Gastric cancer among American Indians and Alaska Natives in the United States, 1999–2004§

Authors


  • This supplement was sponsored by Cooperative Agreement Number U50 DP424071-04 from the Centers for Disease Control and Prevention, Division of Cancer Prevention and Control and Prevention.

  • The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.

  • §

    This article is a U.S. Government work and, as such, is in the public domain in the United States of America.

Abstract

BACKGROUND.

Gastric cancer incidence rates for American Indians and Alaska Natives (AI/ANs) historically have exceeded those for non-Hispanic whites (NHWs). Previous reports may have underestimated the true burden of gastric cancer in AI/AN populations because of misclassification of AI/AN race in cancer registries.

METHODS.

Population-based cancer registry data from 1999 through 2004 were used to describe gastric cancer incidence in AI/ANs and NHWs in the US. To address misclassification of race, registry data were linked with Indian Health Service administrative records, and analyses were restricted to residents of Contract Health Service Delivery Areas (CHSDA). Disease patterns were assessed for 6 geographic regions and for all regions combined. Rates were expressed per 100,000 population and were age-adjusted to the 2000 US standard population.

RESULTS.

In CHSDA counties, gastric cancer incidence rates for AI/ANs were higher than the rates for NHWs across most regions. For both sexes combined, AI/AN rates ranged from 6.1 in the East region to 24.5 in Alaska; there was relatively little regional variation in NHW rates. Most patients with gastric cancer were diagnosed with late-stage disease, regardless of race, age, or sex. In some regions, cancer rates in the central/distal portions of the stomach were higher among AI/ANs than among NHWs, whereas rates in the proximal stomach were similar between the 2 populations.

CONCLUSIONS.

AI/ANs are generally at greater risk for gastric cancer than NHWs. Relatively high rates of cancer in the central/distal portions of the stomach among AI/ANs in some geographic regions may indicate a disproportional burden of Helicobacter pylori-associated disease. Cancer 2008;113(5 suppl):1225–33. Published 2008 by the American Cancer Society.

There is wide geographic variation in the occurrence of gastric cancer.1–3 Economically developing countries generally have higher incidence and mortality rates of the disease than developed countries,3 with the notable exceptions of Japan and Korea. Gastric cancer claims an estimated 857,000 lives each year worldwide and is second only to hepatocellular carcinoma as a leading cancer cause of death.1–3

For reasons that are not understood fully, gastric cancer incidence and mortality rates declined in many countries over the course of the 20th century.4 Possible reasons that have been hypothesized to explain the decline have included increased fruit and vegetable consumption, decreased intake of foods preserved with salt or by smoking, declines in salt intake, widespread availability of refrigeration, and the reduced prevalence of Helicobacter pylori infection. In the US, gastric cancer mortality rates per 100,000 among males declined from 46.3 in 1930 to 5.9 in 2004, and the rates among females declined from 35.2 in 1930 to 3.0 in 20045, 6 Nonetheless, gastric cancer remains a highly fatal condition, because the majority of patients are diagnosed with late-stage disease that is difficult to treat.7, 8

Previous studies have demonstrated that AI/AN populations are at higher risk for this disease than the general US population.9, 10 In the Southwestern US and in Alaska, incidence rates for gastric cancer among AI/AN populations remain high despite declining rates in other racial/ethnic populations from the same regions.9, 10 However, it is likely that misclassification of AI/ANs as other races in central cancer registries11–14 resulted in an underestimate of the true burden of gastric cancer in these populations.

This report provides a comprehensive overview of the burden of gastric cancer among AI/AN populations in the US from both nationwide and regional perspectives. Rates for NHW populations are presented for comparison. To minimize the effects of the misclassification of race for AI/ANs, cancer registry data were linked with Indian Health Service (IHS) patient services records, and the analyses were restricted to residents of counties where such linkages were most efficacious.

