Adherence to Portfolio diet and risk of type 2 diabetes mellitus in a population-based prospective cohort study

Abstract

Background

Type 2 diabetes mellitus (T2DM) poses significant global health challenges, with its rising prevalence linked closely to lifestyle and dietary factors. The Portfolio Diet, a plant-based dietary regimen known to lower cholesterol and cardiovascular disease, may also offer protective effects against T2DM, though limited research has explored this potential association. This study investigates the relationship between adherence to the Portfolio Diet and T2DM risk within the Tehran Lipid and Glucose Study (TLGS) during 8.9 years of follow-up, a population-based prospective study.

Methods

This analysis included 2,188 T2DM-free adults, aged ≥ 21. Dietary intake was assessed using a validated and reliable FFQ, and adherence to the Portfolio Diet was evaluated through a Portfolio Diet Score (PDS) derived from the dietary intake data. Cox proportional hazard models, adjusted for demographic and lifestyle factors, were employed to assess the association between PDS and T2DM risk.

Results

Overall PDS were not statistically significantly associated with T2DM risk across the tertiles in fully adjusted models (HR = 0.91; 95%CI: 0.64–1.31, P trend = 0.610). Analysis of individual PDS components indicated an inverse association between nuts intake and T2DM risk in the highest tertile (HR = 0.72, 95% CI: 0.51–0.99; P trend = 0.050). Other components, including plant protein, viscose fiber, monounsaturated fatty acids (MUFA), and total phytosterols, did not show significant associations with T2DM risk in adjusted models.

Conclusions

While overall adherence to the portfolio diet did not significantly reduce T2DM risk in this Middle Eastern cohort, nut consumption was inversely associated with T2DM incidence. These findings indicate that both region-specific dietary patterns and the weighting of components within the Portfolio Diet Score may affect its ability to capture associations with type 2 diabetes risk.

Clinical trial number

Not applicable.

Introduction

Type 2 diabetes mellitus (T2DM) is a persistent medical condition characterized by high blood glucose levels and reduced insulin sensitivity, leading to long-term organ damage. It represents a significant public health concern due to its irreversible complications, high disability rates, and mortality [1]. In 2021, 537 million adults worldwide were affected by T2DM, with projections indicating an increase to 700 million by 2045 [2]. The Middle East and North Africa (MENA) region holds the highest global prevalence at 16.2% [3]. The prevalence of T2DM in Iran has been reported to range from 8.1 to 14.2%, depending on various studies [4–6]. Different risk factors have been considered concerning T2DM, and diet plays a crucial role [7].

The Portfolio Diet is a plant-based dietary pattern developed in the early 2000s with the primary goal of lowering cholesterol and reducing cardiovascular disease [8]. It consists of a combination, or “portfolio,” which lowers cholesterol, such as plant proteins, viscous fiber, nuts, and phytosterols. Another essential component, monounsaturated fats (MUFAs), was later added to the Portfolio Diet to enhance its cholesterol-lowering effects. Previous studies have shown that the consumption of components of the Portfolio Diet including legumes, nuts, and viscose fibers was associated with a reduced risk of T2DM and improved diabetes management [9–14]. In addition, intake of MUFA is effective in improving glycemic control and insulin sensitivity [15]. However, this relationship has been rarely reported in the form of a coherent score.

While the Portfolio Diet has been extensively studied and shown to effectively reduce risk factors associated with cardiovascular disease (CVD), such as lowering low-density lipoprotein cholesterol (LDL-C) levels, improving blood pressure, promoting favorable lipid profiles, and improving markers of inflammation and vascular health [16–18], research on its connection to T2DM remains scarce [19–21]. A patient-level pooled analysis of two randomized dietary trials has shown the beneficial effects of the portfolio diet in adults living with T2DM. The Portfolio Diet was associated with lower hemoglobin A1C (HbA1c) over 6 months, predominantly driven by two pillars: plant protein, nuts, and seeds, particularly dietary pulses [20]. Moreover, in the PREDIMED-Plus trial, lower levels of HbA1c and fasting blood sugar (FBS) were observed in elderly people with increasing adherence to the portfolio diet [19]. Investigating the Women’s Health Initiative (WHI) study indicated that a portfolio diet might decrease the occurrence of T2DM by approximately 23% [21].

