Relationship between dietary intake and atherogenic index of plasma in cardiometabolic phenotypes: a cross-sectional study from the Azar cohort population

Abstract

Background

Cardiovascular diseases are a leading cause of global mortality, with diet playing a key role in their progression. The Atherogenic Index of Plasma (AIP) is a predictive marker for cardiovascular risk, but its association with dietary intake across cardiometabolic phenotypes remains underexplored. This study investigates the relationship between dietary intake and AIP, hypothesizing that energy intake and macronutrients influence AIP and, consequently, cardiovascular risk.

Methods

This cross-sectional study analyzed data from 9,515 participants aged 35–55 in the Azar cohort study. Based on Body Mass Index (BMI) and metabolic syndrome (MetS), participants were classified into four phenotypes: metabolically healthy normal weight (MHNW), metabolically unhealthy normal weight (MUHNW), metabolically healthy obese (MHO), and metabolically unhealthy obese (MUHO). Dietary intake was evaluated using a semi-quantitative food frequency questionnaire (FFQ), and AIP was calculated. Adjustments were made for age, gender, socioeconomic status, and physical activity.

Results

A notable difference was observed in demographic and clinical status between cardiometabolic groups of males and females. The AIP was highest in the MUHNW (0.42 for males; 0.28 for females) and lowest in the MHNW (0.05 for males; -0.05 for females, P < 0.001). There was a statistically significant difference in the mean energy intake and the percentage of energy intake from protein among the cardiometabolic phenotypes (p < 0.001). After adjusting for confounders, only weak but meaningful correlations remained for energy, carbohydrate, and protein intake in the MUHO (r = 0.048, P = 0.01; r = 0.057, P = 0.003; and r = 0.050, P = 0.01) and for carbohydrate and lipid intake in the MHO (r = 0.034, P < 0.01 and r = -0.055, P < 0.001).

Conclusion

The study found weak but meaningful correlations between energy, carbohydrate, and protein intake and AIP in the MUHO phenotype and between carbohydrate and lipid intake and AIP in the MHO phenotype. This highlights the role of energy and carbohydrates in AIP within specific subgroups. Future research should focus on the effects of macronutrient combinations on AIP and long-term dietary impacts on metabolic health instead of BMI.

Introduction

Cardiovascular diseases (CVDs) are a leading cause of death globally, influenced by genetic, environmental, and behavioral factors. Key contributors include dyslipidemia, hypertension, diabetes, and obesity, with oxidative stress and inflammation increasing CVD risk [1]. Globally, CVD caused 17.9 million deaths in 2016, with prevalence rising from 257 million in 1990 to 550 million in 2019 [2]. In Iran, the prevalence of CVD among individuals aged 40–64 years was reported as 11.2% in males and 9.0% in females [3]. BMI is a common tool for assessing cardiovascular disease risk, but individuals within the same BMI category can have different cardiometabolic risks. BMI does not consider variations in body fat distribution, which impacts metabolic health, leading to the identification of distinct cardiometabolic risk markers [4, 5].

To address this variation, classifications like metabolically healthy normal weight (MHNW), metabolically unhealthy normal weight (MUHNW), metabolically healthy obese (MHO), and metabolically unhealthy obese (MUHO) have been developed. These phenotypes focus on metabolic health and cardiometabolic risk factors rather than just BMI. Additionally, lifestyle factors, particularly dietary intake, significantly influence these phenotypes [6,7,8].

Diet plays a crucial role in cardiovascular health by affecting lipid profiles and cholesterol levels, which are linked to the risk of CVD [9, 10]. Modifying diets, such as substituting carbohydrates with certain fats, can alter triglycerides (TG) and high-density lipoprotein-cholesterol (HDL-C) levels, impacting lipoprotein metabolism. These changes are vital as they influence the production of very low-density lipoprotein (VLDL) and low-density lipoprotein-cholesterol (LDL-C), key factors in cholesterol metabolism and cardiovascular risk [11].

