Results from the current study indicated that among micronutrients, sodium, potassium, calcium, magnesium, zinc, iodine, carotene, vitamin D, tryptophan and vitamin B12 had an inverse relationship with the incidence of T2D while, phosphate, iron, chloride, and riboflavin intake were positively associated with the risk of diabetes. In the macronutrient group, water, protein, and fiber intake had a protective effect against the incidence of diabetes while carbohydrate saturated fat, mono fat, and cholesterol had a positive correlation with the incidence of T2D. It seems that the increasing prevalence of type 2 diabetes is due to increased obesity, decreased physical activity, and dietary changes. Many observational studies have shown that macronutrient (protein, carbohydrate, fiber, and fat) and micronutrient (vitamins and minerals) intake are associated with the incidence of diabetes [27,28,29,30].

Our DT model indicated that potassium, tryptophan, iodine, sodium, and riboflavin intake had the most significant effect on the incidence of diabetes, respectively. In the subgroup with higher potassium, lower iodine, higher magnesium, and lower tryptophan intake, the risk of diabetes was approximately 98%. According to our results, in the subgroup with lower potassium intake, lower tryptophan intake, and lower sodium intake, the incidence percentage of diabetes for both subgroups of higher magnesium and lower magnesium and lower zinc intake was 96%.

Magnesium deficiency plays a key role in the pathogenesis of diabetes [31]. Lower magnesium levels can inactivate cellular defenses against oxidative stress effects [32]. It was also shown that In over 300 enzymatic reactions, magnesium is a necessary cofactor, specifically in all phosphorylation reactions and in general in all reactions involving the utilization and transfer of ATP, such as cellular responses to growth factors and cell proliferation [33, 34]. Through its role as an enzyme cofactor, magnesium may directly influence glucose metabolism in cells [35, 36] and may interact with calcium homeostasis to influence insulin secretion [37]. In the present study, magnesium intake was higher in T2D patients, which correlates with other research results [38, 39]. We also found that magnesium intake had an inverse association with the incidence of diabetes, which is consistent with previous investigations [39, 40].

We found that thiamin, vitamin C, vitamin D, riboflavin, folate, and niacin levels were higher in the T2D group while there were no significant differences in vitamin B12 intake between the case and control group. Furthermore, vitamin B12 and vitamin D intake had an inverse correlation with T2D incidence, while riboflavin had a positive association with the incidence of T2D. There was no significant association between thiamin, vitamin C, folate, and niacin intake with T2D incidence. Walter et al. [41] demonstrated that thiamin was higher in diabetic patients, while there were no significant differences between vitamin C, vitamin B12, and vitamin D between cases and controls. In contrast to our results, it was shown that thiamin had lower levels in diabetics [42]. This inconsistency may be due to multi-vitamin consumption or differences in sample size. A cohort study has shown a positive relationship between riboflavin intake and the prevalence of diabetes [43]. Our results are consistent with previous studies which found that there is an inverse correlation between vitamin D levels and the incidence of diabetes [44, 45]. However, in a few studies, there was no significant association between vitamin D intake and the incidence of diabetes [17]. Some studies found that serum vitamin B12 had no significant differences between T2D patients and nondiabetics in agreement with our results [46].

The results of this study indicated that phosphate and calcium levels were higher in the case group. These results agree with previous studies [47]. However, a few studies demonstrated that levels of phosphate and calcium were lower in the T2D group [48]. Furthermore, we indicated that there is an inverse correlation between calcium intake and the incidence of diabetes, while this association was positive between serum phosphate and the risk of diabetes.

This study supports evidence from previous investigations [16, 49].

We indicated that iron intake has a positive association with the occurrence of diabetes. This result reflects those of a prospective cohort study which also found a positive correlation between iron intake and the incidence of T2D in women [50]. However, another prospective cohort study between the men population indicated no significant association between total iron intake and the incidence of diabetes but a positive correlation between Heme-iron intake from red meat and the risk of diabetes [51].

