A cross-sectional study of Dietary patterns and Helicobacter pylori infection among American Indian adults in the Southwest

Abstract

High sodium diets have been shown to promote stomach colonization and induction of tissue damage by Helicobacter pylori (H. pylori), a risk factor for gastric cancer. Among American Indians in the Southwest, where the H. pylori prevalence is 60%, the association between diet and H. pylori infection has not been studied. We conducted a cross-sectional pilot study with adults in the Navajo Nation to assess their diet with self-administered food questionnaires and to detect H. pylori from stool samples using droplet digital PCR. Three diet patterns were identified among 93 adults (51% 18–44 years, 73% female): 1) Western, 2) Soups and Mixed Dishes, and 3) Fruits and Vegetables. Participants in the highest and middle tertiles of the Soups and Mixed Dishes pattern scores had higher odds of having H. pylori (OR_Highest=5.59, 95% CI, 1.50–23.70; OR_Middle=3.48, 95% CI, 1.08–12.32) than those in the lowest tertile. This positive association may be linked to the sodium content of foods in this diet pattern. Soups and Mixed dishes may contribute to H. pylori infection and be a potential modifiable lifestyle factor that cancer prevention efforts in the Navajo Nation may consider incorporating into nutrition education for individuals positive for H. pylori infection.

I. Introduction

Substantial gastric cancer disparities exist for American Indians and Alaska Natives (AI/AN) in nearly all regions of the United States (U.S.).¹ For American Indian adults in the Navajo Nation, the largest tribe in the Southwest, the prevalence of Helicobacter pylori (H. pylori), a known risk factor for gastric cancer, is twice as high as that of the U.S. population (~60% stool-based versus 27%, seroprevalence).^2,3 The high H. pylori prevalence in the Navajo Nation may be a contributing factor to high incidence and mortality rates of gastric cancer in this population.⁴

Diet is an important contributor to the gastric environment and may influence the course of H. pylori infection. Epidemiological studies have highlighted diets high in sodium and processed meats as being associated with H. pylori infection,^5–7 although some studies show no association.^8,9 Proposed mechanisms have linked sodium to increased pathogenicity of H. pylori through changes in the stomach’s protective mucous lining, aiding in the colonization and expression of virulence genes of H. pylori, and initiating inflammatory responses of the gastric epithelium that increase epithelial cell proliferation.^10,11

There are limited studies that examine the relationship between diet and H. pylori infection, specifically within American Indian communities, particularly at the tribal level. The extent of dietary investigation and H. pylori infection among AI/AN people has been limited to assessing the type of water consumed and alcohol consumption.^3,12,13 In particular, a study of 101 Navajo adults found that individuals who consumed unregulated water (i.e. not disinfected or monitored for safety) had a higher odds of H. pylori infection than individuals who consumed regulated water.³ To add to the limited literature on diet and H. pylori among AI/AN people and to assess more than one diet factor, our study used an a posteriori approach to derive whole diet patterns of Navajo adults to examine the association between diet and H. pylori infection. We used data from food questionnaires that included tribal foods from 104 Navajo adults from our 2021–2022 pilot study. Given the high prevalence of H. pylori infection in the Navajo Nation and the mechanistic association of sodium and H. pylori, we hypothesized that Navajo adults with a diet high in sodium-containing foods would have higher odds of H. pylori infection than adults with diets lower in sodium-containing foods.

2. Materials and Methods

Study Participants

Details of the cross-sectional community-based pilot study have been described previously.² Only participants who identified as Navajo, were 18 years old or older, resided in the Navajo Nation, were not pregnant, had not used oral or intravenous antibiotics in the past 3 months, had not used proton pump inhibitors in the past 3 months, were not treated for H. pylori infection in the past 3 months, and were not undergoing any cancer treatment were included. Participant demographic factors included age (categorized as 18–44, 45–44, 55+), sex (female/male), and educational level (≤High school, >High school). Measures were adapted from prior research.^3,14

Dietary assessment

Two self-administered food questionnaires assessed dietary intake. Participants were asked about their diet in the past year using a general food frequency questionnaire (FFQ) and a Navajo Food Questionnaire.

