Association between Helicobacter pylori infection and insulin resistance: Data from the US and Chinese cohorts

Abstract

Background

Helicobacter pylori infection has been associated with diverse extraintestinal morbidities, including cardiometabolic diseases. Insulin resistance is a key pathogenic factor involved in the development of various metabolic diseases. This study aimed to investigate the association between Helicobacter pylori infection and insulin resistance.

Methods

This cross-sectional study analyzed two independent cohorts: 2918 participants from the US National Health and Nutrition Examination Survey (1999–2000) and 912 inpatients from Southern China. Helicobacter pylori infection was determined using serum antibodies (National Health and Nutrition Examination Survey), ¹³C-urea breath test, or rapid urease test (China cohort). Insulin resistance was assessed using the triglyceride–glucose index or homeostasis model assessment of insulin resistance. Multivariate linear regression models were used to analyze the associations.

Results

In the National Health and Nutrition Examination Survey cohort, Helicobacter pylori positivity was initially associated with a higher triglyceride–glucose index (coefficient = 1.17, p = 0.001); however, this association lost statistical significance after full adjustment, and no significant association was observed with the homeostasis model assessment of insulin resistance index (coefficient = 0.03, p = 0.900). Similarly, in the Southern China cohort, no significant relationship was found between Helicobacter pylori infection status and the triglyceride–glucose index (coefficient = 0.06, p = 0.689).

Conclusions

Helicobacter pylori infection is not consistently associated with insulin resistance. Further studies are needed to clarify its role in metabolic diseases.

Keywords

Insulin resistance homeostasis model assessment of insulin resistance triglyceride–glucose metabolic diseases

Introduction

Helicobacter pylori, is one of the most prevalent pathogenic bacteria causing gastrointestinal infections and represents a major global public health concern, with a higher prevalence in developing countries.¹ H. pylori infection can cause several gastric diseases, including peptic ulcers, chronic active gastritis, gastric adenocarcinoma, and type B low-grade mucosa-associated lymphoid tissue lymphoma.² Recent studies have further demonstrated that H. pylori infection is associated with multiple nongastrointestinal tract morbidities, including metabolic syndrome, diabetes mellitus (DM), cardiovascular diseases, and autoimmune disorders.^3–7

Insulin resistance (IR) is a condition characterized by reduced responsiveness of insulin-targeted tissues to physiological insulin levels. IR is considered a pathogenic driver of numerous conditions, such as metabolic syndrome, atherosclerosis, DM, and nonalcoholic fatty liver disease.^8–11 The hyperinsulinemic–euglycemic clamp is the gold standard diagnostic method for diagnosing IR; however, its complexity and cost limit its widespread use in clinical practice.¹² The triglyceride–glucose (TyG) index and the homeostasis model assessment of IR (HOMA-IR) are novel, simple, accessible, and reproducible clinical alternative indices for studying IR.^13,14

However, the relationship between H. pylori infection and IR remains unclear. Some studies suggest that successful H. pylori eradication may improve insulin sensitivity and glycemic control, supporting a potential pathophysiological link.^15–17 Conversely, other studies have reported no significant association between H. pylori infection and IR.^18–20 This inconsistency may arise from racial differences, varying methodologies for IR assessment, and, crucially, unaccounted confounding. A major confounder is type 2 DM (T2DM), a condition whose pathogenesis is fundamentally driven by IR. The interplay is complex: T2DM is associated with an increased risk of H. pylori infection,^21–25 and antidiabetic medications may differentially influence this risk; for example, metformin use has been associated with a dose-dependent reduction in infection risk,²⁶ whereas the use of insulin and calcium channel blockers in patients with diabetes has been significantly associated with a higher incidence of H. pylori eradication.²⁷ These observations underscore the necessity of accounting for T2DM status and metabolic pharmacotherapy as potential confounders when examining the relationship between H. pylori and IR. Furthermore, to further investigate the downstream effects of IR, it is crucial to examine its key downstream manifestations. IR is a central driver of dyslipidemia, characterized primarily by elevated triglyceride (TG) levels. Therefore, examining the association between H. pylori and TG provides a direct link to the atherogenic lipid profile that contributes to the increased cardiovascular risk observed in infected individuals.

Hence, this study was designed to examine the association between H. pylori infection and IR across different ethnic populations while accounting for potential confounders as well as extend this investigation by assessing the association between H. pylori infection and TG levels as a direct marker of IR-related dyslipidemia. We used data from the 1999–2000 cycle of the US National Health and Nutrition Examination Survey (NHANES) and a cohort from Southern China.