Zinc deficiency is related to increased oxidative stress and many clinical complications, such as impaired wound healing and taste acuity in diabetic patients [52]. We indicated that zinc intake was higher in T2D patients. We also showed that there is an inverse relationship between zinc intake and the incidence of diabetes. These results are in line with those of previous studies [53, 54]. However, a study in the United States doesn’t support the thesis that zinc status is inversely associated with the incidence of diabetes [55]. There may be a few explanations for this finding: (1) the mean age for this study was 27.03 which is lower than the mean age of our study (52.18); (2) average zinc intake was higher (16.7 mg/day) in comparison to our study (9.77 mg/day).

There are few studies about the relationship between iodine intake and the incidence of diabetes. A cohort study has shown a positive correlation between iodine intake and the incidence of diabetes [56]. However, the current study findings do not support the previous research. We indicated an inverse relationship between iodine intake and the incidence of diabetes. A possible explanation for this might be that only women were investigated in this study. Another possible explanation for this is that the average iodine intake (155.6 µg/day) was higher than our study (108.18 µg/day). However, further work is required to investigate the relationship between dietary iodine intake and the incidence of diabetes. We found that tryptophan had a protective effect against the incidence of diabetes. This result was in accord with a recent study showing that dietary tryptophan was associated with decreased risk of T2D [57]. However, a few studies do not support our results [58].

We demonstrated that potassium intake had an inverse correlation with the risk of diabetes. This result is in agreement with previous studies [59]. Cardiovascular disease (CVD) risk is higher in T2D patients [60]. It was shown that higher sodium intake was associated with hypertension and an increased risk of CVD [61, 62]. Iran had one of the highest salt intakes worldwide at 9.52 g/day [63]. The current study indicated that sodium intake was higher in T2D patients. Furthermore, there was no significant association between sodium intake and the development of diabetes. These results are consistent with previous studies [64]. We found that chloride intake had a positive relationship with the incidence of T2D. Some studies have shown that there is a positive correlation between salt intake and the incidence of diabetes [65]. While some studies indicated that lower dietary salt intake was associated with an increased risk of diabetes [66]. Further research should be undertaken to investigate the association between salt intake and the incidence of diabetes.

According to our DT algorithm, protein, carbohydrate, cholesterol, and fiber intake had the most crucial effect on the risk of T2D, respectively. We indicated that in the subgroup with higher protein and carbohydrate intake the risk of diabetes development was 94%. In the subgroup with lower protein, lower cholesterol, lower carbohydrate, lower saturated fat, and lower fiber intake, the incidence of diabetes was 86%.

In this study, water intake was higher in T2D patients than in the control group. We also indicated that water intake had a protective effect against T2D incidence. These results are consistent with a recent study showing an inverse relationship between water intake and the incidence of diabetes among men and women [67]. The hyperglycemia produce the symptoms of repeated urination, increased thirst, and increased hunger [68]. Also, we found that carbohydrate intake was higher in T2D patients. Another result was that carbohydrate intake positively affected the incidence of diabetes. These results agree with previous studies [28]. Another result of our study was a small inverse relationship between fiber and protein intake and the incidence of diabetes. These results support other studies [38, 69]. However, A cohort study of the European population indicated that high protein intake was associated with a small elevated risk of T2D [66, 70]. It seems possible that this inconsistency may be due to differences in dietary habits and animal protein consumption or plant protein intake.

We demonstrated that cholesterol intake had no significant differences between the case and control groups. Furthermore, we indicated that there is a positive correlation between dietary cholesterol intake and the incidence of diabetes. A prospective cohort study of French women showed the same result as our result [71].

To our knowledge, this is the largest study so far documenting the correlation between micro/macro nutrients and the incidence of diabetes. Another strength of our study was that it has been one of the first studies using the Gini impurity index to construct a DT machine learning model to indicate the correlation between nutrient factors and the incidence of T2D. Furthermore, the sample size and long follow-up duration might be considered the strengths of our study. We also used analysis models (LR and DT) to evaluate the association between nutrient factors and the risk of diabetes more accurately.

This study was limited by the absence of adjusting the incidence estimates by smoking, drug use, body mass index (BMI), family history of diabetes, and sex that may influence the result. More research using control trials is needed to evaluate the effect of micro/macro nutrients on diabetes, especially in iodine and sodium intake.