Previous studies have been conducted in American and European countries, and no study has been conducted in the MENA region. In addition, the dietary habits, lifestyle, and genetics of people living in the MENA region differ from other regions (e.g. higher refined grain intake may modulate outcome) [22]. Therefore, this study aimed to examine the link between adherence to the Portfolio Diet and the development of T2DM in the Tehran Lipid and Glucose Study (TLGS), a population underrepresented in prior research, addressing gaps in regional dietary guidelines.

Materials and methods

Study design and population

This prospective study was conducted as part of the TLGS, a longitudinal population-based cohort investigation in District 13 of Tehran, involving 15,005 men and women aged 3 years or older. The TLGS methodology has been described extensively in previous studies [23]. The study began in 1999, with measurements repeated every three years. For this analysis, Phase 3 (2005–2008) (n = 12519, age > 3 years) was used as the baseline, and participants were followed up through Phase 6 (2016–2018).

As shown exclusion criteria in Fig. 1, participants aged ≥ 21 were considered for this project (n = 9645). Individuals with T2DM at baseline (n = 1341) and missing data of age, sex, body mass index (BMI), and variables of diabetes risk score (n = 923) were excluded, so 7381 subjects remain.

Fig. 1.

Outline of study participants’ selection

Dietary intakes of participants were assessed in Phase 3 in a representative subsample of cohorts’ participants (n = 4920) due to the complexity of dietary data collection, time, and cost savings. The Food Frequency Questionnaire (FFQ) was completed by 3462 subjects and the characteristics of participants who completed the FFQ were similar to those of the total population in Phase 3.

The study adhered to ethical guidelines and was approved by the Institutional Review Board of the Research Institute for Endocrine Sciences at Shahid Beheshti University of Medical Sciences, Tehran, Iran. All the participants provided written informed consent.

Dietary assessment and portfolio diet score

Dietary intake was assessed through face-to-face interviews conducted by expert dietitians using a validated semi-quantitative FFQ [24–26]. The frequency of consumption of each food item was recorded based on portion sizes from the previous year, with portion sizes converted into grams. The Iranian Food Composition Table was used to calculate the macronutrient and micronutrient intake [27]. The Portfolio Diet Score (PDS) was computed following the Glenn et al. methodology, identifying FFQ items related to six Portfolio Diet components: plant protein, viscous fiber, nuts, phytosterols, MUFAs, and saturated fats/cholesterol [28]. Daily servings of relevant foods were combined for each component. Higher scores were assigned to participants who consumed more recommended and less restricted food. Scoring was based on quintiles for each component: five points for the highest quintile of recommended foods (e.g., nuts) and one point for the lowest. For restricted foods and saturated fats/cholesterol, the scoring was inverted. The total scores ranged from six to 30, with higher scores indicating greater Portfolio Diet adherence [28].

Clinical and laboratory measurements

A standard questionnaire was used to collect demographic data including age, education level, physical activity, smoking status, marital status, medication use, and family history of T2DM. Weight was measured using a digital scale and height was measured using a tape meter. BMI was calculated as weight divided by height in meters squared. Waist circumference was measured at the umbilical level by using an upstretched tape meter. Physical activity was assessed using the Modifiable Activity Questionnaire, with results expressed as metabolic equivalent (MET) minutes per week [29]. After a 15-minute resting period, systolic and diastolic blood pressure measurements were taken twice using a mercury sphygmomanometer, with the average of the two readings used in the analysis. Fasting blood samples were collected after–a 12–14 h fast to measure fasting plasma glucose (FPG), lipid, and other biomarkers. Plasma glucose and triglyceride levels were determined using an enzymatic colorimetric method, while high-density lipoprotein cholesterol (HDL-C) was measured after the precipitation of apolipoprotein B-containing lipoproteins. Participants who were not taking glucose-lowering medications also underwent an oral glucose tolerance test. The intra-assay and inter-assay coefficients of variation were less than 2.2% for FPG, 1.9% for triglycerides, and 2.9% for HDL-C.

Definition of terms

T2DM was defined according to the American Diabetes Association (ADA) criteria: FPG ≥ 126 mg/dl, 2-hour post-challenge plasma glucose (2-h-PCPG) ≥ 200 mg/dl, or a history of glucose-lowering medication use [30].

The diabetes risk score was estimated based on five variables, including systolic blood pressure, a family history of diabetes, waist-to-height ratio, Triglyceride/HDL-C, and FPG [31]. This score could better identify the development of T2DM in a Middle Eastern adult population [31].