Methods

Study cohorts

Two cross-sectional cohorts—one from the NHANES and the other from Southern China—were used for the analysis. For the NHANES cohort, we derived data from the 1999–2000 cycle, as H. pylori status was only assessed during this cycle. A total of 9965 adults were evaluated. Individuals meeting the following criteria were excluded: (a) missing serum H. pylori antibody data; (b) ambiguous H. pylori antibody results (i.e. H. pylori antibody titers between 0.91 and 1.09); and (c) missing fasting blood glucose and fasting triglyceride measurements (see Figure 1 for the sample selection flow diagram). All participants provided written informed consent. The NHANES was approved by the Research Ethics Review Board of the National Center for Health Statistics and the Centers for Disease Control and Prevention. Data on race, age, sex, poverty–income ratio, body mass index (BMI), household size, educational level, smoking and drinking history, and history of stroke and coronary heart disease (CHD) were obtained through interviews and physical examinations. Serum biochemical variables, including C-reactive protein (CRP), fasting glucose, low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), total cholesterol (TC), and fasting TG levels, were obtained from laboratory records. Information on DM and hypertension (HTN) was obtained through interviews and laboratory records.

Figure 1.

Flow chart of participant selection.

For the Southern China cohort, we enrolled patients hospitalized at the Cardiovascular Center of Shunde Hospital, Southern Medical University, between January 2019 and December 2021. This cohort was originally established to investigate the association between H. pylori infection and CHD. In summary, we enrolled adults aged ≥18 years who underwent H. pylori testing during hospitalization. Exclusion criteria included use of antibiotics or bismuth within 1 month before hospitalization, use of H2 receptor antagonists or proton pump inhibitors within 1 month before hospitalization, gastrointestinal hemorrhage, history of gastrectomy, hematological disease, severe renal or liver dysfunction, autoimmune disease, acute infection, cachexia, malignancy, or thyroid disease. Data on age, sex, BMI, heart rate, smoking and drinking history, history of DM and HTN, laboratory parameters, and carbon breath test outcomes were collected. Laboratory detection parameters, including TG, TC, and fasting glucose concentrations, were obtained from the medical record system.

Assessment of H. pylori infection status

In the NHANES cohort, H. pylori infection was determined by serum H. pylori IgG antibody titers measured using enzyme-linked immunosorbent assays. An antibody titer <0.90 indicated the absence of detectable IgG antibodies (negative), whereas a titer >1.10 indicated the presence of IgG antibodies (positive). Titers ranging from 0.91 to 1.09 were considered ambiguous and excluded from the analysis.

In the Southern China cohort, H. pylori infection was assessed using the ¹³C urea breath test or rapid urease testing during painless electrogastroscopy. The ¹³C urea breath test was performed as follows: an initial breath sample was collected after fasting, a ¹³C-labeled capsule was administered, a second respiratory sample was collected after 30 min, and both samples were assayed. During upper gastrointestinal endoscopy, antral biopsy specimens were obtained for rapid urease testing. H. pylori infection was considered present if the biopsy tissue caused the test strip dip to change from yellow to red within 1 min (strongly positive) or within 3 min (weakly positive). If no color change was observed, the rapid urease test was considered negative, and the patient was classified as negative for H. pylori infection.

Assessment of IR

We used two metrics, the TyG index and HOMA-IR index, to assess the IR status among participants in the NHANES cohort. The TyG index was calculated using the following formula: TyG = ln (fasting triglyceride concentration (mg/dL) × fasting glucose concentration (mg/dL)/2).²⁸ The HOMA-IR index was calculated using the following formula: HOMA-IR = (fasting glucose concentration (mmol/L) × fasting insulin concentration (μU/mL)/22.5).²⁹ In the Southern China cohort, only the TyG index was calculated because fasting insulin concentrations were not detected.

Statistical analysis

First, to evaluate potential selection bias resulting from the high rate of exclusions, we compared the baseline characteristics between participants included in the final analysis and those who were excluded. Subsequently, we described the baseline characteristics of the study participants from the NHANES database and the Southern China cohort, respectively. Participants were categorized into two groups based on H. pylori infection status (positive vs. negative). To account for the complex survey design and unanswered questions in the NHANES, appropriate sample weights were used in the current analysis. In the NHANES, continuous variables were summarized as means and SD, and categorical variables were classified as counts and proportions. We employed multivariable linear regression models to assess the associations between H. pylori infection and the outcome variables. The primary analysis aimed to examine the relationship between H. pylori infection and surrogate indices of IR, including the TyG index and HOMA-IR. As a complementary analysis, we further evaluated the association between H. pylori infection and TG levels to preliminarily investigate a key downstream manifestation of IR. For all regression analyses, we applied an identical sequential modeling strategy to ensure comparability using three models: (a) Model 1: unadjusted; (b) Model 2: adjusted for key clinical characteristics (age, sex, race, and BMI); and (c) Model 3: further adjusted for lifestyle and clinical covariates (smoking status, drinking status, CRP, TC, and history of DM, HTN, stroke, and CHD). TG levels were adjusted for only in the models where TyG or HOMA-IR served as the outcome, but not in the models where TG itself served as the outcome variable. Furthermore, due to substantial missing data precluding direct adjustment for antidiabetic and antihypertensive medications, we addressed potential confounding by these agents through stratified analyses based on T2DM status.