Some research suggests that certain dietary patterns, such as high-fat or low-carbohydrate diets, are associated with beneficial lipid profiles. In contrast, others report no significant change or even worsening lipid levels [12, 13]. The atherogenic index of plasma (AIP), calculated from the logarithmic ratio of triglyceride (TG) to HDL-C, has recently been utilized in studies to evaluate the relationship between dietary intake and CVD risk [14, 15]. AIP reflects lipoprotein particle size and atherogenicity, providing more information than traditional lipid profiles. It shows an inverse relationship with LDL-C particle size, where higher AIP correlates with smaller, more atherogenic LDL-C particles linked to an increased cardiovascular risk [16]. Macronutrients, particularly those high in saturated fats, may contribute to increased TG levels and reduced HDL-C levels. Conversely, diets rich in unsaturated fats, such as the Mediterranean diet, are associated with decreased TG levels and increased HDL-C levels [17]. Dietary-induced changes in lipid profiles may impact the AIP, a key marker of CVD. Previous studies have explored the relationship between AIP and dietary intake using various methods but have not directly or simultaneously examined this relationship within the context of cardiometabolic phenotypes [14, 15, 18,19,20]. For instance, these studies often employed varying designs, focused on single-sex populations, included small sample sizes or lacked stratification by BMI or metabolic syndrome indicators [20,21,22]. However, the role of dietary intake in influencing AIP within different cardiometabolic phenotypes remains poorly understood. These phenotypes, which integrate metabolic and cardiovascular risk factors, provide a valuable framework for developing targeted dietary interventions to enhance the predictive value of AIP in cardiovascular disease risk assessment. Furthermore, the limited focus on energy intake and macronutrients in relation to AIP across cardiometabolic phenotypes highlights significant knowledge gaps. To the best of our knowledge, this is the first study to explore the association between dietary intake and AIP across cardiometabolic phenotypes, utilizing a large population from the Azar cohort and hypothesizing that dietary intake impacts AIP in these phenotypes.

Methods and materials

Study design and setting

This cross-sectional analysis is derived from the Azar Cohort Study, which is a part of the Prospective Epidemiological Research Studies in Iran (PERSIAN) [23]. The detailed cohort profile outlines the initiation of the Azar Cohort in 2014, which included three phases: a pilot phase, an enrollment phase, and a 15-year continuous follow-up of participants [24]. For this study, data from the pilot and enrollment phases were utilized. Eligibility for the Azar Cohort Study required participants to be aged 35 to 70 years, to have resided in Shabestar City for at least nine months, and to have at least one Azeri parent. Individuals with mental or physical disabilities were excluded. All participants provided written informed consent and were thoroughly briefed on the study procedures. In the present study, the inclusion criteria were limited to individuals aged 35 to 55 years. The exclusion criteria were as follows: pregnant females, individuals with a history of cancer, and those whose total daily energy intake fell outside of the specified ranges: less than 1200 kcal or more than 4200 kcal for males, and less than 1000 kcal or more than 3500 kcal for females. Additionally, individuals with incomplete data were excluded. Based on these criteria, from the total 15,001 participants in the Azar Cohort Study, 9,515 males and females were included in this analysis (Fig. 1). This study was approved by the Ethical Board of the Research Council of Tabriz University of Medical Sciences (No. IR.TBZMED.REC.1402.428).