Statistical analysis

Baseline characteristics were reported as means (standard deviations) or counts (percentages) across tertiles of PDS using ANOVA and chi-square tests. Cox proportional hazard models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for the risk of T2DM across PDS tertiles. Three models were developed: Model 1 was crude analyses; Model 2 was adjusted for age, sex, total energy intake, physical activity, dietary cholesterol, and diabetes risk score, while Model 3 was adjusted for BMI. Statistical analyses were performed using SPSS version 18.

Results

The mean (SD) age and BMI of the total population were 40.6 (13.3) years and 27.0 (4.7) kg/m², respectively. The median of the portfolio diet score was 18 [interquartile range (IQR) = 14, 21] in the total population. The baseline characteristics of the participants across tertiles of PDS are reported in Table 1. Participants in the highest tertile of the PDS were older, more likely to be married, and had higher levels of physical activity. There were no significant differences in gender, smoking status, educational attainment, and BMI across tertiles. A higher PDS was associated with a significantly higher intake of total energy, carbohydrates, protein, fat MUFA, polyunsaturated fatty acids (PUFA), and saturated fatty acids (SAF), and fiber intake as well as fruits, vegetables, legumes, fish and poultry, nuts, and seeds. Participants in the highest PDS tertile also consumed more red meat (P < 0.0001); however, there were no significant differences in consumption of dairy products, whole grain, refined grain, and processed meat across tertiles.

Table 1.

Baseline characteristics of participants across tertiles of portfolio diet score (PDS)

	T1	T2	T3	P-value
Female	413(54.3%)	449(53.3%)	324(56.6%)	0.461
Age (year)	39.17 ± 13.37	40.57 ± 13.52	41.97 ± 13.06	0.001
Academic degrees	190(24.9%)	227(26.9%)	177(30.6%)	0.247
Marital status, married	582(76.3%)	642(76.1%)	473(81.4%)	0.034
Smoker	112(14.7%)	95(11.3%)	65(11.2%)	0.196
Physical activity (MET min-week)	36.3 ± 56.6	31.5 ± 49.7	39.1 ± 59.7	0.028
BMI	26.9 ± 4.8	26.9 ± 4.7	20.3 ± 11.4	0.052
Diabetes risk score	19.3 ± 10.6	20.4 ± 11.4	20.3 ± 11.4
Nutrients
Total energy (kcal/d)	1957 ± 660	2288 ± 687	2565 ± 689	< 0.0001
Carbohydrate (g/d)	276.7 ± 103.7	330.8 ± 108.5	377.5 ± 110.1	< 0.0001
Protein (g/d)	66.5 ± 25.3	77.5 ± 26.2	88.5 ± 27.3	< 0.0001
Fat (g/d)	69.1 ± 28.6	79.3 ± 31.5	87.5 ± 31.3	< 0.0001
PUFA	14.2 ± 6.9	16.5 ± 7.9	18.6 ± 8.5	< 0.0001
MUFA	23.9 ± 10.2	27.4 ± 11.5	30.3 ± 11.7	< 0.0001
SFA	23.9 ± 10.2	27.5 ± 22.5	27.9 ± 11.1	< 0.0001
Fiber	30.7 ± 20.7	37.7 ± 19.9	45.8 ± 19.7	< 0.0001
Diet
Whole grain (g/d)	82.7 ± 107.2	91.4 ± 105.0	95.2 ± 100.0	0.071
Refined grain (g/d)	285.8 ± 153.3	271.3 ± 157.8	265.2 ± 150.5	0.137
Vegetables (g/d)	267.5 ± 169.5	355.9 ± 217.9	441.8 ± 252.5	< 0.0001
Fruits (g/d)	189.5 ± 136	373.9 ± 232	579.8 ± 325	< 0.0001
Dairy products (g/d)	354.5 ± 241	347.7 ± 233.5	366.4 ± 253.1	0.537
Red Meat (g/d)	39.9 ± 33.3	47.7 ± 33.7	57.7 ± 45.3	< 0.0001
Processed meat	2.3 ± 5.0	2.1 ± 4.3	2.1 ± 4.3	0.716
Fish and Poultry	36.9 ± 35.6	36.9 ± 34.3	43.3 ± 31.7	0.012
Legumes (g/d)	11.3 ± 14.0	17.3 ± 18.4	26.9 ± 25.3	< 0.0001
Nuts and seeds (g/d)	7.7 ± 23.6	12.99 ± 25.84	20.95 ± 29.3	< 0.0001
PDS components
Plant Protein (serving/d)	0.12 ± 0.12	0.22 ± 0.28	0.36 ± 0.37	< 0.0001
Viscose Fiber (serving/d)	0.37 ± 0.28	0.78 ± 0.52	1.19 ± 0.74	< 0.0001
Nuts (serving/d)	0.09 ± 0.14	0.21 ± 0.29	0.45 ± 0.58	< 0.0001
Total Phytosterols (mg/d)	43.7 ± 28.8	81.9 ± 33.9	129.9 ± 52.6	< 0.0001
MUFA (serving/d)	0.04 ± 0.09	0.15 ± 0.37	0.37 ± 0.64	< 0.0001
SFA/Cholesterol (serving/d)	2.6 ± 1.5	2.9 ± 1.6	2.8 ± 1.4	0.006