All statistical analyses were performed using R software version 4.3.1 (http://www.R-project.org). Statistical significance was defined as a two-tailed p-value <0.05.

Results

Comparison of baseline characteristics between included and excluded US participants

To assess potential selection bias resulting from the exclusion of a substantial number of participants due to data unavailability, we compared the baseline characteristics between included and excluded populations from the original NHANES cohort (Table S1). The results indicated that, compared to the excluded group, the included participants were older and had better socioeconomic status and higher educational levels. Critically, the included population exhibited a greater cardiometabolic disease burden. The prevalence rates of DM, HTN, stroke, and CHD were all markedly higher in the included group. Consistently, the use of insulin, oral hypoglycemic agents, and antihypertensive drugs was substantially higher in the included population. Regarding biochemical markers, the included group exhibited higher TyG index values and HDL-C concentrations. In summary, the final analytical sample represents an older population with more complex health profiles and a higher cardiometabolic risk burden.

Baseline characteristics of US participants

The final sample of the NHANES cohort included 2918 participants, representing a weighted population of 216,872,581 individuals. The mean age of all participants was 38 years, and 49% were male. The prevalence of H. pylori infection differed significantly across racial groups. Participants with positive H. pylori serology were older and had higher TG concentrations, CRP levels, TyG indices, and HOMA-IR indices as well as lower levels of education, income, and HDL-C concentrations than those with negative H. pylori serology. Moreover, positive serology was significantly associated with a greater burden of comorbidities, specifically higher prevalence of DM, HTN, stroke, and CHD (all p < 0.001). There were no significant differences between the two groups with regard to sex, BMI, household size, smoking and drinking history, or fasting LDL-C and TC levels (Table 1).

Table 1.

Demographic and clinical data of H. pylori–positive and –negative groups (NHANES 1999–2000).

Characteristic	Overall(n = 2918)	Negative(n = 1817)	Positive(n = 1101)	p-value
Age (years)	38.0 (18.3)	37.0 (17.9)	44.0 (18.1)	<0.001
Male, n (%)	1418 (49)	854 (48)	564 (50)	0.325
Race, n (%)				<0.001
Non-Hispanic White	1083.0 (37.1)	877.0 (48.3)	206.0 (18.7)
Mexican American	954.0 (32.7)	453.0 (24.9)	501.0 (45.5)
Non-Hispanic Black	620.0 (21.2)	349.0 (19.2)	271.0 (24.6)
Other Hispanic	174.0 (6.0)	84.0 (4.6)	90.0 (8.2)
Other Race	87.0 (3.0)	54.0 (2.9)	33.0 (2.9)
Poverty–income ratio	2.93 (1.57)	3.18 (1.56)	2.21 (1.51)	<0.001
BMI (kg/m²)	26 (5.13)	26 (5.93)	26 (5.19)	0.039
Household, n (%)				0.527
Large (four or more members)	1399 (37)	866 (37)	533 (38)
Small (1–3 members)	1519 (63)	951 (63)	568 (62)
Educational level, n (%)				<0.001
Less than high school	1572 (32)	867 (27)	705 (45)
High school diploma	521 (24)	350 (25)	171 (21)
More than high school	825 (44)	600 (48)	225 (34)
Smoke, n (%)	1186 (48)	701 (47)	485 (53)	0.110
Drink, n (%)	1981 (72)	1263 (73)	718 (69)	0.110
CRP (mg/dL)	0.19 (0.67)	0.18 (0.68)	0.21 (0.63)	0.023
Fasting glucose (mg/dL)	93 (28)	93 (20)	95 (41)	0.005
TG (mg/dL)	112 (96)	108 (95)	121 (100)	0.005
TC (mg/dL)	194 (42)	192 (41)	198 (43)	0.181
HDL-C (mg/dL)	47 (14)	48 (14)	45 (14)	0.026
LDL-C (mg/dL)	117 (36)	116 (36)	120 (37)	0.106
DM, n (%)	463 (15.8)	235 (12.9)	228 (20.7)	<0.001
HTN, n (%)	1287 (44.1)	650 (35.8)	637 (57.8)	<0.001
Stroke, n (%)	112 (3.8)	45 (2.5)	67 (6.1)	<0.001
CHD, n (%)	184 (6.3)	75 (4.1)	109 (9.9)	<0.001
TyG index	8.57 (0.64)	8.53 (0.62)	8.66 (0.67)	0.002
HOMA-IR index	2.28 (2.90)	2.23 (2.88)	2.45 (2.94)	0.029

Values are the mean (SD) and count (proportion) for continuous and categorical variables, respectively, Wilcoxon rank-sum test for complex survey samples, and chi-square test with Rao and Scott's second-order correction.

