Abstract
The Dietary Index for Gut Microbiota (DI-GM) is a novel metric assessing diet quality in relation to gut microbial composition. However, its association with all-cause mortality (ACM) in coronary heart disease (CHD) patients remains unexplored. This study aimed to evaluate the association between DI-GM and ACM among the CHD population. This retrospective cohort study utilized 2005–2018 National Health and Nutrition Examination Survey data. DI-GM scores were calculated based on 24-hour dietary recall data. The association between DI-GM and ACM in CHD patients was assessed using weighted Cox proportional hazards models and restricted cubic spline analyses. Subgroup analyses and interaction tests were performed to examine the robustness of the findings and potential effect modifiers. A total of 1537 CHD patients were included, representing 81,24,166 US adults. The overall ACM rate was 37.41%. Higher DI-GM scores were significantly associated with lower ACM risk. Compared with participants with scores of 0 to 3, those with scores ≥6 and 5 had significantly reduced ACM risk, with hazard ratios of 0.593 (95% confidence interval: 0.413–0.851; P = .005) and 0.649 (95% confidence interval: 0.432–0.975; P = .037), corresponding to 40.7% and 35.1% reductions in ACM risk. Restricted cubic spline analysis revealed a nonlinear inverse association between DI-GM and ACM. Subgroup analyses indicated that diabetes status may modify this association (P for interaction = .046). Higher DI-GM scores were nonlinearly and inversely associated with ACM among CHD patients. This association may be modified by diabetes status.
Keywords: all-cause mortality, cohort study, coronary heart disease, Dietary Index for Gut Microbiota, National Health and Nutrition Examination Survey
1. Introduction
Coronary heart disease (CHD) is a prevalent cardiovascular condition associated with substantial morbidity and mortality.[1,2] In 2015, of the 400 million cardiovascular disease (CVD) cases globally, 111 million (27%) were attributable to CHD.[3] The World Health Organization (WHO) projects CHD among the top 3 causes of mortality worldwide by 2030, with annual deaths reaching 9.3 million.[1] Its etiology is multifactorial, involving non-modifiable and modifiable risk factors. Non-modifiable factors include age, genetic predisposition, and sex, whereas modifiable factors encompass diabetes mellitus, dyslipidemia, hypertension, overweight or obesity, and smoking.[4-6] Increasing evidence underscores the critical roles of physical inactivity and unhealthy dietary patterns in CHD-related mortality.[7-9] These modifiable factors are particularly crucial, as preventive strategies can be effectively directed toward them. Therefore, understanding factors influencing CHD-related mortality is essential to inform targeted interventions and lower disease burden.
Attention has increasingly been directed to the role of gut microbiota in CHD pathogenesis.[10-13] The gut microbiota regulates lipid metabolism, inflammatory responses, and vascular endothelial function through bioactive metabolites like short-chain fatty acids (SCFAs) and trimethylamine-N-oxide (TMAO). These mechanisms may thereby influence CHD development and progression.[14] Diet is a key determinant of gut microbiota composition and function. For instance, the Mediterranean diet is cardioprotective by reducing blood pressure and insulin resistance, improving lipid profiles, and attenuating systemic inflammation.[15,16] Notably, its cardiometabolic benefits are substantially mediated by gut microbiota activity.[17,18] These findings suggest gut microbiota as a crucial mechanistic link between dietary patterns and CHD outcomes.
The Dietary Index for Gut Microbiota (DI-GM) is a new metric quantifying the impact of diet on gut microbiota composition and function. It integrates multiple dietary components, classifying them as beneficial or adverse to intestinal microbial health. The score is calculated based on the intake of 14 food items or nutrients. Beneficial components are scored 1 if intake meets or exceeds the sex-specific median. Otherwise, 0 was assigned. Adverse components are scored 0 if intake is at or above the median or if fat-derived energy exceeds 40% of total energy. Otherwise, 1 was assigned. The total score is 0 to 14, including up to 10 for beneficial components and 4 for adverse components. Despite its potential utility, the association between DI-GM and all-cause mortality (ACM) in CHD patients remains unclear. A significant association between DI-GM and ACM in CHD patients was hypothesized. Higher DI-GM scores, reflecting a diet more favorable to gut microbiota, may be associated with a lower ACM risk in this high-risk population. Therefore, this study investigated the relation of DI-GM to ACM using 2005–2018 National Health and Nutrition Examination Survey (NHANES) data. The findings may provide novel evidence to inform dietary strategies and improve prognosis in CHD patients.
2. Materials and methods
2.1. Data source
NHANES is a multi-cycle, biennial, cross sectional survey on nationally representative Americans of the National Center for Health Statistics (NCHS).[19] Its protocols gained the approval of the National Center for Health Statistics Research Ethics Review Board (Protocol #2005-06, #2011-17, and #2018-01). Every participant gave informed consent in writing. Therefore, ethical approval was unnecessary. Ethical oversight and consent procedures were detailed at https://wwwn.cdc.gov/nchs/nhanes/Default.aspx.
2.2. Population selection
DI-GM data from 7 NHANES cycles (2005–2018) were analyzed. A total of 70,190 participants were initially identified. Participants with missing mortality data, DI-GM scores, CHD, or covariate information (demographic characteristics, smoking and drinking status, and medical history) were excluded. Ultimately, 1537 participants were analyzed. The selection process is provided in Figure 1.
