Further adjustments for marital status, smoking, education, and physical activity (model 2) yielded similar results (OR = 0.64, 95% CI: 0.52–0.80, P-trend < 0.001), as did additional control for BMI (model 3) (OR = 0.67, 95% CI: 0.53–0.83, P-trend < 0.001). Table 3Multivariable-adjusted odds ratios (95% CIs) for diabetes and prediabetes across LLDS tertiles.Tertiles of LLDSP-trend*T1T2T3LLDS Ranges(14–27)(28–32)(33–45)Diabetes Number of participants151418601499 Number of participants with diabetes262274171 Crude1.000.73 (0.60–0.90)0.60 (0.48–0.74) < 0.001 Model 1†1.000.75 (0.60–0.92)0.63 (0.51–0.78) < 0.001 Model 2‡1.000.77 (0.62–0.95)0.64 (0.52–0.80) < 0.001 Model 3§0.78 (0.63–0.96)0.67 (0.53–0.83) < 0.001 Prediabetes Number of participants125215861328 Number of participants with prediabetes153163118 Crude1.000.75 (0.58–0.97)0.58 (0.44–0.74) < 0.001 Model 1†1.000.79 (0.61–1.02)0.64 (0.49–0.83) 0.001 Model 2‡1.000.82 (0.63–1.07)0.66 (0.51–0.87) 0.003 Model 3§1.000.84 (0.64–1.09)0.70 (0.53–0.91) 0.009 Data are presented as odds ratios (95% CI).†Model 1: Adjusted for age and sex.‡Model 2: Further adjusted for marriage status, smoking, education degree, and physical activity.§Model 3: Further adjusted for BMI. Multivariable-adjusted odds ratios (95% CIs) for diabetes and prediabetes across LLDS tertiles. Data are presented as odds ratios (95% CI). †Model 1: Adjusted for age and sex. ‡Model 2: Further adjusted for marriage status, smoking, education degree, and physical activity. §Model 3: Further adjusted for BMI. For prediabetes (n = 434), higher adherence to the LLDS was associated with significantly lower odds of prediabetes. In the crude model, participants in T3 of LLDS had 42% lower odds compared to those in T1 (OR = 0.58, 95% CI: 0.44–0.74, P-trend < 0.001). This inverse association remained significant in model 1 (OR = 0.64, 95% CI: 0.49–0.83, P-trend = 0.001) and model 2 (OR = 0.66, 95% CI: 0.51–0.87, P-trend = 0.003). Also, after further adjustment for BMI in model 3, the associations remained (OR = 0.70, 95% CI: 0.53–0.91, P-trend = 0.009). Overall, higher LLDS was inversely linked to both T2DM and prediabetes, with trends holding after multivariable adjustments. Discussion This cross-sectional study examined the association between greater adherence to the LLDS and the odds of developing T2DM and prediabetes. To our knowledge, this is the first study that investigates the association between LLDS and odds of T2DM and prediabetes in Iran. After multivariable adjustment for age, sex, marital status, smoking status, education level, physical activity, and BMI, participants in the highest LLDS tertile had significantly lower odds of diabetes (OR = 0.67, 95% CI: 0.53–0.83) and prediabetes (OR = 0.70, 95% CI: 0.53–0.91) compared with those in the lowest tertile. We believe the main novel contribution of this study is that it is the first to evaluate the LLDS with both prediabetes and T2DM in Iranian adults. This helps fill a gap in the literature on diet quality scores in Middle Eastern settings. Prior studies have demonstrated a strong inverse association between adherence to the LLDS and the risk of chronic diseases in the general population. In a case-control study including 300 pregnant women, Abudari et al.27 reported that participants in the highest LLDS quartile had 76% lower odds of gestational diabetes mellitus (GDM) compared with those in the lowest quartile. Likewise, Fateh et al.7 demonstrated that there is a strong correlation between LLDS and both T2DM and hypertension (HTN) in postmenopausal women. In addition to metabolic effects, Khani et al.28 showed that greater adherence to the LLDS was associated with better sleep quality, with individuals in the highest LLDS tertile exhibiting significantly lower odds of poor sleep quality than those in the lowest tertile. Our findings are consistent with these previous LLDS-based studies conducted in other populations and extend the evidence to an Iranian population for the first time. Furthermore, the available evidence suggests that the protective effects of the LLDS extend beyond metabolic health. Case-control studies have reported a reduced risk of breast cancer in individuals with higher adherence to the LLDS29. This association has also been observed for nonalcoholic fatty liver disease30. Additionally, a prospective cohort study has shown that greater adherence to the LLDS was associated with a reduced risk of lung cancer incidence and death31. Our