BMI: body mass index; CHD; coronary heart disease; CRP: C-reactive protein; DM: diabetes mellitus; HDL-C: high-density lipoprotein cholesterol; HOMA-IR: homeostasis model assessment of insulin resistance; HTN: hypertension; LDL-C: low-density lipoprotein cholesterol; NHANES: National Health and Nutrition Examination Survey; TC: total cholesterol; TG: triglycerides; TyG: triglyceride–glucose.

Association between H. pylori infection and TyG/HOMA-IR in US participants

The association between H. pylori infection status and the TyG index was evaluated in the NHANES 1999–2000 cohort (Table 2). The results indicated that H. pylori seropositivity was associated with a higher TyG index (β = 1.17, 95% confidence interval (CI): 1.07–1.27). After adjustment for age, sex, race, and BMI (Model 2) and additional adjustment for smoking and drinking history, CRP, TC concentration, TG concentration, and cardiometabolic comorbidities including DM, HTN, stroke, and CHD (Model 3), the association was substantially attenuated and was no longer statistically significant (β = 0.02, 95% CI: 0.00–0.05).

Table 2.

Association between H. pylori infection and TyG in US participants

	Model 1			Model 2			Model 3
Characteristic	Beta	95% CI¹	p-value	Beta	95% CI¹	p-value	Beta	95% CI¹	p-value
H. pylori = positive	1.17	1.07, 1.27	0.001	0.09	0.01, 0.17	0.031	0.02	0.00, 0.05	0.101
Race
Other Hispanic				−0.09	−0.22, 0.05	0.200	−0.02	−0.07, 0.03	0.401
Non-Hispanic White				−0.03	−0.18, 0.11	0.600	−0.03	−0.06, 0.00	0.021
Non-Hispanic Black				−0.41	−0.51, −0.31	<0.001	−0.13	−0.16, −0.10	<0.001
Other race				−0.11	−0.29, 0.07	0.200	−0.03	−0.09, 0.04	0.400
Age				0.01	0.01, 0.01	<0.001	0.01	0.00, 0.01	<0.001
Sex = male				0.13	0.04, 0.22	0.012	0.04	0.02, 0.07	<0.001
BMI				0.03	0.03, 0.04	<0.001	0.01	0.01, 0.01	<0.001
Smoke = yes							0.04	0.02, 0.06	<0.001
Drink = yes							−0.03	−0.06, −0.01	0.010
CRP							0.02	0.00, 0.03	0.036
TC							0.01	0.00, 0.00	<0.001
TG							0.01	0.00, 0.01	<0.001
DM = yes							0.01	−0.03, 0.03	0.802
HTN = yes							0.03	0.00, 0.05	0.029
Stroke = yes							−0.02	−0.08, 0.03	0.403
CHD = yes							−0.01	−0.06, 0.03	0.600

Model 1: Unadjusted; Model 2: Adjusted for age, sex, race, and BMI; Model 3: Adjusted for model 2 + smoke and drink history, CRP level, TC level, triglyceride level, DM, hypertension, stroke, and CHD.

95% CI¹: confidence Interval.

BMI: body mass index; CHD: coronary heart disease; CRP: C-reactive protein; DM: diabetes mellitus; HTN: hypertension; TC: total cholesterol; TyG: triglyceride–glucose.

The association between H. pylori infection status and the HOMA-IR index was also examined. Nevertheless, no significant association was observed between H. pylori infection and HOMA-IR index in any of the statistical models (Table 3).

Table 3.

Association between H. pylori infection and HOMA-IR in US participants.