Food Frequency Questionnaire (FFQ)

The FFQ, developed by the Nutrition Assessment Shared Resource (NASR) of Fred Hutchinson Cancer Center, Seattle, WA, assessed dietary intake by asking participants to report the frequency of consumption and portion size of 181 food items over the last year.¹⁵ Each item was defined by a series of foods or beverages that were categorized into 10 major food groups: CEREALS, BREADS, SNACKS; MEAT-FISH-EGGS; SPAGHETTI, MIXED DISHES, SOUPS; DAIRY PRODUCTS; VEGETABLES AND GRAINS; SAUCES AND CONDIMENTS; FRUITS; SWEETS; BEVERAGES AND ALCOHOL. The annual intake of each food item was calculated using the frequency of consumption (never or less than once per month, 1 per month, 2–3 per month, 1 per week, 2 per week, 3–4 per week, 5–6 per week, 1 per day, 2+ per day) and portion size (small, medium, large).

Navajo Food Questionnaire

To assess the consumption of tribal foods over the last year, we developed a Navajo Food Questionnaire by referring to tribal food lists from previous studies^16,17 and following the same consumption categorization as the NASR FFQ. We selected 13 tribal foods based on foods consumed among American Indian and Navajo people.^17–19 The 13 food items were frybread, Navajo taco, Navajo roast mutton, Navajo burger, pueblo bread, kneel down bread (baked white corn), blue corn mush (blue cornmeal and juniper ash pudding), mutton and/or beef stew with dumplings and/or macaroni, mutton and/or beef stew with corn or any vegetables, blood sausages (boiled sheep stomach stuffed with sheep blood and vegetables), achee’ (sheep fat wrapped with sheep stomach intestine), Navajo cake (white corn cornmeal cake), and chiilchin (red sumac pudding). These tribal foods were grouped into five food groups: BREADS; SOUPS OR STEWS; CEREAL; MEATS; DESSERTS AND SNACKS. Navajo and nutrition researchers (PS, FM, MG, NT) reviewed and pilot-tested the questionnaire to determine the appropriateness of selected foods and the length of the questionnaire. The values used for the frequency of consumption were the same as for the general FFQ. Annual intake for each tribal food was calculated using the frequency of consumption and medium portion size values.

Identification of dietary patterns

Of the combined 181-FFQ and 13-Navajo food items, we first removed food items with low consumption (≤10%, resulting in the removal of 13 food items) and then categorized food items into 20 food groups based on culinary value and nutritional content (Table 1).^{6,9,16,20–23} Subsequently, we applied principal component analysis (PCA) with varimax rotation to the 20 food groups using R studio version 4.0.3 (psych package; R Core Team, Vienna, Austria). The Kaiser Criterion and Scree plot methods, along with prior knowledge, were used to guide the selection of components for final analysis.^6,9,24,25 By inspecting the scree plot of the rotated principal components, we selected a subset of principal components with eigenvalues >1.0 as dietary patterns of interest. Among this subset of principal components, food groups with an absolute factor loading ≥ 0.30 were considered the main contributors to the dietary patterns, as was conducted in other studies.^6,9 Dietary patterns were named according to the food groups listed in Table 1 with an absolute factor loading ≥ 0.30 and informed by other studies.^6,9 The factor scores for each diet pattern were categorized into tertiles based on the tertile distribution for H. pylori negative participants.

Table 1.

Food groups used in the principal component analysis.