Figure 1.
Flowchart of study participant selection.
2.3. DI-GM assessment
DI-GM scores were derived from 2005 to 2018 24-hour dietary recall data based on 14 beneficial or adverse dietary components.[20,21] Beneficial components included avocado, broccoli, chickpeas, coffee, cranberries, fermented dairy products, dietary fiber, green tea, soybeans, and whole grains. Adverse components encompassed refined grains, red meat, processed meat, and a high-fat diet (≥40% of total energy intake from fat). For each beneficial component, 1 was assigned if intake met or exceeded the sex-specific median, and 0 otherwise. For adverse components, 0 was assigned if intake met or exceeded the median (or ≥40% for fat), and 1 otherwise. The total DI-GM score was 0 to 14, with beneficial components contributing up to 10 and adverse components up to 4. Participants were categorized into 4 groups based on DI-GM scores.[21]
2.4. Definition of CHD
The medical condition section of the NHANES questionnaire is identified by the prefix “MCQ.” It contains self- and proxy-reported data on a variety of health conditions and medical histories for participants aged 20 and older. CHD status was determined by the question, “Has a doctor or other health professional told you that you had CHD?” Responses to this item, labeled MCQ160c in the household interview questionnaire, were used to identify CHD participants. “Yes” meant having CHD.[22]
2.5. Covariates
Collected variables included demographic characteristics, comorbidities, and laboratory parameters: sex, age, race/ethnicity, education, marital status, poverty-to-income ratio (PIR), smoking, drinking, body mass index (BMI), hypertension, diabetes, hemoglobin A1c, fasting blood glucose (FBG), systolic blood pressure (SBP), diastolic blood pressure, lymphocyte count, monocyte count, neutrophil count, hemoglobin, platelet count (PC), albumin, and creatinine. Fasting venous blood samples were obtained and analyzed according to NHANES laboratory protocols. Fasting laboratory data were collected from participants aged ≥12 who had fasted for 8.5–24 hours. Diabetes was defined as self-reported physician diagnosis, hypoglycemic medication use, hemoglobin A1c ≥6.5%, or fasting blood glucose ≥7.0 mmol/L.[23] Hypertension was defined by a reported medical diagnosis or antihypertensive drug use. It was also defined by elevated measurements. Elevated blood pressure meant a mean systolic blood pressure ≥130 mm Hg or diastolic blood pressure ≥80 mm Hg across 3 consecutive readings.[24] BMI was calculated as weight in kilograms divided by height in meters squared (kg/m2). BMI was classified as normal weight (<25.0 kg/m2), overweight (25.0–29.9 kg/m2), or obese (≥30.0 kg/m2).[25] PIR was low (<1.30), middle (1.30–3.49), or high (≥3.50).[26] Continuous variables with over 20% missing data were excluded. Those with <20% missingness were imputed using multiple imputation with a random forest algorithm. Five imputed datasets were generated. Follow-up time was calculated from baseline interview to death or December 31, 2019, whichever occurred first.
2.6. Statistical analysis
All analyses followed guidelines from the Centers for Disease Control and Prevention. Given the complex, multistage, stratified sampling of NHANES, appropriate sampling weights were applied to ensure nationally representative estimates. Participants were divided into 4 groups by DI-GM scores. Trends across groups were assessed. Normality of continuous variables was evaluated using the Shapiro–Wilk test. All continuous variables were non-normally distributed. They are presented as medians with interquartile ranges. Group differences were assessed using the Wilcoxon rank-sum test. Categorical variables are expressed as frequencies and percentages. They were compared using weighted chi-square tests. Multicollinearity among covariates was assessed before modeling. Variables with a variance inflation factor ≥5 were excluded. The association between DI-GM and ACM in CHD participants was evaluated using weighted Cox proportional hazards models. The results are reported as hazard ratios and 95% confidence intervals. Three models were constructed: model 1 adjusted for age and sex; model 2 additionally adjusted for drinking, education, marital status, PIR, race, and smoking; model 3 further adjusted for BMI, lymphocyte count, monocyte count, neutrophil count, hypertension, diabetes, hemoglobin, platelet count, albumin, and creatinine. Potential nonlinear associations were assessed using restricted cubic spline (RCS) models with 4 knots. Subgroup analyses were conducted by age, sex, PIR, smoking, drinking, BMI category, hypertension, and diabetes. All analyses were enabled by R 4.3.2 (The R Foundation for Statistical Computing). Two-tailed P < .05 denoted statistical significance.
3. Results
3.1. Baseline characteristics
The weighted baseline characteristics are presented in Table 1. A total of 1537 CHD patients, representing 81,24,166 US adults, were included. Their median age was 68 (interquartile ranges, 61–77). Males accounted for 66.34% and females for 33.66%. The overall ACM rate was 37.41%.
Table 1.