	Model 1			Model 2			Model 3
Characteristic	Beta	95% CI¹	p-value	Beta	95% CI¹	p-value	Beta	95% CI¹	p-value
H. pylori = positive	1.27	0.86, 1.87	0.200	0.03	−0.37, 0.42	0.900	0.03	−0.33, 0.39	0.900
Race
Other Hispanic				0.17	−0.93, 1.3	0.700	−0.10	−0.80, 0.59	0.803
Non-Hispanic White				−0.14	−0.70, 0.42	0.600	−0.46	−0.88, −0.05	0.028
Non-Hispanic Black				−0.22	−0.62, 0.18	0.200	0.42	−0.02, 0.87	0.061
Other Race				−0.18	−0.66, 0.30	0.400	0.42	−0.52, 1.4	0.400
Age				0.00	−0.01, 0.02	0.400	0.01	0.00, 0.02	0.006
Sex = male				0.39	0.09, 0.69	0.020	0.31	−0.04, 0.65	0.085
BMI				0.23	0.17, 0.29	<0.001	0.20	0.17, 0.22	<0.001
Smoke = yes							0.08	−0.26, 0.43	0.602
Drink = yes							−0.49	−0.87, −0.12	0.010
CRP							0.19	−0.05, 0.42	0.121
TC							−0.01	−0.02, −0.01	<0.001
TG							0.01	0.01, 0.01	<0.001
DM = yes							0.32	−0.13, 0.77	0.201
HTN = yes							0.01	−0.32, 0.34	>0.9
Stroke = yes							−0.02	−0.85, 0.81	>0.9
CHD = yes							−0.76	−1.4, −0.10	0.054

BMI: body mass index; CHD: coronary heart disease; CRP: C-reactive protein; DM: diabetes mellitus; HOMA-IR: homeostasis model assessment of insulin resistance; HTN: hypertension; TC: total cholesterol.

Association between H. pylori infection and TyG in Southern China

Overall, 912 inpatients who underwent H. pylori testing were enrolled in the Southern China cohort. The average age was 58.4 years, and 51.6% of the participants were male. A total of 603 participants were H. pylori–positive (66.1%). Individuals in the H. pylori–positive group were more likely to be female and had higher BMI, TG concentrations, and fasting glucose concentrations. There were no statistically significant differences between the two groups in terms of heart rate, TC concentrations, history of DM, and HTN (Table 4). In the unadjusted model, H. pylori positivity showed a borderline association with the TyG index (β = 0.55, p = 0.056). This association became nonsignificant after sequential adjustment for covariates (Table 5).

Table 4.

Demographic and clinical data of H. pylori–positive and –negative groups in Southern China.

Characteristic	Overall(n = 912)	Negative(n = 309)	Positive(n = 603)	p-value
Age (years)	58.44 (10.87)	57.93 (11.16)	58.70 (10.72)	0.315
Male, n (%)	471.0 (51.6%)	142.0 (46.0%)	329.0 (54.6%)	0.017
BMI (kg/m²)	24.48 (3.40)	24.14 (3.17)	24.65 (3.50)	0.028
Smoke, n (%)	187.0 (20.5%)	55.0 (17.8%)	132.0 (21.9%)	0.173
Drink, n (%)	31.0 (3.4%)	10.0 (3.2%)	21.0 (3.5%)	0.999
Heart rate (bpm)	74.71 (11.31)	74.26 (10.62)	74.95 (11.65)	0.374
HTN, n (%)	463.0 (50.8%)	151.0 (48.9%)	312.0 (51.7%)	0.452
DM, n (%)	153.0 (16.8%)	50.0 (16.2%)	103.0 (17.1%)	0.802
TC (mmol/L)	4.98 (1.23)	4.89 (1.14)	5.02 (1.27)	0.112
TG (mmol/L)	1.56 (1.02)	1.47 (0.97)	1.61 (1.05)	0.047
Fasting glucose (mmol/L)	5.80 (1.75)	5.64 (1.42)	5.88 (1.89)	0.030
TyG index	4.66 (4.14)	4.29 (4.11)	4.84 (4.14)	0.055

Continuous variables were presented as mean ± SD or median and interquartile range, and categorical variables were expressed as numbers and percentages.

BMI: body mass index; DM: diabetes mellitus; HTN: hypertension; TC: total cholesterol; TG: triglycerides; TyG: triglyceride–glucose.

Table 5.

Association between H. pylori infection and TyG in Chinese participants

	Model 1			Model 2			Model 3
Characteristic	Beta	95% CI¹	p-value	Beta	95% CI¹	p-value	Beta	95% CI¹	p-value
H. pylori = positive	0.55	−0.01, 1.12	0.056	0.50	−0.07, 1.07	0.086	0.06	−0.23, 0.35	0.689
Age				−0.01	−0.04, 0.01	0.399	0.01	−0.00, 0.02	0.071
Sex = male				0.76	0.21, 1.30	0.006	−0.03	−0.36, 0.29	0.851
BMI							0.02	−0.02, 0.07	0.259
Smoke = yes							−0.14	−0.53, 0.26	0.492
Drink = yes							1.02	0.23, 1.80	0.011
TC							0.01	−0.11, 0.13	0.827
TG							3.45	3.30, 3.60	<0.001

Model 1: Unadjusted; Model 2: Adjusted for age and sex; Model 3: Adjusted for model 2 + smoke and drink history, BMI, TC, triglyceride. 95% CI¹: confidence Interval.