Fig. 1

Flow chart of participant selection

Data collection

The participants’ socioeconomic status (SES) was assessed using the Wealth Score Index (WSI), which was derived through multiple correspondence analysis. The WSI for each participant was calculated by evaluating their ownership of various permanent assets (such as a TV, dishwasher, and car), the condition of their living space (including ownership type and number of rooms), and their level of education [23]. Participants were categorized into five WSI quintiles, from the lowest to the highest (first to fifth quintile). To measure physical activity levels, the metabolic equivalent of task (MET) was used. MET quantifies the energy expenditure of an individual based on their body weight, with 1 MET representing the oxygen consumption of 3.5 mL per kilogram of body weight per minute at rest [25, 26]. Participants were classified into three activity levels: low, moderate, and high. Smoking status was assessed in two categories: non-smoker and smoker. The smoking scale consisted of two options: if respondents reported that they had never smoked, had smoked fewer than 100 cigarettes in their lifetime, or had quit smoking more than a year ago, they were defined as “non-smokers.” If they reported that they were currently smoking or how many cigarettes a day they smoked, they were defined as “smokers [27].” Alcohol consumption was assessed using two options (yes/no).

Food intake assessment

The participants ‘dietary intake was evaluated using a semi-quantitative FFQ developed specifically for the PERSIAN cohort study. This semiquantitative, interviewer-administered questionnaire comprised 130 items, designed to capture participants’ usual intake of various food items over the past year. Participants reported the frequency of consumption (daily, weekly, monthly, or yearly) and the portion size for each item. To improve the accuracy of portion size estimation, actual household measures such as dishes, cups, and utensils, as well as several portion size models, were shown to participants during the interview. Additionally, a 64-picture album featuring standard portions of common foods, including bread, fruits, and vegetables, was provided when necessary, and based on the traditional Iranian and United States Department of Agriculture (USDA) food composition tables, daily intakes of energy and nutrients were determined. The validity and reproducibility of this FFQ were rigorously evaluated within the PERSIAN cohort’s pilot study. The Spearman correlation coefficients (DEA-SCC) for food group intake ranged from 0.23 to 0.79. Strong correlations were observed for staple foods such as grains, oils, and sugar-rich items, reflecting their frequent consumption in the Iranian diet [28].

Biochemical and clinical measurement

To evaluate biochemical markers, blood samples were collected after a fasting period of 12 h. TG and HDL-C levels were measured using enzymatic colorimetric methods with commercially available kits from Pars Azmoun, Tehran, Iran. Fasting blood glucose (FBG) was measured using the enzymatic colorimetric method with glucose oxidase.

Blood pressure was assessed using a Richter sphygmomanometer. Trained personnel conducted the measurements while participants were seated and rested for 10 min. Two readings were taken from each arm, with a one-minute interval between consecutive measurements. The average of the second readings from the right and left arms was calculated to determine the blood pressure level.

Anthropometric measurements

The evaluated anthropometric parameters included height, weight, BMI, and Waist circumference (WC).

Weight was measured using a Seca scale (Hamburg, Germany) with an accuracy of 0.1 kg, with the participants standing without shoes and wearing minimal clothing. Height was measured using a tape measure, with the participants standing without shoes and their shoulders in a neutral position. WC was measured with a tape measuring at the level between the lower costal edge and the iliac crest. BMI was calculated by dividing weight (kg) by the square of height (m²).

Definition of different cardiometabolic phenotypes

Metabolic syndrome (MetS) is defined when a person meets at least three of the following criteria, as specified by the National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) [29]:

• FBG ≥ 100 mg/dL or use of hypoglycemic drugs;

• Serum TG ≥ 150 mg/dL or use of triglyceride-lowering drugs;

• HDL-C ≤ 40 mg/dL in males or ≤ 50 mg/dL in females, or use of -HDL-C-raising drugs;

• Systolic blood pressure (SBP) ≥ 130 mmHg or diastolic blood pressure (DBP) ≥ 85 mmHg, or use of blood pressure-lowering drugs;

• Waist circumference > 102 cm in males or > 88 cm in females.

We defined a BMI of < 25 kg/m² as normal weight and a BMI of ≥ 25 kg/m² as overweight or obese. In our study, we grouped the subjects into four cardiometabolic phenotypes based on their BMI and the presence of MetS. The categories are as follows:

Metabolically healthy normal weight (MHNW)

defined by a BMI under 25 kg/m² and having fewer than three risk factors associated with MetS.