Weighted baseline characteristics of the study population.
| Characteristic | N* | Overall, N = 81,24,166† | DI-GM = 0–3, N = 10,19,863† | DI-GM = 4, N = 15,98,658† | DI-GM = 5, N = 19,51,705† | DI-GM = 6–14, N = 35,53,941† | P-value‡ |
|---|---|---|---|---|---|---|---|
| Gender, n (%) | 1537 | .702 | |||||
| Male | 1072 (66.34%) | 158 (70.15%) | 209 (62.36%) | 265 (68.28%) | 440 (65.98%) | ||
| Female | 465 (33.66%) | 50 (29.85%) | 104 (37.64%) | 113 (31.72%) | 198 (34.02%) | ||
| Age (yr) | 1537 | 68.00 (61.00, 77.00) | 63.52 (55.70, 72.28) | 69.00 (61.00, 76.00) | 67.00 (60.00, 74.00) | 71.00 (64.00, 78.00) | <.001 |
| Race, n (%) | 1537 | .011 | |||||
| Mexican | 144 (3.70%) | 12 (3.31%) | 22 (3.39%) | 43 (3.67%) | 67 (3.97%) | ||
| Other Hispanic | 102 (2.73%) | 13 (1.79%) | 24 (3.42%) | 23 (3.71%) | 42 (2.16%) | ||
| Non-Hispanic White | 990 (80.86%) | 131 (77.37%) | 192 (79.79%) | 254 (84.14%) | 413 (80.53%) | ||
| Non-Hispanic Black | 200 (6.14%) | 39 (9.85%) | 64 (10.32%) | 38 (4.34%) | 59 (4.19%) | ||
| Others | 101 (6.57%) | 13 (7.69%) | 11 (3.08%) | 20 (4.15%) | 57 (9.15%) | ||
| Education level, n (%) | 1537 | .369 | |||||
| <9th grade | 228 (7.89%) | 37 (9.00%) | 43 (7.58%) | 61 (7.86%) | 87 (7.72%) | ||
| 9–11th grade | 244 (13.57%) | 36 (14.48%) | 51 (11.68%) | 63 (16.37%) | 94 (12.62%) | ||
| High school grad/GED or equivalent | 372 (27.14%) | 53 (32.32%) | 76 (28.17%) | 100 (29.12%) | 143 (24.11%) | ||
| Some college or AA degree | 408 (28.70%) | 56 (28.01%) | 93 (29.94%) | 98 (30.07%) | 161 (27.59%) | ||
| College graduate or above | 285 (22.70%) | 26 (16.19%) | 50 (22.63%) | 56 (16.58%) | 153 (27.96%) | ||
| Marital status, n (%) | 1537 | .282 | |||||
| Married | 907 (61.07%) | 125 (64.79%) | 166 (56.36%) | 236 (65.58%) | 380 (59.64%) | ||
| Widowed | 287 (17.37%) | 27 (14.37%) | 51 (13.71%) | 67 (14.38%) | 142 (21.52%) | ||
| Divorced | 188 (11.00%) | 31 (12.36%) | 49 (13.62%) | 42 (10.45%) | 66 (9.75%) | ||
| Separated | 39 (2.37%) | 6 (1.74%) | 13 (3.11%) | 8 (2.64%) | 12 (2.07%) | ||
| Never married | 78 (5.56%) | 13 (4.07%) | 25 (10.85%) | 11 (4.17%) | 29 (4.36%) | ||
| Living with partner | 38 (2.63%) | 6 (2.67%) | 9 (2.36%) | 14 (2.78%) | 9 (2.67%) | ||
| PIR | 1537 | 2.50 (1.43, 4.78) | 2.13 (1.20, 3.92) | 2.39 (1.32, 3.97) | 2.16 (1.39, 4.53) | 3.03 (1.52, 5.00) | .037 |
| Smoking, n (%) | 1537 | .726 | |||||
| Never | 552 (35.89%) | 64 (33.61%) | 109 (34.94%) | 126 (33.30%) | 253 (38.39%) | ||
| Current | 272 (19.25%) | 56 (22.06%) | 66 (19.96%) | 70 (22.89%) | 80 (16.13%) | ||
| Former | 713 (44.86%) | 88 (44.33%) | 138 (45.10%) | 182 (43.81%) | 305 (45.49%) | ||
| Drinking, n (%) | 1537 | .577 | |||||
| No | 399 (23.96%) | 48 (26.89%) | 74 (25.18%) | 111 (26.45%) | 166 (21.19%) | ||
| Yes | 1138 (76.04%) | 160 (73.11%) | 239 (74.82%) | 267 (73.55%) | 472 (78.81%) | ||
| Hypertension, n (%) | 1537 | .134 | |||||
| No | 258 (19.67%) | 28 (17.18%) | 45 (14.33%) | 62 (17.92%) | 123 (23.75%) | ||
| Yes | 1279 (80.33%) | 180 (82.82%) | 268 (85.67%) | 316 (82.08%) | 515 (76.25%) | ||
| Diabetes, n (%) | 1537 | .06 | |||||
| No | 870 (57.04%) | 104 (47.38%) | 171 (51.28%) | 213 (58.91%) | 382 (61.38%) | ||
| Yes | 667 (42.96%) | 104 (52.62%) | 142 (48.72%) | 165 (41.09%) | 256 (38.62%) | ||
| BMI (kg/m2) | 1537 | 29.52 (26.00, 33.80) | 30.32 (26.31, 34.87) | 31.10 (26.61, 34.79) | 29.88 (27.12, 34.19) | 28.30 (24.54, 32.29) | <.001 |
| Lymphocyte count (1000 cells/µL) | 1537 | 1.80 (1.40, 2.40) | 2.00 (1.50, 2.60) | 1.80 (1.40, 2.30) | 1.80 (1.50, 2.40) | 1.80 (1.40, 2.30) | .058 |
| Monocyte count (1000 cells/µL) | 1537 | 0.60 (0.50, 0.70) | 0.60 (0.50, 0.80) | 0.60 (0.50, 0.80) | 0.60 (0.50, 0.70) | 0.60 (0.50, 0.70) | .814 |