BMI: body mass index; TC: total cholesterol; TG: triglycerides; TyG: triglyceride–glucose.

Association between H. pylori infection and TG

Analysis of the NHANES cohort revealed that H. pylori positivity was associated with significantly higher TG levels in the unadjusted model. However, after sequential adjustment for demographic, clinical, and lifestyle factors in Models 2 and 3, the association was substantially attenuated and no longer statistically significant (Table S2). In Chinese participants, H. pylori positivity showed a borderline association with TG levels (β = 0.14, p = 0.053), which was attenuated to nonsignificance after covariate adjustment (β = 0.04, p = 0.518) (Table S3).

Association between H. pylori infection and TyG/HOMA-IR in participants with T2DM

In stratified analyses conducted to address potential confounding by antidiabetic and antihypertensive medications, associations between H. pylori infection and IR indices were examined among participants with T2DM. Within the US T2DM subgroup, H. pylori positivity was initially associated with a higher HOMA-IR index in the unadjusted model (β = 1.2, 95% CI: 0.21–2.2, p = 0.018); however, this association was substantially attenuated and no longer statistically significant after multivariable adjustment (Model 3: β = 0.32, 95% CI: −0.75 to 1.4, p = 0.601). A similar pattern was observed for the TyG index (Tables S4 and S5). In the Chinese T2DM cohort, a similar trend was observed for the TyG index (Table S6).

Discussion

In an analysis of a representative sample of US adults with data collected between 1999 and 2000, H. pylori infection showed a tendency toward an association with the TyG index. However, this association became nonsignificant (p = 0.101) after full adjustment for confounders. We strongly suspect that this association may be coincidental and requires further confirmation. We also found no association between H. pylori infection and HOMA-IR and TG levels. Furthermore, we did not find a significant correlation between H. pylori infection and TyG levels in the southern Chinese population. In the subgroup of US participants with T2DM, the associations between H. pylori infection and both the TyG index and HOMA-IR were similarly attenuated and lost significance after multivariable adjustment.

The most plausible explanation for these findings is the potent confounding effect of sociodemographic and cardiometabolic factors. As demonstrated by the baseline characteristics, H. pylori–positive individuals exhibited a markedly different risk profile, including lower socioeconomic status and a heavier burden of traditional cardiometabolic comorbidities. These factors are independently associated with both an increased risk of H. pylori acquisition in early life and the development of IR and dyslipidemia later in life. When such confounders are inadequately accounted for, they may generate spurious associations. Our use of sequential modeling and stratified analyses demonstrates how this apparent association dissolves after rigorous adjustment. This underscores the methodological imperative in observational studies of H. pylori to thoroughly adjust for this complex web of confounders, a step that may have been insufficiently addressed in prior studies reporting positive associations.

Our results provide a compelling framework for reconciling inconsistencies in the existing literature. The positive associations reported in some earlier studies, often based on smaller sample sizes, may reflect residual confounding rather than a direct pathophysiological link.^30–34 Conversely, our large, well-characterized cohorts with comprehensive adjustment lend greater weight to the null findings reported in other large-scale studies. Furthermore, the absence of an association even within the T2DM subgroup—where confounding by medication use and disease severity is most pronounced—strengthens the conclusion that a clinically relevant independent association is unlikely. The controversial evidence regarding eradication therapy further supports this interpretation.^17,18,33 If the association were causal and robust, more consistent benefits from eradication would be expected.

These results have several important clinical implications. First, they provide robust evidence against diverting clinical attention toward H. pylori as a significant modifiable risk factor for IR. The consistency of findings across populations argues against routine screening for or treatment of H. pylori for the specific purpose of improving insulin sensitivity. Furthermore, this evidence supports the avoidance of unnecessary antibiotic use, thereby mitigating public health threats of antibiotic resistance and potential treatment-related side effects. Thus, clinicians should continue to prioritize lifestyle interventions and evidence-based pharmacotherapy targeting conventional risk factors, as these remain the most effective strategies for mitigating IR and its associated cardiometabolic consequences.