Criteria for group	Food group	Food items
Nutrient profile (sodium)	Processed meats & fried meats	Bacon, sausage, hot dogs, bratwurst, lunch meat, spam, fried fish, fried chicken
Nutrient profile (sodium/carbohydrates)	Refined grains	Cereals, breads, pancakes, white rice, frybread
Nutrient profile (carbohydrates)	Sweets & pastries	Jams, cereal bars, cookies, candy, donuts, Navajo cake
Nutrient profile (carbohydrates)	Sweets	Ice cream, puddings
Nutrient profile (sodium)	Salty snacks	Chips, popcorn, crackers
Nutrient profile (protein)	Meat	Eggs, red meat, white meat, fish, and seafood
Culinary use	Organ meat	Liver, organ meats, achee, blood sausage
Nutrient profile (dairy)	Dairy	Cream, milk, yogurt, sour cream
Nutrient profile (fats)	Fats	Butter, margarine, lard, meat drippings, oils
Nutrient profile (sodium)	Soups	Meat stews, chili, ramen, Navajo stews (dumplings or vegetable)
Nutrient profile (sodium)	Mixed dishes	Spaghetti, pasta, Asian noodles, burritos, enchiladas
Nutrient profile (carbohydrates)	Potatoes	Fried, boiled, baked, mashed, salads
Nutrient profile (phytochemicals)	Vegetables	Green lettuce/spinach/collards, carrots, tomatoes, peppers/chili, broccoli, cabbage, peas, beans, squash, sweet potatoes, onions, garlic, avocado, kneel down bread
Nutrient profile (phytochemicals)	Fruits	Apples, bananas, peaches, apricots, dried fruits, oranges, berries, melons, grapes, cherries, pineapple, fruit cocktail
Nutrient profile (caffeine)	Coffee	Coffee, latte, tea
Nutrient profile (vitamins)	Juices	Fruit juices, vegetable juices
Nutrient profile (carbohydrates)	Soft drinks	Diet and regular soft drinks
Nutrient profile (alcohol)	Alcohol	Beer, wine, liquor
Nutrient profile (sodium, fats)	Sauces	Cheese or cream sauces, gravies, salad dressing, salsa, mayonnaise
Nutrient profile (fiber)	Whole grain/High fiber	Whole-grain crackers, brown rice, whole-grain bread, high fiber cereal

Measurement of Helicobacter pylori

H. pylori was determined from stool samples collected by the participant. Self-sampling stool collection instructions were adapted from previous research and provided to participants.²⁶ Participants placed stool samples (approximately 1 teaspoon) in a vial with 5 mL of 95% ethanol preservative (Fisher Scientific) and stored at room temperature before mailing to the Salama Lab at the Fred Hutchinson Cancer Center (FHCC) in Seattle, WA. Detailed methods for droplet digital PCR (polymerase chain reaction) analysis of samples and determination of H. pylori status (positive or negative) have been described in a separate publication.²

Statistical analyses

Logistic regression models assessed the association between diet factor score tertiles and H. pylori infection. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using participants in the lowest tertile (T1) as the referent group. ORs were calculated in two models, an unadjusted model and an adjusted model controlling for age, sex, and total energy intake (kcal/day, log-transformed due to uneven distribution). Due to potential confounding and the effects of measurement error by energy intake, total energy intake was included in the adjusted model.²⁷ To test for trends, we created a continuous variable from the median factor scores of each tertile and regressed H. pylori infection on this continuous variable in a logistic regression model. All analyses were performed using R Studio version 4.0.3 (R Core Team, Vienna, Austria). A p<0.05 was considered statistically significant.

3. Results

Study population

A total of 104 participants completed questionnaires and provided stool samples. From these, we excluded participants with incomplete dietary data (n=6) or extreme energy intakes based on previous study cutoff values (<500 (n=4) or >5,500 (n=1) kcal/day for females and <500 (n=0) or >8,000 (n=0) for males).^28,29 Final analyses were based on 93 participants.

The H. pylori prevalence was 57.0% (95% CI, 46.3–67.2; Table 2). Participant characteristics by H. pylori status are detailed in Table 2. A total of 50.5% of participants were 18–44 years old, 22.6% 45–54 years old, 26.9% 55+ years old. Of the participants, 73.1% were female, and 77.4% had attained more than a high school education.

Table 2.

Characteristics of tribal adults by Helicobacter pylori (H. pylori) status.

Characteristics	Total n (%)	H. pylori-negative n (%)	H. pylori-positive n (%)
Total	93 (100.0)	40 (43.0)	53 (57.0)
Age group, years
18–44	47 (50.5)	17 (42.5)	30 (56.6)
45–54	21 (22.6)	10 (25.0)	11 (20.8)
55+	25 (26.9)	13 (32.5)	12 (22.6)
Sex
Female	68 (73.1)	30 (75.0)	38 (71.7)
Male	25 (26.9)	10 (25.0)	15 (28.3)
Education
≤High school	21 (22.6)	12 (30.0)	9 (17.0)
>High school	72 (77.4)	28 (70.0)	44 (83.0)
FFQ represented dieta
Completely	24 (25.8)	13 (32.5)	11 (20.8)
Mostly	49 (52.7)	21 (52.5)	28 (52.8)
Somewhat or Not at all	18 (19.4)	5 (12.5)	13 (24.5)

Abbreviation: FFQ=Food Frequency Questionnaire.