| Neutrophil count (1000 cells/µL) | 1537 | 4.40 (3.50, 5.50) | 4.50 (3.60, 6.00) | 4.40 (3.50, 5.40) | 4.30 (3.60, 5.50) | 4.40 (3.40, 5.43) | .38 |
| Hemoglobin,g/dL | 1537 | 14.30 (13.20, 15.10) | 14.60 (13.40, 15.20) | 14.20 (13.20, 14.80) | 14.30 (12.90, 15.10) | 14.30 (13.14, 15.10) | .445 |
| Platelet count (1000 cells/µL) | 1537 | 210.00 (174.00, 254.00) | 216.65 (183.00, 253.27) | 214.00 (184.53, 254.00) | 208.00 (176.00, 259.00) | 205.00 (168.00, 245.00) | .185 |
| Albumin (g/L) | 1537 | 41.41 (39.00, 44.00) | 41.36 (39.00, 44.00) | 41.00 (39.00, 44.00) | 41.00 (40.00, 43.00) | 42.00 (40.00, 44.00) | .194 |
| Creatinine (µmol/L) | 1537 | 89.28 (75.14, 107.85) | 88.40 (74.26, 106.73) | 91.05 (77.79, 110.83) | 90.17 (76.88, 110.95) | 88.40 (74.26, 105.49) | .183 |
| DI-GM | 1537 | 5.00 (4.00, 7.00) | 3.00 (2.00, 3.00) | 4.00 (4.00, 4.00) | 5.00 (5.00, 5.00) | 7.00 (6.00, 7.00) |
AA = Associate in Arts, BMI = body mass index, DI-GM = Dietary Index for Gut Microbiota, GED = General Educational Development, PIR = poverty–income ratio.
N not missing (unweighted).
n (unweighted) (%); median (25%,75%).
Chi-squared test with Rao & Scott’s second-order correction; Wilcoxon rank-sum test for complex survey samples.
3.2. Association of DI-GM with ACM in CHD
The association between DI-GM and ACM in CHD patients is summarized in Table 2. When DI-GM was analyzed as a continuous variable, a significant inverse association with ACM was observed in model 1. However, this association was attenuated after further adjustment and became insignificant in models 2 and 3.
Table 2.
The association between Dietary Index for Gut Microbiota and all-cause mortality of coronary heart disease.
| ACM | Model 1 | Model 2 | Model 3 | ||||
|---|---|---|---|---|---|---|---|
| HR (95% CI) | P | HR (95% CI) | P | HR (95% CI) | P | ||
| DI-GM | Continuous | 0.894 (0.806–0.991) | .033 | 0.905 (0.810–1.011) | .077 | 0.926 (0.830–1.033) | .167 |
| DI-GM group | 0–3 | – | – | – | – | – | – |
| 4 | 0.655 (0.433–0.991) | .045 | 0.635 (0.403–1.001) | .05 | 0.659 (0.428–1.015) | .059 | |
| 5 | 0.631 (0.448–0.888) | .008 | 0.573 (0.393–0.835) | .004 | 0.593 (0.413–0.851) | .005 | |
| ≥6 | 0.577 (0.397–0.838) | .004 | 0.575 (0.380–0.872) | .009 | 0.649 (0.432–0.975) | .037 | |
Model 1 was adjusted for age and gender. Model 2 was adjusted for age, gender, race, education level, marital status, PIR, drinking status and smoking status. Model 3 was adjusted for age, gender, race, education level, marital status, PIR, drinking status, smoking status, hypertension status, diabetes status, BMI, lymphocyte count, monocyte count,neutrophil count, hemoglobin, platelet count, albumin and creatinine.
ACM = all-cause mortality, BMI = body mass index, CI = confidence interval, DI-GM = Dietary Index for Gut Microbiota, HR = hazard ratio, PIR = poverty–income ratio.
In categorical analyses (0–3 as reference; 4, 5, and ≥6), DI-GM scores of 5 and ≥6 were significantly associated with reduced ACM risk in model 1. These inverse associations remained significant in model 2 and persisted in model 3.
3.3. RCS analysis
The association of DI-GM with ACM in CHD patients was examined via RCS curves (Fig. 2). After full adjustment, a nonlinear association of DI-GM with ACM of CHD was noted in our RCS analysis (P for nonlinear = .036).
Figure 2.
RCS analysis of the association between DI-GM and ACM in CHD. ACM = all-cause mortality, CHD = coronary heart disease, CI = confidence interval, DI-GM = Dietary Index for Gut Microbiota, HR = hazard ratio, RCS = restricted cubic spline.