Several limitations should be acknowledged when analyzing our results. First, the cross-sectional design precludes the establishment of causal or temporal relationships. Second, substantial missing data on specific antidiabetic and antihypertensive medications in the NHANES database prevented direct adjustment for these potential confounders; this limitation was addressed as methodologically feasible through stratified analyses by T2DM status. Third, the interpretability and generalizability of the findings are further constrained by potential selection biases. These include the high exclusion rate in the NHANES cohort, which resulted in an analytical sample with a different risk profile, and the hospital-based setting of the Southern China cohort, which limits the extrapolation of results to the general population. Finally, the absence of data on specific bacterial virulence factors (e.g. cytotoxin-associated gene A (CagA) and vacuolating cytotoxin gene A (VacA)) and systemic inflammatory mediators precluded investigation into the underlying biological mechanisms.

Conclusions

In summary, our findings indicate that H. pylori infection is not associated with IR. Further research is needed to better identify the relationship between H. pylori infection and IR, which may provide novel insights into the prevention and treatment of cardiometabolic diseases.

Supplemental Material

sj-pdf-1-imr-10.1177_03000605261416741 - Supplemental material for Association between Helicobacter pylori infection and insulin resistance: Data from the US and Chinese cohorts

Supplemental material, sj-pdf-1-imr-10.1177_03000605261416741 for Association between Helicobacter pylori infection and insulin resistance: Data from the US and Chinese cohorts by Shali Hao, Youbing Zhang, Lingxiao Li, Zerong Chen, Jiahuan Li, Yue Cao, Libin Mo, Yangguang Liu, Ling Zhao, Xiaohui Huang, Yuli Huang and Xiaoyan Cai in Journal of International Medical Research

Footnotes

Acknowledgments

Not applicable.

Author contributions

SH, YZ, LL, XC, and YH contributed to the study conception and design. YC, JL, YL, ZC, XH, LM, and LZ interpreted, analyzed, and acquired the data. Manuscript drafting was performed by SH, whereas critical revisions were performed by XC and YH. All authors agree to be accountable for the study aspects to guarantee accuracy and integrity.

Availability of data

Data supporting the study outcomes would be provided by corresponding author on reasonable demand.

Consent for publication

Not applicable.

Declaration of conflicting interests

The authors declare no competing interests.

Ethics approval and consent to participate

The NHANES was accredited by the Research Ethics Review Board of the National Center for Health Statistics, Center of Disease Control. The cohort from South China was registered at the China Clinical Trial Registration Center (NO: 202100912).

Funding

This project was supported by the Guangdong Medical Research Foundation (grant number BB2023061); the Scientific Research Start Plan of Shunde Hospital, Southern Medical University (grant number SRSP2024004); and Guangdong Basic and Applied Basic Research Fund (2019B1515120044; 2024A1515140190).

Supplemental material

Supplemental material for this article is available online.

ORCID iD

Yuli Huang

References

Zamani

Ebrahimtabar

Zamani

, et al. Systematic review with meta-analysis: the worldwide prevalence of Helicobacter pylori infection. Aliment Pharmacol Ther 2018; 47: 868–876.

Naito

Yoshikawa

Molecular and cellular mechanisms involved in Helicobacter pylori-induced inflammation and oxidative stress. Free Radic Biol Med 2002; 33: 323–336.

Liu

, et al. Helicobacter pylori infection combined with non-alcoholic fatty liver disease increase the risk of atherosclerosis: focus in carotid artery plaque. Medicine (Baltimore) 2019; 98: e14672.

Reuter

Raspe

Uebner

, et al. Treatment with Helicobacter pylori-derived VacA attenuates allergic airway disease. Front Immunol 2023; 14: 1092801.

Doheim

Altaweel

Elgendy

, et al. Association between Helicobacter pylori infection and stroke: a meta-analysis of 273,135 patients. J Neurol 2021; 268: 3238–3248.

Wong

Rayner-Hartley

Byrne

MF.

Extraintestinal manifestations of Helicobacter pylori: a concise review. World J Gastroenterol 2014; 20: 11950–11961.

Santos

MLC

De Brito

Da Silva

FAF

, et al. Helicobacter pylori infection: beyond gastric manifestations. World J Gastroenterol 2020; 26: 4076–4093.

Zimmet

Alberti

Shaw

Global and societal implications of the diabetes epidemic. Nature 2001; 414: 782–787.

Adeva-Andany

Martínez-Rodríguez

González-Lucán

, et al. Insulin resistance is a cardiovascular risk factor in humans. Diabetes Metab Syndr 2019; 13: 1449–1455.

10.

Stumvoll

Goldstein

Van Haeften

TW.

Type 2 diabetes: principles of pathogenesis and therapy. Lancet 2005; 365: 1333–1346.

11.