MET, metabolic equivalent; BMI, body mass index; PUFA, polyunsaturated fatty acids; MUFA, mono-unsaturated fatty acids; SFA, saturated fatty acids

Over a median follow-up of 8.9 years, 234 new cases of T2DM were identified. Table 2 presents the HRs for the risk of T2DM across PDS tertiles. In Model 1, participants in the highest PDS tertile had no association with T2DM (HR = 1.08; 95%CI = 0.78–1.51, P trend = 0.645) compared to those in the lowest tertile. After adjusting for age, sex, total energy intake, physical activity, and diabetes risk score (Model 2), adherence to PDS was not associated with T2DM risk (HR = 0.97; 95% CI: 0.68–1.39, P trend = 0.863). Further adjustment for BMI (Model 3) did not change the findings (HR = 0.91; 95% CI: 0.64–1.31, P trend = 0.610).

Table 2.

Multivariable adjusted hazard ratio (95% CI) for type 2 diabetes across tertiles of portfolio diet score

	T1	T2	T3	P trend
Median (IQR)	13 (11–14)	18 (17–19)	23 (22–24)
Model 1	1	1.00 (0.73–1.36)	1.08 (0.77–1.51)	0.645
Model 2	1	0.91 (0.66–1.39)	0.97 (0.68–1.39)	0.863
Model 3	1	0.90 (0.65–1.23)	0.91 (0.63–1.30)	0.610

Model 1: crude

Model 2: adjustment for age, sex, total energy, physical activity, and diabetic risk score

Model 3: further adjustment for BMI at baseline

Analysis of the individual PDS components (Table 3) revealed that greater nut consumption was associated with a significantly lower risk of developing T2DM in the adjusted models. In the fully adjusted model, the highest tertile of nut consumption was associated with a 28% reduction in diabetes risk (HR = 0.72, 95% CI: 0.51–0.99, P = 0.050) compared to the lowest tertile, suggest component-specific effects. Other dietary components, including plant protein, viscose fiber, MUFA, and total phytosterols, showed no significant association with T2DM risk in the fully adjusted models.

Table 3.

Multivariable adjusted hazard ratio (95% CI) for type 2 diabetes across tertiles of portfolio diet score component