findings are consistent with previous studies using other dietary quality indices, such as the Alternative Healthy Eating Index (AHEI), the DASH diet, the dietary inflammation index (DII), the Healthy Eating Index (HEI), and the Mediterranean diet, which have similarly reported inverse associations with the risk of T2DM. Umbrella reviews suggest that greater adherence to these dietary indices is associated with a reduced T2DM32. The present study demonstrates an inverse correlation between LLDS and diabetes. This association may be explained by the divergent dietary patterns observed across the LLDS tertiles. Individuals who exhibit elevated LLDS scores tend to demonstrate a dietary pattern characterized by increased consumption of fruits, vegetables, whole grains, legumes, and nuts. The ingestion of these foodstuffs may assist in the regulation of glucose levels, a phenomenon that can be attributed to their high fiber content, antioxidant properties, and the presence of other beneficial plant compounds. It has been demonstrated that there is a correlation between these factors and a reduction in systemic inflammation, as well as an enhancement in the gut microbiota balance27. Fruits and vegetables have been shown to play an important role in reducing the risk of T2DM due to their bioactive compounds, polyphenols, antioxidants, and fiber. The substances in question enhance insulin response by reducing oxidative stress and improving glucose absorption function. The presence of fiber in these food groups also works through different mechanisms, such as slowing the rate of carbohydrate absorption and thus controlling postprandial glucose and insulin33. In addition, higher dietary fiber intake is also associated with increased satiety34. This mechanism could reduce the risk of overweight and obesity, a known risk factor for T2DM. In addition to plant-based sources, low-fat dairy products have been demonstrated to have a beneficial effect on diabetes status. Several biological pathways may explain this association, which collectively facilitate the beneficial effects on insulin resistance. Calcium, vitamin D, casein, and whey protein found in these products may regulate body fat accumulation and waist circumference while improving insulin sensitivity. Mechanistically, low calcium intake reduces extracellular calcium and increases intracellular calcium, stimulating fat storage in adipocytes and increasing insulin resistance risk35,36. Conversely, evidence suggests that high intake of red and processed meat is associated with adverse metabolic profiles. Research has indicated that increased consumption of these foods is associated with elevated levels of C-reactive protein (CRP), ferritin, fasting insulin, and hemoglobin A1c (HbA1c (and diminished levels of adiponectin. Despite the attenuation of certain associations following adjustment for BMI, obesity emerged as a substantial contributing factor to these outcomes, except for ferritin. It is noteworthy that the substitution of red meat for healthier protein sources, such as poultry, fish, legumes, and nuts, has been linked to enhanced inflammatory and metabolic profiles. Specifically, there is a consistent body of evidence indicating a positive correlation between the consumption of processed meats and an elevated risk of T2DM. This association is hypothesized to be attributable to the presence of iron, which has been demonstrated to induce oxidative stress and elevate levels of ferritin. These phenomena have been shown to have a deleterious effect on insulin sensitivity37,38. Taken together, in addition to dairy and fruits, whole grains have also been consistently associated with a reduced risk of T2DM. Conversely, higher consumption of red and processed meats and sugar-sweetened beverages has been linked to an increased risk of T2DM39. Supporting this pattern, results from a meta-analysis indicated that higher nut consumption was associated with a significantly lower risk of diabetes, whereas legume consumption did not show a significant association with diabetes risk40,41. Finally, beverage consumption patterns may also contribute to diabetes risk. In a prospective cohort study, higher tea and coffee consumption was associated with a significantly reduced risk of T2DM. This association was independent of caffeine, magnesium, potassium, and blood pressure. It is possible that other bioactive compounds may mediate this effect42. In contrast, higher margarine consumption