MATERIALS AND METHODS

Cancer Cases

Incident gastric cancer cases diagnosed during 1999 through 2004 were identified from population-based registries that participate in the National Cancer Institute Surveillance, Epidemiology, and End Results (SEER) Program15 and/or the Centers for Disease Control and Prevention National Program of Cancer Registries (NPCR).16 During the period covered by this study, participating registries classified tumor histology, tumor behavior, and primary cancer site (ie, topography) according to the Third Edition of the International Classification of Diseases for Oncology (ICD-O).17

Eligible cases included all malignant neoplasms of the stomach (ICD-O topography codes C16.0-C16.9 and ICD-O behavior code 3). Lymphomas (ICD-O histology codes 9590-9769), mesothelioma (ICD-O histology codes 9050-9055), and Kaposi sarcoma (ICD-O histology code 9140) were excluded from the analysis. Benign and in situ tumors (ICD-O behavior codes 0 and 2, respectively) also were excluded along with tumors of uncertain or unknown behavior (ICD-O behavior code 1).

Cancer registries usually document AI/AN ancestry from statements in medical and vital records. However, previous studies have demonstrated that registries often misclassify AI/ANs as other races.11–14 For the current analysis, cancer registry records were linked with IHS patient services files to address this problem, because AI/AN individuals must provide proof of membership in a federally recognized tribe to receive healthcare from the IHS.18

IHS provision of healthcare to AI/AN populations is considered to be greatest in Contract Health Service Delivery Area (CHSDA) counties, which generally are defined as those counties that contain or are adjacent to federally recognized tribal reservations and/or trust lands. There is evidence that AI/AN race misclassification occurs less often in CHSDA counties.19 For this reason, 1 set of incidence rates was calculated for residents of all US counties, and a second set of rates was calculated for residents of CHSDA counties. Figure 1 illustrates the incidence rates calculated for each of 6 IHS regions (Alaska, Pacific Coast, Northern Plains, Southern Plains, Southwest, and East) and for all regions combined. These IHS regions were chosen because they are consistent with previous reports of regional patterns of specific health outcomes and disease risk factors for AI/ANs.20–22 Approximately 56% of the US AI/AN population resides in CHSDA counties. This proportion varies by IHS region: Alaska, 100%; East, 13.1%; Northern Plains, 59%; Southern Plains, 64.1%; Pacific Coast, 55.6%; Southwest, 87.5%. Additional details regarding CHSDA and IHS and the data sources and methods used for this analysis are provided elsewhere.18

Figure 1.

States and Contract Health Service Delivery Area counties by Indian Health Service Region.

Standards for coding stage of disease at diagnosis changed during the period of this study (1999-2004). To avoid incomparability among the staging schemes,23 the analysis of stage of disease at diagnosis was restricted to incident cases diagnosed during the years 2001 through 2003 and coded according to SEER Summary Stage 2000.24 The staging categories were as follows: localized for disease that was restricted to the stomach, regional for disease that extended directly into organs and areas adjacent to the stomach, and distant for disease that had metastasized to portions of the body not directly adjacent to the stomach. The undetermined category was for those whose disease stage at diagnosis could not be determined.

The site of primary tumor within the stomach may be related to the underlying etiology of disease.25 In general, cancers that arise in central/distal regions of the stomach are associated more closely with H. pylori infection than cancers in the proximal stomach. To characterize the topographic distribution of gastric cancers among AI/ANs and NHWs, cases were grouped as follows: proximal (cardia and fundus), distal (gastric body, lesser curvature, greater curvature, antrum, and pylorus), and overlapping/unknown (overlapping sites or unknown primary site). Because most studies of gastric cancer etiology were focused on adenocarcinoma, our analysis of cases by anatomic subsite was restricted solely to patients with adenocarcinoma.

Statistical Analyses

Average annual age-adjusted incidence rates were calculated by using the direct method.26 Rates were expressed per 100,000 population and age-adjusted to the 2000 US standard population.27 Age-specific rates were calculated for 4 categories: aged <40 years, ages 40 to 49 years, ages 50 to 64 years, and aged ≥65 years. Ratios comparing incidence rates among AI/ANs with those among NHWs were calculated by dividing the former by the latter, confidence intervals (CIs) for age-adjusted rates and standardized rate ratios (RR) were calculated based on methods described by Tiwari et al28 using SEER*Stat version 6.3.6.29 Denominators for rate calculations were derived from population estimates from the US Bureau of the Census.18 Differences between AI/ANs and NHWs by categories of disease stage were evaluated with the chi-square statistic using standard modules in SAS software.30