^aParticipants were asked if the Fred Hutchinson Cancer Center FFQ used in the study represented the foods they usually ate

Dietary patterns

After varimax rotation, the principal component analysis identified six dietary components with eigenvalue >1.0 (Supplemental Figure 1). Three dietary components were subsequently selected because they explained the largest variance, which was 50.4% of the total variation. Component 1 explained 31.8% followed by 10.4% for component 2, and 8.2% for component 3. The factor loadings of each component are shown in Table 3. Based on the food items with the highest loadings within each component, the name of the component was based on these foods or the overall theme of the pattern. Component 1 was labeled the “Western” diet pattern characterized by the intake of processed meats, refined grains, sweets and pastries, salty snacks, meat, dairy, potatoes, and vegetables; component 2 was the “Soups & Mixed Dishes” diet pattern which included the intake of soups and mixed dishes such as spaghetti, pasta, Asian noodles, burritos, and enchiladas; component 3 was identified as the “Fruits & Vegetables” diet pattern characterized by the intake of fruits and vegetables.

Table 3.

Rotated factor-loading matrix for the 3 dietary patterns^*.

Food groups	Western	Soups/Mixed Dishes	Fruits/Vegetables
Processed meats	0.866	-	-
Refined grains	0.696	-	-
Sweets & pastries	0.668	-	-
Sweets	0.776	-	-
Salty snacks	0.745	-	-
Meat	0.847	-	-
Organ meat	-	-	-
Dairy	0.376	-	-
Fats	-	-	-
Soups	-	0.941	-
Mixed dishes	-	0.924	-
Potatoes	0.580	-	-
Vegetables	0.341	-	0.791
Fruits	-	-	0.872
Coffee	-	-	-
Juices	-	-	-
Soft drinks	-	-	-
Alcohol	-	-	-
Sauces	0.603	-	-
Whole grain/High fiber	0.740	-	-
Variance explained (%)	31.8	10.4	8.2

^*For simplicity, absolute values <0.3 were excluded.

- Absolute factor-loading values less than 0.3 were suppressed.

Association between dietary patterns and H. pylori infection

The associations between dietary patterns and H. pylori are shown in Table 4. The Soups and Mixed Dishes pattern was positively associated with H. pylori in the adjusted model for tertile 2 (OR=3.48, 95% CI, 1.8–12.32) and tertile 3 (OR=4.84, 95% CI, 1.37–18.87) compared to tertile 1. The trend for increasing odds of H. pylori across the Soups and Mixed Dishes pattern tertiles was significant (p=0.03). No significant associations were observed between the Western or the Fruits and Vegetable patterns and H. pylori in the unadjusted and adjusted models.

Table 4.

Logistic Regression Models of Dietary Patterns and Helicobacter pylori (H. pylori) infection.

	Tertilesa of factor scores (n=93)
Dietary patterns	T1	T2	T3	p for trend
Western	(−1.07, −0.588)	(−0.589, 0.0253)	(0.0254, 4.93)
H. pylori infection (n)	12	21	20
Model 1	1.00	1.88 (0.67–5.41)	1.79 (0.64–5.17)	0.372
Model 2	1.00	1.72 (0.54–5.59)	1.52 (0.33–6.94)	0.750
Soups/Mixed Dishes	(−1.05, −0.398)	(−0.399, −0.143)	(−0.144, 1.40)
H. pylori infection (n)	6	19	28
Model 1	1.00	3.41 (1.07–11.87)b	5.03 (1.64–17.10)b	0.014b
Model 2	1.00	3.48 (1.08–12.32)b	4.84 (1.37–18.87)b	0.033b
Fruits/Vegetables	(−1.78, −0.654)	(−0.655, 0.293)	(0.0294, 2.31)
H. pylori infection (n)	9	21	23
Model 1	1.00	2.51 (0.86–7.68)	2.75 (0.95–8.35)	0.113
Model 2	1.00	2.87 (0.96–9.16)	2.80 (0.91–9.10)	0.134

^atertile categories based on the distribution of H. pylori-negative participants

^bstatistical significance p<0.05;

Model 1 is unadjusted; Model 2 is adjusted for age (continuous), gender (female, male), calories (log-transformed)

4. Discussion

In this study, we identified three dietary patterns and found that a Soups and Mixed Dishes diet pattern was positively associated with H. pylori infection in Navajo adults. No significant associations were observed between the Western or the Fruits and Vegetable diet pattern and H. pylori infection.