3.4. Subgroup analysis
Subgroup analyses and interaction tests were conducted to assess the robustness of the findings (Fig. 3). Participants with DI-GM scores ≤5 served as the reference. The inverse association between DI-GM and ACM was generally consistent across subgroups by age, BMI, diabetes, drinking status, hypertension, PIR, sex, and smoking. Interactions were insignificant for age, BMI, drinking, hypertension, PIR, sex, and smoking (all P for interaction > .05). However, a significant interaction was identified for diabetes status (P for interaction = .046). Therefore, diabetes may modify the association between DI-GM and ACM in CHD patients.
Figure 3.
Subgroup analysis of the association between DI-GM and ACM in CHD. ACM = all-cause mortality, BMI = body mass index, CHD = coronary heart disease, CI = confidence interval, DI-GM = Dietary Index for Gut Microbiota, HR = hazard ratio, PIR = poverty-to-income ratio.
4. Discussion
This study demonstrated that higher DI-GM scores were independently nonlinearly associated with lower ACM among CHD patients. Subgroup and interaction analyses supported the robustness of these findings. Diabetes status appeared to modify the association. Dietary patterns conducive to gut microbiota health may contribute to improved outcomes in CHD. Optimizing dietary structure and addressing gut microbial dysbiosis may represent practical preventive strategies.
RCS analysis confirmed a significant nonlinear association of DI-GM with ACM. This pattern helps explain the differences between continuous and categorical models. The continuous model showed an attenuated and insignificant association after full adjustment. In contrast, categorical analysis revealed a stable and significant risk reduction in the high DI-GM group. The benefit appeared after a threshold level was reached. CHD patients with DI-GM scores ≥5 derived the clearest mortality benefit. Additional increases above this level yielded only modest gains, resulting in a plateau effect. One possible explanation is that once diet quality maintains microbial diversity and stability, further improvement offers limited additional benefit.[27] Prospective studies are needed to refine the optimal DI-GM threshold for clinical use.
Regarding the diet-gut microbiota interaction, DI-GM captures the effects of dietary fiber, protein, fat, and probiotics on the intestinal microbial ecosystem. Higher DI-GM scores reflect greater intake of fiber-rich foods, high-quality protein, and microbiota-supportive components. Several biological mechanisms underpin this relationship. Dietary fiber is fermented by gut microbiota into SCFAs, such as acetate, propionate, and butyrate. These metabolites are key energy sources for intestinal epithelial cells and are anti-inflammatory. They regulate lipid metabolism and improve vascular endothelial function. Butyrate suppresses inflammatory cytokine expression and mitigates atherosclerotic inflammation. Propionate modulates hepatic lipid metabolism and lowers serum triglyceride and low-density lipoprotein cholesterol (LDL-C) levels, thereby reducing CHD risk.[28-30] SCFAs reduce the translocation of harmful microbial products like lipopolysaccharide (LPS), which helps limit systemic inflammation.[31] Moreover, SCFAs activate peroxisome proliferator-activated receptors, key lipid metabolism and vascular inflammation regulators.[32] Probiotic-rich foods, including yogurt and fermented soy products, promote beneficial microbial growth and inhibit pathogenic species. Therefore, microbial homeostasis can be maintained.[33] A balanced microbiota reduces TMAO production. It is a pro-atherogenic metabolite derived from dietary precursors like choline and L-carnitine.[34] TMAO contributes to atherosclerosis by reducing the bioavailability of nitric oxide, an essential molecule for endothelial function. It causes endothelial dysfunction, which is an early hallmark of atherogenesis.[35,36] It also activates inflammatory signaling pathways such as mitogen-activated protein kinase and nuclear factor-kappa B. This leads to oxidative stress and inflammatory damage to the vascular endothelium, thereby facilitating plaque rupture.[37] Gut microbiota imbalance is implicated in CHD and other CVDs. In heart failure, dysbiosis increases intestinal permeability and systemic inflammation.[38] Microbiota-related gut–brain axis signaling influences blood pressure regulation via neural and inflammatory mechanisms.[39,40] Dysbiosis correlates with disordered inflammation, impaired lipid and glucose metabolism, and vascular dysfunction. This increases the risk of acute myocardial infarction.[41] Therefore, gut microbiota imbalance significantly influences the development and progression of CVDs, including CHD, heart failure, hypertension, and acute myocardial infarction. This occurs through effects on inflammation, metabolism, and vascular integrity. The mechanistic foundation possibly explains the inverse association of DI-GM with ACM. Our results align with prior studies emphasizing the cardioprotective effects of healthy dietary patterns. The Mediterranean diet, featuring high vegetable, fruit, whole grain, olive oil, and fish intake, correlates with favorable microbiota profiles and lower CVD risk.[42] Prospective studies have also demonstrated the relation of higher dietary fiber intake to lower cardiovascular mortality.[43] These observations align with our results. Higher DI-GM scores, reflecting healthier dietary patterns, were associated with reduced ACM among CHD patients.
Subgroup analyses revealed the association of diabetes with a markedly increased ACM risk. A significant interaction between diabetes and DI-GM was observed. Higher DI-GM scores were generally protective across most subgroups. However, this protective effect was greater in nondiabetic individuals and attenuated in diabetes patients. Therefore, diabetes modifies the association between DI-GM and ACM in CHD.