RESULTS

In total, 701 incident gastric cancer cases were diagnosed among AI/ANs in participating cancer registries during the period 1999 through 2004 (Table 1). A majority of these cases (ie, 606 cases representing 86.5% of all incident cases) were diagnosed among AI/AN residents of CHSDA counties. Gastric cancer incidence rates for AI/AN populations residing in CHSDA counties were uniformly higher than rates based on AI/AN residents of all counties combined (except in Alaska, where all counties are designated as CHSDAs). In contrast, there was little difference in NHW rates between CHSDA counties and all counties combined. These findings were consistent with improved classification of AI/AN cancer cases within CHSDA counties, which increased rates for AI/ANs but had minimal effect on rates for NHWs. Consequently, all remaining findings were based on rates that were calculated for residents of CHSDA counties.

Table 1. Gastric Cancer Incidence by Indian Health Service Region for American Indians/Alaska Natives and Non-Hispanic Whites: US, 1999-2004a
IHS RegionSexCHSDA CountiesAll Counties
AI/AN CountAI/AN Rateb95% CI for AI/AN RateNHW RatebRR (AI/AN:NHW)95% CI for RRAI/AN CountAI/AN Rateb95% CI for AI/AN RateNHW RatebRR (AI/AN:NHW)95% CI for RR
  • Source: Cancer registries in the Center for Disease Control's National Program of Cancer Registries (NPCR) and the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) Program.

  • CHSDA indicates Contract Health Services Delivery Area; IHS, Indian Health Service; AI/AN, American Indians/Alaska Natives; 95% CI, 95% confidence interval; NHW, non-Hispanic whites; RR, rate ratio.

  • a

    AI/AN race is reported by NPCR and SEER registries or through linkage with the IHS patient registration database. AI/AN persons of Hispanic origin are included.

  • b

    Rates are per 100,000 persons and are age-adjusted to the 2000 US standard population.

  • c

    The RR is statistically significant (P < .05).

  • d

    Rates and RRs for Alaska in the CHSDA Counties section are the same as those in the All Counties section, because all counties in Alaska are CHSDA counties.

  • Years of data and registries used: 1999-2004 (41 states and the District of Columbia): Alaska,* Ala,* Ark, Ariz,* Calif,* Colo,* Conn,* DC, Del, Fla,* Ga, Hawaii, Iowa,* Idaho,* Ill, Ind,* Ky, La,* Mass,* Me,* Mich,* Minn,* Mo, Mont,* NC,* Neb,* NH, NJ, NM,* Nev,* NY,* Ohio, Okla,* Ore,* Pa,* RI,* Tex,* Utah,* Wash,* Wis,* WVa, and Wyo*; 1999 and 2002-2004: ND*; 2001-2004: SD*; 2003-2004: Miss* and Va; 2004: Tenn (asterisks indicate states with at least 1 county designated as a CHSDA).

  • Percent regional coverage of AI/AN in CHSDA counties versus AI/AN in all counties: Alaska, 100%; East, 13.1%; Northern Plains, 59%; Southern Plains, 64.1%; Pacific Coast, 55.6%; Southwest, 87.5%.