Metabolically unhealthy normal weight (MUHNW)

defined by a BMI under 25 kg/m² but with at least three risk factors indicative of MetS.

Metabolically healthy obese (MHO)

defined by a BMI of 25 kg/m² or above, and having fewer than three risk factors associated with MetS.

Metabolically unhealthy obese (MUHO)

defined by a BMI of 25 kg/m² or above who exhibits three or more risk factors for MetS.

Calculation of atherogenic index of plasma (AIP)

$$\:Atherogenic\:Index\:Of\:Plasma=log10\frac{\left[Triglyceride\right]}{\left[HDL\:cholesterol\right]}$$

AIP value from 0.11 to 0.21 indicates an intermediate risk, while AIP > 0.21 suggests an increased risk [16].

Statistical analysis

Data analysis was performed using SPSS 23.0 software (SPSS Inc., Chicago, IL, USA). The normality of the data was assessed using the Kurtosis and Skewness index, along with descriptive statistics. Quantitative variables with a normal distribution were reported as mean and standard deviation, while those with a non-normal distribution were reported as median and interquartile range. Qualitative variables were reported as frequency (percentage). A one-way analysis of variance (ANOVA) was used to compare quantitative variables between groups, and the LSD post hoc test was applied for multiple comparisons to determine statistical significance. The Chi-square test and Kruskal-Wallis test were used to compare qualitative variables. The Pearson correlation method was applied to examine the unadjusted association between dietary intake and AIP across cardiometabolic phenotypes. Subsequently, partial correlation analysis was performed to control for potential confounding variables. For the first adjustment, SES and physical activity were accounted for. In the second adjustment, age and gender were further included alongside SES and physical activity. P-values < 0.05 were considered statistically significant.

Results

The baseline clinical and biochemical characteristics of participants, categorized by gender and cardiometabolic phenotypes, are summarized in Table 1. The mean age of male and female participants was 45.36 ± 5.90 years and 44.74 ± 5.86 years, respectively. Among males, those with metabolically healthy phenotypes (MHNW, MHO) had a higher prevalence of college education compared to those with metabolically unhealthy phenotypes (MUHNW, MUHO). For females, college education was most prevalent in the MHNW phenotype. In terms of physical activity, males in the MHNW group reported the highest levels of high-intensity activity (56.6%), followed by the MUHNW phenotype (55.3%) and the MHO phenotype (52.1%). Among females, the MHO phenotype exhibited the highest proportion of high-intensity activity (23.2%). Significant differences in physical activity levels were observed across cardiometabolic phenotypes in both genders (p < 0.001).

Table 1 Baseline, clinical, and biochemical characteristics stratified by cardiometabolic phenotype and gender among participants of the Azar cohort population

Anthropometric and clinical measures, including WC, FBG, and TG, progressively increased across cardiometabolic phenotypes from MHNW to MUHO, with the highest values observed in the MUHNW and MUHO phenotypes. The mean WC was highest in the MUHO phenotype (104.33 ± 8.13 cm for males and 99.18 ± 8.81 cm for females). FBG levels were significantly elevated in the MUHNW phenotype (120.40 ± 52.55 mg/dl for males and 123.80 ± 54.78 mg/dl for females) compared to the MHNW phenotype (88.13 ± 17.14 mg/dl for males and 88.03 ± 17.53 mg/dl for females). Among males, the highest TG levels were observed in the MUHO phenotype (232.43 ± 122.25 mg/dl), while for females, the highest TG levels were found in the MUHNW phenotype (193.79 ± 68.06 mg/dl). A significant difference in clinical and biochemical was observed among the cardiometabolic phenotypes in both genders (p < 0.05).