Diabetes is a well-established independent risk factor for CHD and its adverse outcomes. It elevates mortality through endothelial dysfunction, accelerated atherosclerosis, and coagulation abnormalities. Epidemiological studies report a 2- to 4-fold higher CHD and ACM risk among diabetic patients.[44,45] Emerging evidence highlights the central role of gut microbiota in diabetes and CHD. Diabetes-associated dysbiosis and reduced microbial adaptability diminish the beneficial effects of microbiota-friendly dietary patterns. This partially explains the attenuated protective effect of higher DI-GM scores in diabetic CHD patients. Therefore, diabetes-specific dietary strategies are needed when applying DI-GM in CHD patients to optimize subgroup outcomes.[46-48]
This study revealed a direct association of DI-GM with ACM in CHD patients. These findings address an important research gap and extend understanding of the prognostic value of gut microbiota-related dietary patterns in this high-risk population. Clinically, the results suggest an expanded framework for dietary management in CHD. Nutritional strategies should go beyond fat or sodium restriction and consider diet-gut microbiota interactions. Diets that improve DI-GM, including dietary fiber, probiotics, and high-quality protein, with reduced processed, high-sugar, and high-fat foods, should be encouraged. For public health, improving dietary quality and restoring gut microbial balance help reduce CHD-related morbidity and mortality. There are some limitations. First, the retrospective design introduces potential bias and prevents causal inference. Second, diabetes and hypertension diagnoses were based on self-reports, which may cause misclassification bias. Dietary intake was assessed through a single 24-hour recall. This potentially introduced measurement bias. Third, the study population was American and mainly comprised non-Hispanic White individuals, which may limit generalizability. Fourth, despite adjustment for multiple confounders, residual confounding cannot be excluded. Factors like occupational status, unmeasured lifestyle behaviors (detailed dietary patterns and physical activity intensity), genetic predisposition, concurrent medication use (antibiotics or proton pump inhibitors), and psychological stress may have influenced the observed associations. Future large-scale prospective studies are warranted to validate these findings. Integrating multi-omics approaches, including gut microbiome sequencing and metabolomics, may further clarify the biological mechanisms linking DI-GM to ACM in diverse CHD populations.
5. Conclusion
A diet with a higher DI-GM score, beneficial to microbiota flora health, was associated with lower ACM in CHD patients. DI-GM-increasing diets provide a practical direction for nutritional management in CHD. Population-level dietary pattern improvement and gut microbiota imbalance correction may be effective for lowering ACM risk in CHD patients.
Author contributions
Conceptualization: Fen Cao, Zhongwu Bao.
Formal analysis: Fen Cao, Jun Liu.
Funding acquisition: Fen Cao, Yongzhi Zhu.
Investigation: Fen Cao, Jun Liu.
Methodology: Gui Zhang, Sheng Zhang, Feng Hou, Zhiqiang Liu.
Resources: Yang Tian, Wei Zhang, Ping Xiao.
Supervision: Ping Xiao.
Writing – original draft: Fen Cao.
Writing – review & editing: Junjun Jiang, Yongzhi Zhu, Kun Wu.
Abbreviations:
- ACM
- all-cause mortality
- BMI
- body mass index
- CHD
- coronary heart disease
- CVD
- cardiovascular disease
- DI-GM
- Dietary Index for Gut Microbiota
- NHANES
- National Health and Nutrition Examination Survey
- PIR
- poverty-to-income ratio
- RCS
- restricted cubic spline
- SCFA
- short-chain fatty acid
- TMAO
- trimethylamine-N-oxide.
This research was funded by the Youth Research Fund Project of Hunan University of Medicine General Hospital (Grant No. QNJJ202507) and the Health Appropriate Technology Promotion Project of Hunan Provincial Health Commission in 2023 (Grant No. 202319018835).
Ethical review and approval were waived, as this study utilized publicly available, de-identified NHANES data. Because these data are anonymized and collected with informed consent by the NCHS, additional ethical approval was not required for secondary analysis.
The authors have no conflicts of interest to disclose.
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
How to cite this article: Cao F, Zhang G, Liu J, Tian Y, Zhang W, Zhang S, Hou F, Bao Z, Liu Z, Xiao P, Jiang J, Zhu Y, Wu K. Association between the Dietary Index for Gut Microbiota and all-cause mortality in coronary heart disease: A retrospective cohort analysis of NHANES (2005–2018). Medicine 2026;105:27(e49532).
PX, JJ, YZ, and KW contributed to this article equally.
Contributor Information
Fen Cao, Email: caofen742342554@163.com.
Gui Zhang, Email: 1053830329@qq.com.
Jun Liu, Email: 330959314@qq.com.
Yang Tian, Email: 992593251@qq.com.
Wei Zhang, Email: 1053830329@qq.com.
Sheng Zhang, Email: 1053830329@qq.com.
Feng Hou, Email: 863144233@qq.com.
Zhongwu Bao, Email: Baozhongwu3805@qq.com.
Zhiqiang Liu, Email: 330959314@qq.com.
Ping Xiao, Email: 1012772017@qq.com.
Junjun Jiang, Email: 28110415@qq.com.
Yongzhi Zhu, Email: zhuyongzhi2811@163.com.