The positive association between the Soups and Mixed Dishes diet pattern and H. pylori infection may be explained by foods in this diet pattern with known high sodium content. High sodium content foods can damage the gastric mucosal and facilitate colonization of the bacteria.^30,31 The proportion of high sodium daily intake (>3,765 mg/day) among Navajo adults with H. pylori was 46.4%, which was higher than Navajo adults without H. pylori (32.6%).² This sodium daily intake is more than the Dietary Guidelines for Americans (2,300 mg/day).³² The Soups and Mixed Dishes diet pattern includes meat/chili/Navajo stews, ramen, spaghetti, pasta, Asian noodles, burritos, and enchiladas, which may be processed and prepared with added sodium³³ in restaurants or from local food stands and food trucks in the Navajo Nation.³⁴ Consistent with our findings, Tsugane et al. found a positive association between daily consumption of miso soup, a high sodium food, and H. pylori infection risk in Japanese men.⁵ Similarly, a large cross-sectional study in China identified a positive association between a “high salt” diet pattern that included noodles and instant noodles and H. pylori infection in 3,014 older adults (45–59 years old).⁶ In contrast, a large cross-sectional study in China found a negative association between a “high protein/cholesterol” diet pattern that included instant noodles and H. pylori in 10,407 adults; however, instant noodles were not the main contributors to that diet pattern.⁹ Another explanation is the link between carbohydrates and H. pylori risk, as the foods in the Soups and Mixed Dishes pattern can be considered refined carbohydrates (e.g., noodles, pasta).^9,20,35 Although the mechanism is unclear, it is also hypothesized that high consumption of carbohydrates can play a significant role in H. pylori infection through abnormal glucose metabolism.³⁵

We found no association between our Western diet pattern and H. pylori infection, which was also found in another cross-sectional study of about 3,000 Chinese adults.⁶ Shu et al. reported that a Western dietary pattern characterized by high consumption of red and/or processed meat and energy-dense foods was not associated with H. pylori infection.⁶ However, our results differed from those of a large cross-sectional study based in China of 10,407 adults.⁹ Xia et al. identified two diet patterns similar to our Western diet pattern and found that a high carbohydrate/sweet pattern was positively associated with H. pylori and a high protein/cholesterol pattern was inversely associated with H. pylori.⁹ Because the foods in our Western diet pattern aligned with foods in both the high carbohydrate/sweet and high protein/cholesterol patterns of the Xia study and the results were opposing, this may be evidence for our non-significant finding. The mixed effects of the individual foods within our Western diet pattern may have attenuated the effect of the association between our Western diet pattern and H. pylori. On the one hand, high consumption of processed meat, salty snacks, and refined grains contains high sodium content and studies have shown high sodium intake promotes the colonization and viability of H. pylori through deterioration and damage to the gastric mucosa.³⁶ On the other hand, animal foods such as red meat are rich in selenium, vitamin A, and vitamin D, which can decrease the risk for H. pylori infection. Specifically, selenium is an essential nutrient that protects membranes from oxidative damage³⁷, and vitamins A and D are known to play a critical role in immune response and aid in suppressing H. pylori inflammation by reducing inflammatory signaling.³⁸ Furthermore, the lack of association between this Western pattern and H. pylori could also be due to reverse causality. Participants who may be experiencing gastrointestinal symptoms may have limited their intake of foods in the Western diet pattern.