Figure 2A and B illustrate the mean AIP for cardiometabolic phenotypes in males and females. The MHNW group consistently exhibits the lowest mean AIP (0.05 for males; -0.05 for females), while the MUHNW phenotype has the highest (0.42 for males; 0.28 for females, P < 0.001). The MUHO phenotype shows slightly lower AIP values than the MUHNW phenotype, and the MHO phenotype demonstrates intermediate values, higher than the MHNW phenotype but lower than the MUHNW and MUHO phenotypes. In males (Fig. 2A), the MUHNW and MUHO phenotypes have significantly higher AIP values compared to the MHNW and MHO phenotypes. A similar trend is observed in females (Fig. 2B). Significant differences were noted among cardiometabolic phenotypes in both genders (P < 0.001).

Fig. 2

Comparison of the mean AIP among different cardiometabolic phenotypes in male participants. (*) Indicates a significant difference between two groups as determined by LSD post hoc analysis following one way ANOVA (p<0.001)

Table 2 shows the mean energy and macronutrient intake among the cardiometabolic phenotypes. The energy intake was highest in the MHNW group (2683.50 ± 653.82 kcal/day) and lowest in the MUHNW group (2473.92 ± 730.72 kcal/day). Metabolically healthy phenotypes (MHNW, MHO) exhibited higher energy intake and a higher percentage of energy intake from carbohydrates (62.09 ± 5.37 and 62.18 ± 5.17, respectively) compared to metabolically unhealthy phenotypes (MUHNW, MUHO) (61.75 ± 5.06 and 61.84 ± 5.19, respectively). There was a statistically significant difference in the mean energy intake and the percentage of energy intake from protein among the cardiometabolic phenotypes (p < 0.001), but the mean percentage of energy intake from carbohydrates and fat among the cardiometabolic phenotypes did not show a statistically significant difference.

Table 2 Dietary intake stratified by cardiometabolic phenotypes in the Azar cohort population (n: 9515)

As shown in Table 3, the correlation analysis between dietary intake and AIP across cardiometabolic phenotypes revealed varying weak correlations depending on the level of adjustment. In the unadjusted stage, weak but meaningful correlations were observed across cardiometabolic phenotypes for energy intake and macronutrients. After adjusting for MET and SES, weak but meaningful correlations were still observed between dietary intake and AIP, except for fat intake in the MUHNW phenotype. Energy and protein intake showed higher correlation coefficients in the MUHNW phenotype (r = 0.226, P = 0.02 and r = 0.237, P = 0.017, respectively). Following the final adjustment, which included MET, SES, age, and gender, most correlations were substantially attenuated or became non-meaningful, except for energy intake, carbohydrate, and protein in the MUHO phenotype retained weak but meaningful correlations (r = 0.048, P = 0.01; r = 0.057, P = 0.003; and r = 0.05, P = 0.01, respectively). Similarly, carbohydrate and lipid intake showed weak yet meaningful correlations in the MHO phenotype (r = 0.034, P = 0.01; r = -0.055, P < 0.001, respectively). Protein intake also retained a weak but meaningful correlation in the MHNW phenotype (r = 0.064, P = 0.005).

Table 3 Correaltion between diferrent cardiometabolic and their diatary intake with AIP

Discussion

This cross-sectional study investigated the association between energy intake, macronutrients, and AIP in different cardiometabolic phenotypes. Limited research and the narrow range of dietary factors analyzed have contributed to uncertainty regarding the nutritional determinants of cardiovascular risk. Comparing results across studies is challenging due to inconsistencies in defining cardiometabolic phenotypes and nutritional factors. Previous studies exploring dietary intake and AIP have often lacked stratification by cardiometabolic phenotypes and yielded conflicting results [30, 31]. Our findings complement existing research by providing novel insights into how dietary factors influence AIP across cardiometabolic phenotypes, enhancing our understanding of cardiometabolic health.