was associated with a significantly increased risk of cardiovascular disease and T2DM; consumption of more than 7 g per day was associated with a 29% increased risk of cardiovascular disease and a 41% increased risk of T2DM43. In the present study, the overall LLDS score showed a significant inverse association with prediabetes and T2DM, suggesting that the combined effect of multiple dietary components, rather than individual food groups, may contribute to the observed association. A notable strength of this study is its large sample size. The use of a validated and culturally appropriate FFQ enhances the reliability of our findings. Furthermore, potential confounding factors were controlled for in the analyses. However, this study has several limitations that warrant consideration. One of the main limitations is related to the use of the FFQ, which is subject to self-reporting biases, including under- or over-reporting of food intake. It should also be noted that data on 2-hour postprandial glucose and HbA1c were not available in the MASHAD cohort, so the diagnosis of prediabetes and T2DM relied solely on fasting blood glucose. Furthermore, our study population was derived from a cohort study in Mashhad (northeastern Iran). Given the differences in food consumption patterns, lifestyle behaviors, and genetic predispositions across regions of Iran and across other countries, our findings may not be generalizable to other populations. Future studies in diverse populations are needed to confirm the external validity of these results. Additionally, the cross-sectional design of this study prevents any causal inference regarding the association between LLDS and prediabetes or T2DM. Despite adjusting for several potential confounders, the possibility of residual or unmeasured confounding such as family history of diabetes cannot be ruled out. Also, categorizing LLDS into tertiles may reduce statistical sensitivity and obscure potential non-linear or dose-response relationships. Conclusion Our findings suggest that greater adherence to the LLDS is associated with a reduced risk of developing diabetes and prediabetes. Further prospective and interventional studies are needed to confirm these associations and elucidate the underlying mechanisms. Abbreviations Alternative Healthy Eating Index One-way analysis of variance Body mass index Basal metabolic rate Confidence intervals C-reactive protein Dietary Approaches to Stop Hypertension Diastolic blood pressure dietary inflammation index Food frequency questionnaire Fasting plasma glucose Gestational diabetes mellitus Hemoglobin A1c high-density lipoprotein cholesterol hypertension International Diabetes Federation Impaired fasting glucose impaired glucose tolerance Low-density lipoprotein cholesterol Lifelines Diet Score Mashhad Atherosclerotic Heart Disease and Stroke Cohort Study Odds ratios Physical activity level Standard deviation Systolic blood pressure type 2 diabetes Total energy expenditure Triglyceride Author contributions SEM, YA, NH, and FE wrote the paper; SEM performed the statistical analyses; SEM, NH, NS, and MGM interpreted the results. YA, NS, SSS, MGM, and GAF revised the article, and MGM had full access to all the data in the study and took responsibility for the integrity and accuracy of the data. The final manuscript was approved by all authors. Funding No funding. Data availability Authors declare that the data of this study are provided in this article, and all the data in the study will be available with the opinion of the corresponding author ( [GhayourM@mums.ac.ir](mailto: GhayourM@mums.ac.ir) ). Declarations Competing interests The authors declare no competing interests. Ethics approval and consent to participate Ethics approval for the study protocol was confirmed by The Human Ethics Committee of Mashhad University of Medical Sciences (Ethics Number: IR.MUMS.MEDICAL.REC.1395.458). All participants signed a written informed consent approved by the Ethics committee. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: Najmeh Seifi and Majid Ghayour-Mobarhan. Contributor Information Najmeh Seifi, Email: SeifiN@mums.ac.ir. Majid Ghayour-Mobarhan, Email: GhayourM@mums.ac.ir. References Associated Data Data Availability Statement Authors declare that the data of this study are provided in this article, and all the data in the study will be available with the opinion of the corresponding author ( [GhayourM@mums.ac.ir](mailto: GhayourM@mums.ac.ir) ).