Further, we found no significant association between a Fruits and Vegetable diet pattern and H. pylori. This supports previous studies with a large sample of Asian adults that showed no association between a “balanced” and “health-conscious” diet pattern and H. pylori.^6,9 The healthy dietary constituents from fruits and vegetables have shown protective effects against H. pylori.^22,40,41 The fruits and vegetables in our Fruits and Vegetables diet pattern contain flavonoids that protect the gastric lining³⁹, and phytochemicals, such as isothiocyanates. Isothiocyanate sulforaphane has been shown to have antibacterial activity and reduce H. pylori growth.^21,40 Yet, we found no association in our study. Moreover, these null findings may be explained by a narrow range of intakes of fruits and vegetables. Navajo people with or without H. pylori have previously reported a lower intake of vegetables (63.6%) and fruits (58.6%) than that recommended in the Dietary Guidelines for Americans (>3.0 cups of vegetables/day; >2.0 cups of fruits/day), which may explain our findings.^2,32 How individuals prepare their vegetables and fruits, which was not captured in this study. Eating raw versus cooked vegetables may have also influenced the quality of nutrients participants consume, resulting in non-significant findings.

These results increase our understanding of diet and H. pylori; however, it is important to shed light on the historical injustices and policies that have shaped the food choices and diet of the Navajo people. Indigenous food systems have been displaced by colonization, and the resulting economic, political, and environmental changes have disrupted the food environments of Indigenous people, including the Navajo Nation.^41,42 Structural barriers exist, such as large distances to grocery stores, income disparities, and lack of transportation, which present a challenging food environment in the Navajo Nation.⁴³ There are 15 grocery stores and 85 small convenience stores selling food across the vast 17 million acres of the Navajo reservation.^44,45 Therefore, low access to affordable healthy foods can result in low intake of fruits and vegetables and high intake of fats, calorie-dense, and high-sodium foods, which are highly processed and easier to access and store in the Navajo Nation. As such, the Navajo people may have moved away from a traditional way of life, including a traditional diet, due to the influence of Western culture and structural barriers.⁴⁶ Further, the continued environmental pollution due to mining and resource extraction has compromised the essential water sources in the Navajo Nation, impacting farming and growing healthy foods.^47,48

Our study presents some notable strengths. Firstly, this study is the only study, to date, investigating the association between dietary patterns and H. pylori infection in American Indian populations. Secondly, our sample of participants comes from the community rather than a hospital-based population. Thirdly, this study identifies the dietary patterns of adults in the Navajo Nation, as well as captures the intake of tribal foods. The dietary patterns in our study may drive further research questions about specific foods or nutrients and H. pylori and gastric cancer risk.

Some limitations of the study should be considered when interpreting these results. First, a convenience sampling method was used to select participants, limiting the generalizability of results to the full Navajo Nation. Second, the small sample sizes in some of our analyses may have limited the power to detect statistical significance. The relatively small sample of cases across dietary patterns may have resulted in imprecise estimates and should be interpreted with caution, and may limit the generalizability of findings. Third, the a posteriori approach is a data-driven approach to derive diet patterns specific to the Navajo study population, and results are less likely to be comparable to other study populations. Fourth, participants may not have accurately recalled dietary information such as specific foods they consumed, frequency, or portion sizes, thus limiting our ability to obtain accurate estimates, even though we accounted for energy intake to reduce potential measurement error and included tribal foods. Fifth, the FFQ used in our study may not have fully captured food intake for certain participants. In particular (Table 2), 19.4% of participants (n = 18) indicated that the FFQ ‘somewhat or not at all’ represented their diet. Of the 19.4%, 13 of the 18 participants had H. pylori. Therefore, developing a culturally appropriate FFQ with relevant foods, beverages, and portion sizes with the tribe and individuals with H. pylori can improve the dietary assessment in the Navajo Nation. Lastly, because our study is cross-sectional, and diet and H. pylori were assessed simultaneously, reverse causation is possible, where individuals with H. pylori may have altered their diet due to stomach symptoms rather than the diet being directly associated with H. pylori. This makes it challenging to determine the direction of the relationship and results must be interpreted with caution.

Conclusion

Our study found a Soups and Mixed Dishes dietary pattern significantly associated with higher odds of H. pylori infection among Navajo adults. Further studies are needed to understand the relationship between diet and H. pylori infection in American Indian populations with high H. pylori prevalence. These results can inform nutrition counselors and clinicians about the diet of the Navajo population and improve cancer prevention education about diets with high sodium for individuals positive for H. pylori.

Data Availability Statement

The Navajo Nation owns the data collected and analyzed in this study. Requests to access the data can be made to the Navajo Nation Human Research Review Board (https://nnhrrb.navajo-nsn.gov/index.html).