Our findings revealed significant differences in energy intake and the percentage of energy derived from protein among cardiometabolic phenotypes. These results are consistent with a study conducted in Iran, which also found significant differences in energy intake and the percentage of energy derived from protein among participants [15]. However, in contrast, Phillips et al. in Ireland indicated that total calorie intake and dietary macronutrient composition were similar between metabolically healthy and unhealthy phenotypes, regardless of BMI [32]. Similarly, Mirmiran et al. in Iran [33] and Kim and Song in Korea [34] reported no significant differences in energy intake or macronutrient distribution throughout cardiometabolic phenotypes.

In contrast to trends observed in other studies, our study found that energy intake and the percentage of energy from carbohydrates were lower in metabolically unhealthy phenotypes (MUHNW and MUHO) compared to metabolically healthy phenotypes (MHNW and MHO). For example, Abolnezhadian et al. [35] and Nikniaz et al. [36] reported higher energy intake in both metabolically healthy and unhealthy obese participants compared to their normal-weight counterparts when examining cardiometabolic phenotypes. To explain this discrepancy, we hypothesize that individuals with metabolically unhealthy phenotypes, often diagnosed with conditions such as type 2 diabetes or hypertension, may have been advised to reduce food intake and modify their lifestyle under medical supervision. This hypothesis aligns with previous findings suggesting that such individuals may adopt dietary patterns aimed at managing metabolic risk factors [37]. Interestingly, our study observed that metabolically healthy phenotypes (MHNW and MHO) with higher energy and carbohydrate intake were also associated with greater physical activity levels. While this may suggest a healthier lifestyle, we propose that the high energy and carbohydrate intake in MHO individuals could signal a potential transition toward an unhealthy metabolic state. Previous longitudinal studies have highlighted the transient nature of the MHO phenotype, with a significant proportion progressing to a metabolically unhealthy phenotype within a decade, accompanied by an elevated risk of cardiovascular diseases. Key contributors to this transition include low-grade inflammation and insulin resistance, particularly in individuals with poor dietary habits [38].

The findings of this study indicated that, after adjusting for confounding factors, correlations between energy intake and macronutrients with AIP across cardiometabolic phenotypes were non-significant, except for weak but meaningful correlations observed between energy, carbohydrate, and protein intake and AIP in the MUHO phenotype, as well as between carbohydrate and lipid intake and AIP in the MHO phenotype. These results suggest that dietary factors may have a differential impact on AIP depending on the cardiometabolic phenotypes. Therefore, it cannot be said that our study is consistent with previous studies. For instance, Shin et al. reported no significant differences in energy and macronutrient intake across AIP quartiles in males [14]. Similarly, Edwards et al. in US adults, as well as Bajerska et al., and Salehi et al., found no significant differences in dietary intake and AIP among female subjects [18, 19, 22]. The disparities between our study and previous research may arise from differences in study design, sample size, population characteristics, and AIP cut-off values. Many earlier studies often focused on a single gender or had smaller sample sizes, which limited their statistical power. In contrast, our study included both genders and a larger sample size, enabling us to detect potential correlations more effectively.

What sets our study apart is its emphasis on dietary intake within the framework of cardiometabolic phenotypes, rather than generalized categories like BMI or gender. By focusing on cardiometabolic phenotypes, we captured relationships between diet and AIP that may have been overlooked in studies prioritizing BMI alone. Additionally, the inconsistency in findings across studies could result from differences in adjusting for confounding factors. Our approach, which accounted for these factors, suggests that dietary influences on AIP may vary across cardiometabolic phenotypes. This underscores the importance of considering confounders in dietary and AIP research. Without proper adjustments, studies may fail to identify true associations, leading to misleading or incomplete conclusions.