	T1	T2	T3	P trend
Plant Protein (serving/d)
Median (IQR)	0.08 (0.04–0.13)	0.14 (0.07–0.25)	0.24 (0.13–0.49)
Model 1	1	0.83 (0.61–1.14)	0.81 (0.59–1.11)	0.899
Model 2	1	0.80 (0.58–1.10)	0.94 (0.67–1.32)	0.712
Model 3	1	0.81 (0.58–1.12)	1.00 (0.71–1.39)	0.955
Viscose Fiber (serving/d)
Median (IQR)	0.30 (0.18–0.49)	0.67 (0.43–0.99)	1.07 (0.76–1.43)
Model 1	1	0.88 (0.63–1.22)	1.06 (0.77–1.45)	0.691
Model 2	1	0.93 (0.67–1.30)	1.05 (0.74–1.47)	0.765
Model 3	1	0.90 (0.64–1.25)	0.90 (0.64–1.27)	0.574
Nuts (serving/d)
Median (IQR)	0.06 (0.02–0.11)	0.14 (0.07–0.25)	0.29 (0.15–0.50)
Model 1	1	0.88 (0.64–1.21)	0.85 (0.623–1.17)	0.336
Model 2	1	0.80 (0.58–1.10)	0.73 (0.52–1.02)	0.067
Model 3	1	0.84 (0.61–1.15)	0.71 (0.51–0.99)	0.050
MUFA (serving/d)
Median (IQR)	0.01 (0.001–0.05)	0.05 (0.01–0.14)	0.19 (0.07–0.43)
Model 1	1	1.20 (0.87–1.67)	1.29 (0.93–1.79)	0.121
Model 2	1	1.09 (0.78–1.51)	1.19 (0.85–1.65)	0.302
Model 3	1	1.08 (0.78–1.50)	1.13 (0.81–1.57)	0.462
Total Phytosterols (mg/d)
Median (IQR)	40.3 (28.4–54.6)	75.9 (60.2–98.1)	120.8 (95.9-151.4)
Model 1	1	0.91 (0.66–1.26)	0.97 (0.71–1.34)	0.893
Model 2	1	0.85 (0.61–1.18)	0.87 (0.61–1.24)	0.455
Model 3	1	0.82 (0.59–1.15)	0.78 (0.55–1.11)	0.176
SFA/Cholesterol (serving/d)
Median (IQR)	2.25 (1.65–3.35)	2.54 (1.79–3.56)	2.66 (1.64–3.64)
Model 1	1	0.76 (0.55–1.03)	0.66 (0.48–0.91)	0.011
Model 2	1	0.93 (0.67–1.28)	0.92 (0.63–1.35)	0.676
Model 3	1	0.92 (0.66–1.27)	0.89 (0.61–1.31)	0.553

IQR, inter-quartile range

Model 1: crude

Model 2: Model 1 + adjustment for age, sex, total energy, physical activity, and diabetic risk score

Model 3: Model 2 + adjustment for BMI at baseline

Discussion

In the present population-based cohort study, we assessed the adherence to the Portfolio Diet among 2188 T2DM-free Tehranian adults, and the association between PDS and T2DM risk was then investigated through 8.9 years of follow-up. Our findings indicated no significant association between overall PDS and risk of T2DM incidence; however, we found an inverse association between nuts consumption (as a component of portfolio diet) and risk of T2DM.

There is limited data regarding the association between adherence to the portfolio diet and the risk of T2DM. A recent cohort study of 145,299 postmenopausal women, evaluated the adherence to the portfolio diet and the association of the PDS and incident T2DM after 16 years of follow-up [21]. The study showed a significant inverse association between PDS and T2DM risk. In a pooled analysis of two randomized dietary trials on subjects with T2DM, the effects of low glycemic index interventions were compared to high cereal fiber control diets. Adherence to the Portfolio Diet and its components was assessed using the validated clinical Portfolio Diet Score. The results indicated that the Portfolio Diet was linked to a reduction in HbA1c levels over six months, primarily influenced by two key components: nuts and seeds, as well as plant protein, especially from dietary pulses [20].

In the current study, PDS was not associated with T2DM risk in adults. The differences between our findings and the aforementioned studies could be attributed to unique local dietary habits. Notably, we observed that higher Portfolio score was associated with increased consumption of red meat, which may have partially counteracted the potential benefits of the Portfolio diet. Furthermore, certain dietary food groups, such as grains, are not included in the PDS, which could also influence the results. In the current research, we found no significant differences in the intake of whole and refined grains across the tertiles of the PDS. The Iranian population tends to consume more carbohydrates compared to Western populations, and this higher carbohydrate intake has been linked to an increased risk of T2DM [32, 33]. Consequently, excluding whole and refined grains from the PDS may explain why no association was observed between this dietary pattern and T2DM. Evidence indicates that consuming whole grains, vegetables, and fruits while reducing intake of refined grains, sugary beverages, and processed meats is more effective in preventing T2DM. Notably, these important dietary aspects are not incorporated into the Portfolio diet [34]. Moreover, the scoring method used in the PDS may also contribute to the lack of observed correlation. The PDS assigns scores from 1 (least adherent) to 5 (most adherent) based on quintile intake of six components, without considering the actual amount consumed within each subgroup. Interestingly, the median intake in our study are lower than those reported in previous research [20, 21]. For example, a study involving menopausal women showed higher consumption of legumes, plant sterols, and MUFA compared to our findings [20, 21]. Conversely, our participants had higher intakes of SFA than those previous studies [20, 21]. Therefore, it appears that the PDS might not be the most effective dietary tool for preventing T2DM in the Tehranian population, though further research is necessary.