Wang

Guo

, et al. Insulin resistance, blood glucose and inflammatory cytokine levels are risk factors for cardiovascular events in diabetic patients complicated with coronary heart disease. Exp Ther Med 2018; 15: 1515–1519.

12.

Cersosimo

Solis-Herrera

Trautmann

, et al. Assessment of pancreatic β-cell function: review of methods and clinical applications. Curr Diabetes Rev 2014; 10: 2–42.

13.

Mirr

Skrypnik

Bogdański

, et al. Newly proposed insulin resistance indexes called TyG-NC and TyG-NHtR show efficacy in diagnosing the metabolic syndrome. J Endocrinol Invest 2021; 44: 2831–2843.

14.

Schmiegelow

Hedlin

Stefanick

, et al. Insulin resistance and risk of cardiovascular disease in postmenopausal women: a cohort study from the women's health initiative. Circ Cardiovasc Qual Outcomes 2015; 8: 309–316.

15.

Cheng

Yang

Hung

, et al. Helicobacter pylori eradication improves glycemic control in type 2 diabetes patients with asymptomatic active Helicobacter pylori infection. J Diabetes Investig 2019; 10: 1092–1101.

16.

Song

Cai

Jin

, et al. The efficacy of Helicobacter pylori eradication in diabetics and its effect on glycemic control: a systematic review and meta-analysis. Helicobacter 2021; 26: e12781.

17.

Imai

Yamada

Saito

, et al. Eradication of insulin resistance. Lancet 2009; 374: 264.

18.

Park

Jeon

Kim

, et al. Helicobacter pylori eradication has no effect on metabolic and inflammatory parameters. J Natl Med Assoc 2005; 97: 508–513.

19.

Demir

Gokturk

Ozturk

, et al. Helicobacter pylori prevalence in diabetes mellitus patients with dyspeptic symptoms and its relationship to glycemic control and late complications. Dig Dis Sci 2008; 53: 2646–2649.

20.

Malamug

Karnchanasorn

Samoa

, et al. The role of Helicobacter pylori seropositivity in insulin sensitivity, beta cell function, and abnormal glucose tolerance. Scientifica (Cairo) 2014; 2014: 870165.

21.

Devrajani

Shah

SZA

Soomro

, et al. Type 2 diabetes mellitus: a risk factor for Helicobacter pylori infection: a hospital based case-control study. Int J Diabetes Dev Ctries 2010; 30: 22–26.

22.

Bener

Micallef

Afifi

, et al. Association between type 2 diabetes mellitus and Helicobacter pylori infection. Turk J Gastroenterol 2007; 18: 225–229.

23.

Yang

Xuan

Influence of Helicobacter pylori infection on metabolic syndrome in old Chinese people. Gastroenterol Res Pract 2016; 2016: 6951264.

24.

Chen

Blaser

MJ.

Association between gastric Helicobacter pylori colonization and glycated hemoglobin levels. J Infect Dis 2012; 205: 1195–1202.

25.

Hsieh

Wang

Hsieh

, et al. Helicobacter pylori infection associated with high HbA1c and type 2 diabetes. Eur J Clin Invest 2013; 43: 949–956.

26.

Tseng

Metformin and Helicobacter pylori infection in patients with type 2 diabetes. Diabetes Care 2018; 41: e42–e43.

27.

Tseng

Diabetes, insulin use and Helicobacter pylori eradication: a retrospective cohort study. BMC Gastroenterol 2012; 12: 46.

28.

Guerrero-Romero

Simental-Mendía

González-Ortiz

, et al. The product of triglycerides and glucose, a simple measure of insulin sensitivity. Comparison with the euglycemic-hyperinsulinemic clamp. J Clin Endocrinol Metab 2010; 95: 3347–3351.

29.

Matthews

Hosker

Rudenski

, et al. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985; 28: 412–419.

30.

Aydemir

Bayraktaroglu

Sert

, et al. The effect of Helicobacter pylori on insulin resistance. Dig Dis Sci 2005; 50: 2090–2093.

31.

Eshraghian

Hashemi

Hamidian

, et al. Helicobacter pylori infection as a risk factor for insulin resistance. Dig Dis Sci 2009; 54: 1966–1970.

32.

Ozdem

Akcam

Yilmaz

, et al. Insulin resistance in children with Helicobacter pylori infection. J Endocrinol Invest 2007; 30: 236–240.

33.

Gen

Demir

Ataseven

Effect of Helicobacter pylori eradication on insulin resistance, serum lipids and low-grade inflammation. South Med J 2010; 103: 190–196.

34.

Aslan

Horoz

Nazligul

, et al. Insulin resistance in H pylori infection and its association with oxidative stress. World J Gastroenterol 2006; 12: 6865–6868.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.40 MB