References
- [1].Xia X, Tian X, Xu Q, et al. Global trends and regional differences in mortality of cardiovascular disease and its impact on longevity, 1980-2021: Age-period-cohort analyses and life expectancy decomposition based on the Global Burden of Disease study 2021. Ageing Res Rev. 2025;103:102597. [DOI] [PubMed] [Google Scholar]
- [2].Ahmadi M, Lanphear B. The impact of clinical and population strategies on coronary heart disease mortality: an assessment of Rose’s big idea. BMC Public Health. 2022;22:14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Roth GA, Johnson C, Abajobir A, et al. Global, regional, and national burden of cardiovascular diseases for 10 causes, 1990 to 2015. J Am Coll Cardiol. 2017;70:1–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Francula-Zaninovic S, Nola IA. Management of measurable variable cardiovascular disease’ risk factors. Curr Cardiol Rev. 2018;14:153–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Gutiérrez-Cuevas J, Santos A, Armendariz-Borunda J. Pathophysiological molecular mechanisms of obesity: a link between MAFLD and NASH with cardiovascular diseases. Int J Mol Sci. 2021;22:11629. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Kase NG, Gretz Friedman E, Brodman M, Kang C, Gallagher EJ, LeRoith D. The midlife transition and the risk of cardiovascular disease and cancer. Part I: magnitude and mechanisms. Am J Obstet Gynecol. 2020;223:820–33. [DOI] [PubMed] [Google Scholar]
- [7].Sun L, Xu C, Zhang Z, Tang L, Liu X. Physical activity lowers all-cause and cardio-cerebrovascular mortality in adults with coronary heart disease. Int J Cardiol. 2024;410:132225. [DOI] [PubMed] [Google Scholar]
- [8].Wang E, Fang C, Zhang J, Wang Y. Association between dietary inflammatory index and all-cause mortality risk in adults with coronary heart disease in the United States. Sci Rep. 2024;14:23998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Niu PP, Zhang C, Li S, Li YS. Association between egg consumption and mortality risk in United States adults with established coronary heart disease or stroke: a cohort study using data from NHANES 1999-2018. Arch Gerontol Geriatr. 2024;124:105475. [DOI] [PubMed] [Google Scholar]
- [10].Hills RD, Jr, Pontefract BA, Mishcon HR, Black CA, Sutton SC, Theberge CR. Gut Microbiome: profound implications for diet and disease. Nutrients. 2019;11:1613. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Bijla M, Saini SK, Pathak AK, et al. Microbiome interactions with different risk factors in the development of myocardial infarction. Exp Gerontol. 2024;189:112409. [DOI] [PubMed] [Google Scholar]
- [12].An L, Wirth U, Koch D, et al. The role of gut-derived lipopolysaccharides and the intestinal barrier in fatty liver diseases. J Gastrointest Surg. 2022;26:671–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Han Y, Gong Z, Sun G, et al. Dysbiosis of gut microbiota in patients with acute myocardial infarction. Front Microbiol. 2021;12:680101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Fiori J, Turroni S, Candela M, Brigidi P, Gotti R. Simultaneous HS-SPME GC-MS determination of short-chain fatty acids, trimethylamine, and trimethylamine N-oxide for gut microbiota metabolic profile. Talanta. 2018;189:573–8. [DOI] [PubMed] [Google Scholar]
- [15].Franchini C, Biasini B, Sogari G, et al. Adherence to the Mediterranean diet and its association with sustainable dietary behaviors, sociodemographic factors, and lifestyle: a cross-sectional study in US University students. Nutr J. 2024;23:56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Silva AR, Moraes BPT, Gonçalves-de-Albuquerque CF. Mediterranean diet: lipids, inflammation, and malaria infection. Int J Mol Sci. 2020;21:4489. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Wang DD, Nguyen LH, Li Y, et al. The gut microbiome modulates the protective association between a Mediterranean diet and cardiometabolic disease risk. Nat Med. 2021;27:333–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Rinott E, Meir AY, Tsaban G, et al. The effects of the Green-Mediterranean diet on cardiometabolic health are linked to gut microbiome modifications: a randomized controlled trial. Genome Med. 2022;14:29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Chen TC, Clark J, Riddles MK, Mohadjer LK, Fakhouri THI. National health and nutrition examination survey, 2015-2018: sample design and estimation procedures. Vital Health Stat 2. 2020:1–35. [PubMed] [Google Scholar]
- [20].Kase BE, Liese AD, Zhang J, Murphy EA, Zhao L, Steck SE. The development and evaluation of a literature-based dietary index for gut microbiota. Nutrients. 2024;16:1045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Zhang X, Yang Q, Huang J, Lin H, Luo N, Tang H. Association of the newly proposed dietary index for gut microbiota and depression: the mediation effect of phenotypic age and body mass index. Eur Arch Psychiatry Clin Neurosci. 2025;275:1037–48. [DOI] [PubMed] [Google Scholar]
- [22].Xu H, Xie P, Liu H, Tian Z, Zhang R, Cui M. The relationship between dietary inflammatory index in adults and coronary heart disease: from NHANES 1999-2018. Front Nutr. 2025;12:1564580. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Lawrence JM, Divers J, Isom S, et al. ; SEARCH for Diabetes in Youth Study Group. Trends in prevalence of type 1 and type 2 diabetes in children and adolescents in the US, 2001-2017. JAMA. 2021;326:717–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Ostchega Y, Fryar CD, Nwankwo T, Nguyen DT. Hypertension prevalence among adults aged 18 and over: United States, 2017-2018. NCHS Data Brief. 2020:1–8. [PubMed] [Google Scholar]