Northern PlainsBoth sexes9713.210.5-16.35.62.35c1.86-2.911169.47.6-11.45.91.59c1.29-1.94
 Males5818.713.6-24.98.52.21c1.60-2.957113.510.1-17.68.81.54c1.15-2.01
 Females399.26.4-12.73.42.69c1.86-3.74456.44.5-8.63.71.72c1.22-2.33
AlaskadBoth sexes8824.519.3-30.45.34.61c3.34-6.388824.519.3-30.45.34.61c3.34-6.38
 Males5234.624.8-46.67.74.46c2.89-6.845234.624.8-46.67.74.46c2.89-6.84
 Females3617.712.2-24.73.05.90c3.41-10.383617.712.2-24.73.05.90c3.41-10.38
Southern PlainsBoth sexes1179.07.4-10.85.11.77c1.44-2.151287.25.9-8.55.21.39c1.15-1.66
 Males6310.58.0-13.57.21.46c1.10-1.91688.16.2-10.37.61.060.81-1.37
 Females547.65.7-9.93.42.23c1.64-2.97606.34.8-8.13.31.91c1.45-2.47
Pacific CoastBoth sexes907.96.2-9.85.61.40c1.11-1.751085.24.2-6.36.00.870.70-1.06
 Males6112.29.0-16.18.31.48c1.09-1.95727.95.9-10.18.80.900.68-1.15
 Females294.73.1-6.73.51.340.88-1.94363.22.2-4.43.80.840.58-1.17
EastBoth sexes206.13.6-9.56.90.880.52-1.38612.51.9-3.36.40.39c0.29-0.51
 Males117.93.6-14.710.10.790.36-1.46383.72.5-5.39.40.40c0.27-0.56
 Females94.82.2-9.24.41.090.48-2.07231.71.0-2.54.20.40c0.25-0.60
SouthwestBoth sexes19411.59.9-13.34.92.37c2.02-2.7620010.69.1-12.34.82.19c1.88-2.54
 Males11415.312.5-18.57.12.17c1.75-2.6411814.011.5-17.07.02.00c1.62-2.42
 Females808.66.8-10.83.02.86c2.22-3.62827.96.2-9.93.02.60c2.03-3.27
TotalBoth sexes60610.810.0-11.85.81.88c1.72-2.057017.26.7-7.86.11.19c1.10-1.29
 Males35914.713.1-16.48.51.74c1.55-1.954199.98.8-10.98.91.110.99-1.23
 Females2477.96.9-9.03.62.18c1.90-2.492825.24.6-5.93.91.34c1.18-1.51

Among residents of CHSDA counties, gastric cancer incidence rates varied by geographic region, race, and sex (Table 1) (Fig. 2). Rates for AI/AN males exceeded those for NHW males in all areas except the East. Rates among AI/AN males ranged from 7.9 in the East to 34.6 in Alaska. In contrast, there was relatively little regional variation in the rates for NHW males (range, 7.1-10.1). Rates for AI/AN females exceeded those for NHW females in all IHS regions. Similar to males, there was large geographic variation in incidence among AI/AN females (range, 4.7-17.7) but not among NHW females (range, 3.0-4.4). Incidence rates were higher for males than females among both AI/ANs and NHWs in all IHS regions.

Figure 2.

Gastric cancer incidence rates (per 100,000 population and age-adjusted to the 2000 US standard population) and corresponding 95% confidence limits for American Indians/Alaska Natives (AI/AN) and non-Hispanic whites (NHW) by sex and Indian Health Service Region and Contract Health Services Delivery Area counties, US, 1999-2004 (source: cancer registries in the Centers for Disease Control and Prevention National Program of Cancer Registries and/or the National Cancer Institute Surveillance, Epidemiology, and End Results Program).

Gastric cancer incidence rates increased with age (Table 2). For all IHS regions combined, AI/AN rates exceeded NHW rates at every age. In the Northern Plains and the Southwest, incidence rates for AI/ANs were greater than rates for NHWs at every age, and the differences in Alaska were particularly pronounced. In the remaining IHS regions, differences between AI/ANs and NHWs were less pronounced, but AI/AN rates tended to be higher. Males rates exceeded females rates at all ages among both AI/ANs and NHWs (data not shown).

Table 2. Gastric Cancer Incidence Rates and Rate Ratios by Age and Indian Health Service Region for American Indians/Alaska Natives and Non-Hispanic Whites in Contract Health Service Delivery Area Counties: US, 1999-2004
IHS Region<40 Years40-49 Years50-64 Years≥65 Years
AI/AN RateaNHW RateaRR95% CIAI/AN RateaNHW RateaRR95% CIAI/AN RateaNHW RateaRR95% CIAI/AN RateaNHW RateaRR95% CI
  • Source: Cancer registries in the Center for Disease Control's National Program of Cancer Registries and the National Cancer Institute's Surveillance, Epidemiology, and End Results Program.

  • IHS indicates Indian Health Service; AI/AN, American Indians/Alaska Natives; NHW, non-Hispanic whites; RR, rate ratio; 95% CI, 95% confidence interval.

  • a

    Rates are per 100,000 persons and are age-adjusted to the 2000 US standard population.