Energy and macronutrient intake influence AIP through several interconnected mechanisms, including triglyceride metabolism, insulin resistance, and lipid dysregulation. For example, a high intake of saturated fatty acids (SFA) has been associated with increased triglyceride levels and reduced HDL-C, contributing to a higher AIP. However, some findings suggest that the type and source of dietary fats may modulate this relationship. Shin et al. observed that consumption of milk and dairy products was associated with a reduced risk of AIP [14], which may indicate a potential protective role of certain fat sources. Similarly, Moussavi et al. found that the quality of dietary fat, combined with overweight or obesity status, significantly correlated with AIP and lipid profiles, indicating the importance of fat fractions in determining AIP levels [20]. Hamułka et al. also found a positive correlation between dietary SFA intake and AIP in young, Caucasian, and overweight [39]. Collectively, these findings highlight that both the quantity and quality of dietary fat influence AIP, although further research is needed to elucidate the complex interactions among these factors. Dietary guidelines frequently emphasize reducing total or saturated fat intake over carbohydrates. However, low-fat diets may sometimes harm lipid profiles due to a seesaw effect between fat and carbohydrate intake [13, 40]. For instance, replacing saturated fats with carbohydrates has been shown to negatively affect lipid levels, exacerbating atherogenic dyslipidemia characterized by elevated triglycerides, small LDL-C particles, and reduced HDL cholesterol commonly associated with insulin resistance and obesity [41]. Similarly, short-term low-carbohydrate interventions demonstrated reductions in triglycerides and improvements in HDL-C, leading to a favorable impact on AIP [42]. Overconsumption of energy, regardless of macronutrient composition, contributes to fat storage, increased triglycerides, and reduced HDL-C [43], ultimately leading to a higher AIP.

The study’s findings suggest that dietary interventions should focus on improving lipid profiles by adjusting energy and macronutrient intake, which have significant implications for dietary recommendations and cardiometabolic health management. Also, these insights suggest that personalized dietary strategies based on individual metabolic health and habits are essential for managing cardiometabolic health effectively. Incorporating more targeted dietary recommendations within the existing framework of nutritional counseling could help reduce the progression of atherogenic dyslipidemia, especially in individuals with cardiometabolic phenotypes, insulin resistance, or adverse lipid profiles, who are at higher risk for cardiometabolic diseases. However, due to its cross-sectional design, the study cannot establish causal relationships, underscoring the necessity for future longitudinal research to clarify the impact of dietary intake on cardiovascular disease development.

Strengths and limitations

This study is the first to utilize a large sample size to identify the correlation between dietary intake and the AIP across cardiometabolic phenotypes, offering valuable insights into CVD risk factors. However, the cross-sectional design limits the ability to infer causal relationships. The small number of participants with a metabolically unhealthy normal weight phenotype restricts the analysis of dietary intake and AIP correlations in this subgroup. Reverse causality is also a possibility, as individuals with metabolically unhealthy phenotype may have altered their diets before the study. Additionally, the use of a semi-quantitative FFQ, while practical, may introduce recall errors, affecting the generalizability of the findings. The study’s results, based on a specific population in northwestern Iran, may not apply to other geographic groups. Future research should use a diverse, prospective design to validate these findings.

Conclusion

In this cross-sectional study, after adjusting for confounding factors, correlations between energy intake and macronutrients with AIP across cardiometabolic phenotypes were non-significant, except for weak but meaningful correlations observed between energy, carbohydrate, and protein intake and AIP in the MUHO phenotype, as well as between carbohydrate and lipid intake and AIP in the MHO phenotype. Future studies should examine dietary patterns in these phenotypes, considering prior lifestyle interventions, to confirm these findings. Moreover, prospective studies are needed to establish a causal relationship between dietary intake and AIP in cardiometabolic phenotypes and to explore the impact of specific macronutrient compositions on AIP while accounting for metabolic status variations. Targeted dietary interventions, such as optimizing macronutrient intake and incorporating functional foods, could improve AIP in metabolically unhealthy phenotypes. Such insights may contribute to the development of personalized nutrition strategies aimed at reducing cardiovascular risk across different cardiometabolic phenotypes.