Despite the limited number of studies on the association between adherence to the Portfolio Diet and the incidence of T2DM, some previous observational studies investigated the association between this dietary pattern and the risk of CVD and other intermediate risk factors for T2DM. For instance, the findings from three cohort studies (the Nurses’ Health Study, the Nurses’ Health Study II, and the Health Professionals Follow-up Study), with 73,924 participants during up to 30 years of follow-up, indicated a significantly lower risk of CVD for participants who had higher adherence to the portfolio diet [35]. Regarding other cardio-metabolic risk factors associated with T2DM, adherence to the Portfolio diet had beneficial effects on blood pressure among hyperlipidemic subjects [36]. Moreover, higher PDS was associated with lower blood levels of HbA1c, fasting glucose, triglycerides, waist circumference, and BMI among older women with metabolic syndrome [19].

The Portfolio dietary pattern is specifically tailored to lower total cholesterol levels. This plant-based diet incorporates specific nutrients—such as phytosterols, viscous fibers, and monounsaturated fatty acids (MUFA)—that help to lower cholesterol and thereby reduce the risk of cardiovascular disease (CVD) [8, 36]. While this dietary pattern and its components are associated with a lower risk of CVD, meta-analyses have found no significant link between these components and the development of T2DM [37–39]. It should be noted that within plant-based diets, certain components—such as bioactive-rich plant-source foods—are more effective in reducing the risk of chronic diseases like T2DM, whereas others, including dairy, eggs, fish, and poultry, may have less influence [40]. In a randomized controlled trial, the Portfolio Diet reduced HbA1c levels over six months in subjects with T2DM, mainly due to two key components: nuts and seeds, and plant protein [20]. Our findings also suggest that among the components of the Portfolio diet, only nut consumption was associated with a decreased risk of T2DM. This aligns with previous studies we conducted as part of the TLGS, which reported that consuming more than four servings of nuts per week decreased the risk of chronic diseases such as T2DM and metabolic syndrome [41, 42]. Conversely, some meta-analyses of prospective cohort studies found no significant association between total nuts consumption and T2DM risk [10, 43]. The variability in in nut consumption across different countries, as noted by Russell J. de Souza et al., explain these conflicting findings [44]. For instance, in countries with high nut consumption like the UAE, Zimbabwe, Iran, Canada, Poland, Turkey, Tanzania, and Palestine, about 64.5% of participants consumed at least 30 g of nuts weekly, with a median intake of 7.4 g per day (IQR: 2.0–17.6 g). In comparison, in our current study, the median weekly nut intake was 41.9 g (IQR: 18.6–101.6 g) over a follow-up of 8.9 years. These findings align with nutritional epidemiological evidence, which underscores that even small but consistent long-term intakes of specific food groups can exert significant physiological effects. For example, Liu et al. showed that increasing nut consumption by at least 0.50 servings/day over four years was associated with reduced risks of CVD, coronary heart disease, and stroke, regardless of initial intake levels and lifestyle factors [45].

There are some potential mechanisms underlying the protective effects of nuts consumption on T2DM risk, including the nutrient profile of nuts which contributed to a low glycemic index, high content of soluble fibers [46], high amounts of high-quality fats including unsaturated fatty acids, in particular, long-chain omega-6 and omega-3 PUFA [47, 48], high amounts of vitamin E, magnesium, and selenium, and bioactive compounds, including polyphenols of various types (e.g., flavonoids, phenolic acids, stilbenes, lignans, other polyphenols), that modulate gut microbiota [49]. On the other hand, according to the results of an SRMA of six prospective cohort studies (n = 569,910) and 86 randomized control trials (n = 5873), nuts consumption was associated with reduced risk of overweight and obesity, as well as reductions in body weight and body fat [50], which may explain the inverse association observed between nuts consumption and risk of T2D. Moreover, the bioactive compounds of the nuts, such as fibers and polyphenols, could play as a prebiotic and modulate the composition and function of the gut microbiome, and induce their beneficial effects on glucose homeostasis via short-chain fatty acids metabolites of fermentation [51]. Finally, there is evidence to suggest that some components of nuts could modify gene expression related to insulin resistance or T2DM [52, 53].