- [25].An S, Qin J, Gong X, et al. The mediating role of body mass index in the association between dietary index for gut microbiota and biological age: a study based on NHANES 2007-2018. Nutrients. 2024;16:4164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Zheng Y, Hou J, Guo S, Song J. The association between the dietary index for gut microbiota and metabolic dysfunction-associated fatty liver disease: a cross-sectional study. Diabetol Metab Syndr. 2025;17:17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Song W, Liu D, Xing Z, et al. Association of the newly proposed dietary index for gut microbiota and all-cause and cardiovascular mortality among individuals with diabetes and prediabetes. Front Nutr. 2025;12:1621277. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Feng Y, Xu D. Short-chain fatty acids are potential goalkeepers of atherosclerosis. Front Pharmacol. 2023;14:1271001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Ohira H, Tsutsui W, Fujioka Y. Are short chain fatty acids in gut microbiota defensive players for inflammation and atherosclerosis? J Atheroscler Thromb. 2017;24:660–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Wang Y, Dou W, Qian X, et al. Advancements in the study of short-chain fatty acids and their therapeutic effects on atherosclerosis. Life Sci. 2025;369:123528. [DOI] [PubMed] [Google Scholar]
- [31].Nogal A, Valdes AM, Menni C. The role of short-chain fatty acids in the interplay between gut microbiota and diet in cardio-metabolic health. Gut Microbes. 2021;13:1–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Christofides A, Konstantinidou E, Jani C, Boussiotis VA. The role of peroxisome proliferator-activated receptors (PPAR) in immune responses. Metabolism. 2021;114:154338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [33].González S, Fernández-Navarro T, Arboleya S, de Los Reyes-Gavilán CG, Salazar N, Gueimonde M. Fermented dairy foods: impact on intestinal microbiota and health-linked biomarkers. Front Microbiol. 2019;10:1046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [34].Canyelles M, Borràs C, Rotllan N, Tondo M, Escolà-Gil JC, Blanco-Vaca F. Gut microbiota-derived TMAO: a causal factor promoting atherosclerotic cardiovascular disease? Int J Mol Sci. 2023;24:1940. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].Caradonna E, Abate F, Schiano E, et al. Trimethylamine-N-Oxide (TMAO) as a rising-star metabolite: implications for human health. Metabolites. 2025;15:220. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [36].Gui Y, Zheng H, Cao RY. Foam cells in atherosclerosis: novel insights into its origins, consequences, and molecular mechanisms. Front Cardiovasc Med. 2022;9:845942. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [37].Seldin MM, Meng Y, Qi H, et al. Trimethylamine N-oxide promotes vascular inflammation through signaling of mitogen-activated protein kinase and nuclear factor-κB. J Am Heart Assoc. 2016;5:e002767. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [38].Tang WH, Kitai T, Hazen SL. Gut microbiota in cardiovascular health and disease. Circ Res. 2017;120:1183–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [39].Yang Z, Lin S, Liu Y, et al. Targeting intestinal microecology: potential intervention strategies of traditional Chinese medicine for managing hypertension. Front Pharmacol. 2023;14:1171119. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [40].Cai M, Lin L, Jiang F, et al. Gut microbiota changes in patients with hypertension: a systematic review and meta-analysis. J Clin Hypertens (Greenwich). 2023;25:1053–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [41].Novakovic M, Rout A, Kingsley T, et al. Role of gut microbiota in cardiovascular diseases. World J Cardiol. 2020;12:110–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [42].Itsiopoulos C, Mayr HL, Thomas CJ. The anti-inflammatory effects of a Mediterranean diet: a review. Curr Opin Clin Nutr Metab Care. 2022;25:415–22. [DOI] [PubMed] [Google Scholar]
- [43].Miao Z, Du W, Xiao C, et al. Gut microbiota signatures of long-term and short-term plant-based dietary patterns and cardiometabolic health: a prospective cohort study. BMC Med. 2022;20:204. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [44].Li Y, Luo J, Bao K, et al. Association of age at diagnosis of type 2 diabetes mellitus with the risks of the morbidity of cardiovascular disease, cancer, and all-cause mortality: evidence from a real-world study with a large population-based cohort study. Diabetes Res Clin Pract. 2024;217:111870. [DOI] [PubMed] [Google Scholar]
- [45].Aponte Ribero V, Efthimiou O, Abolhassani N, et al. Diabetes and the risk of cardiovascular events and all-cause mortality among older adults: an individual participant data analysis of five prospective studies. Eur J Clin Invest. 2025;55:e14340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [46].Tang N, Chen X, Li H, et al. Association of C-reactive protein, triglyceride, and glucose index with mortality in coronary heart disease and type 2 diabetes from NHANES data. Sci Rep. 2025;15:24687. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [47].Zeng Z, Qiu J, Chen Y, et al. Altered gut microbiota as a potential risk factor for coronary artery disease in diabetes: a two-sample bi-directional mendelian randomization study. Int J Med Sci. 2024;21:376–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [48].Wu W, Cheng J, Hou Z, et al. Association between the dietary index for gut microbiota and all-cause/cardiovascular mortality in patients with metabolic dysfunction-associated steatotic liver disease. Diabetol Metab Syndr. 2025;17:263. [DOI] [PMC free article] [PubMed] [Google Scholar]