Recent studies have increasingly highlighted the protective effects of plant-based dietary patterns against the risk of developing T2DM. It has been reported that individuals with a higher overall plant-based diet index (based on a higher intake of plant-source foods and a lower intake of animal-source foods) had a significantly lower risk of T2DM [40]. Sullivan et al. categorized overall plant-based diets into healthy plant-based diet index (hPDI) which represents greater healthy plant food intake and lower intakes of animal-derived and unhealthy plant foods, and unhealthy plant-based diet index (uPDI) which represents greater unhealthy plant food intake and lower intakes of animal-derived and healthy plant foods [54]. They reported that hPDI scores were inversely associated with diabetes risk, although uPDI scores were not associated with diabetes risk. Notably, the non-significant inverse association between overall PDS and T2DM in our study may be related to the fact that we did not distinguish between healthy and unhealthy plant sources in calculating PDS. As was demonstrated in the aforementioned study, consumption of processed plant-based foods like refined grains, potatoes, sweets, and sugar-sweetened beverages (SSBs) may increase the risk of T2DM, while consumption of fiber- and bioactive-rich plant-based foods like fruits, vegetables, nuts, legumes, and whole grains, has beneficial effects on T2DM risk [54]. This finding is also consistent with our finding regarding the inverse association between nuts consumption and the risk of T2DM.

The strengths of the present study include the prospective cohort design, the relatively large sample size and long-term follow-up period, the use of a validated FFQ to estimate dietary intakes, and the availability of detailed data on potential confounders. Our study also had some limitations. First, we used a FFQ for assessing dietary intakes, so measurement error was not excluded. Although we adjusted for a wide range of potential confounders, we were unable to evaluate the impact of genetic and environmental factors. Further studies should integrate potential factors and biomarkers with dietary data to reduce misclassification.

In this Tehranian cohort, nut intake _but not overall portfolio diet adherence_ was associated with reduced T2DM risk. This underscores the need to evaluate both dietary patterns and their components in context-specific ways. The lack of association for the full dietary pattern may reflect the influence of unique local dietary behaviors—such as higher red meat consumption among individuals with greater adherence to the Portfolio Diet—which could attenuate the intended benefits of this plant-based dietary pattern. Moreover, the Portfolio Diet Score was originally developed to assess cardiovascular risk through lipid-lowering mechanisms and may not be sufficiently sensitive to dietary factors most relevant to glycemic control or diabetes prevention in Middle Eastern populations. These findings suggest that both culturally specific dietary practices and the scoring structure of the PDS may influence the observed associations. Future research should focus on developing and validating regionally adapted dietary indices that better capture context-specific dietary exposures relevant to type 2 diabetes risk.

Abbreviations

T2DM

Type 2 diabetes mellitus

MENA

Middle East and North Africa

MUFA

Monounsaturated fats

PUFA

Polyunsaturated fatty acids

SFA

Saturated fatty acids

CVD

Cardiovascular disease

LDL-C

Low-density lipoprotein cholesterol

HbA1c

Hemoglobin A1C

FBS

Fasting blood sugar

WHI

Women’s Health Initiative

TLGS

Tehran Lipid and Glucose Study

BMI

Body mass index

FFQ

FOOD Frequency Questionnaire

PDS

PORTFOLIO Diet Score

MET

Metabolic equivalent

FPG

Fasting plasma glucose

HDL-C

High-density lipoprotein

ADA

American Diabetes Association

2-h-PCPG

2-hour post-challenge plasma glucose

HRs

Hazard ratios

SSBs

Sugar-sweetened beverages

Author contributions

Hanieh Malmir and Somayeh Hosseinpour-Niazi conceptualized and designed the study. Hanieh Malmir and Somayeh Hosseinpour-Niazi analyzed and interpreted the data; Hanieh Malmir, Somayeh Hosseinpour-Niazi, Zahra Gaeini, Parvin Mirmiran, and Fereidoun Azizi drafted the initial manuscript; Somayeh Hosseinpour-Niazi and Parvin Mirmiran supervised the project; all authors read and approved the final version of the manuscript.

Funding

This work was supported by Shahid Beheshti University of Medical Sciences, Tehran, Iran (43015937).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

The study approved by the Institutional Review Board of the Research Institute for Endocrine Sciences at Shahid Beheshti University of Medical Sciences, Tehran, Iran. All the participants provided written informed consent.

Consent for publication

The authors have not agreed to open access publication of this article.

Competing interests

The authors declare no competing interests.