  • b

    The RR is statistically significant (P < .05).

  • Years of data and registries used: 1999-2004 (41 states and the District of Columbia): Alaska,* Alabama,* Arkansas, Arizona,* California,* Colorado,* Connecticut,* the District of Columbia, Delaware, Florida,* Georgia, Hawaii, Iowa,* Idaho,* Illinois, Indiana,* Kentucky, Louisiana,* Massachusetts,* Maine,* Michigan,* Minnesota,* Missouri, Montana,* North Carolina,* Nebraska,* New Hampshire, New Jersey, New Mexico,* Nevada,* New York,* Ohio, Oklahoma,* Oregon,* Pennsylvania,* Rhode Island,* Texas,* Utah,* Washington,* Wisconsin,* West Virginia, and Wyoming*; 1999 and 2002-2004: North Dakota*; 2001-2004: South Dakota*; 2003-2004: Mississippi* and Virginia; 2004: Tennessee (asterisks indicate states with at least 1 county designated as a Contract Health Service Delivery Area).

Northern Plains1.20.25.69b2.46-11.397.12.33.07b1.60-5.3917.58.42.09b1.35-3.0969.630.82.26b1.65-3.04
Alaska2.20.116.27b3.24-154.7015.02.65.73b2.47-13.1526.16.04.37b2.26-8.26134.531.14.33b2.83-6.61
Southern Plains0.20.20.880.10-3.335.92.22.64b1.39-4.7213.47.01.90b1.27-2.7747.028.01.68b1.29-2.16
Pacific Coast0.20.21.050.21-3.031.82.30.790.29-1.7410.98.51.290.86-1.8646.030.41.51b1.12-2.00
East0.30.31.080.03-6.333.92.81.400.29-4.169.59.80.970.35-2.1230.638.10.800.38-1.50
Southwest0.60.23.43b1.56-6.803.82.11.810.98-3.1217.47.32.39b1.78-3.1463.226.52.39b1.94-2.91
Total0.60.22.69b1.82-3.834.92.42.05b1.57-2.6415.18.41.78b1.52-2.0859.131.71.87b1.67-2.08

A majority of gastric cancer cases in both AI/ANs and NHWs were diagnosed at regional or distant stages of disease (data not shown). For all IHS regions combined, AI/ANs were slightly less likely than NHWs to be diagnosed at early stages of disease, but these differences did not achieve statistical significance (P = .48). Similar patterns were observed by geographic region.

For all IHS regions combined, incidence rates of adenocarcinoma of the proximal stomach generally were similar for AI/ANs and NHWs, whereas rates of adenocarcinoma of the central/distal stomach were higher for AI/ANs than for NHWs (Table 3). However, this pattern was not uniform across IHS regions or by sex.

Table 3. Gastric Adenocarcinoma Incidence Rates and Rate Ratios by Anatomic Subsite, Sex, and Indian Health Service Region for American Indians/Alaska Natives and Non-Hispanic Whites in Contract Health Service Delivery Area Counties: US, 1999-2004
IHS RegionSubsiteMalesFemales
AI/ANNHWAI/AN:NHWAI/ANNHWAI/AN:NHW
Ratea95% CIRatea95% CIRR95% CIRatea95% CIRatea95% CIRR95% CI
  • Source: Cancer registries in the Center for Disease Control's National Program of Cancer Registries and the National Cancer Institute's Surveillance, Epidemiology, and End Results Program.

  • IHS indicates Indian Health Service; AI/AN, American Indians/Alaska Natives; NHW, non-Hispanic whites; 95% CI, 95% confidence interval; RR, rate ratio.

  • a

    Rates are per 100,000 persons and are age-adjusted to the 2000 US standard population.

  • b

    Includes primary sites C16.0-C16.1 (topography codes from the International Classification of Diseases for Oncology).

  • c

    Includes primary sites C16.2-C16.6.

  • d

    Includes primary sites C16.8-C16.9.

  • e

    The RR is statistically significant (P < .05).

  • Years of data and registries used: 1999-2004 (41 states and the District of Columbia): Alaska,* Alabama,* Arkansas, Arizona,* California,* Colorado,* Connecticut,* the District of Columbia, Delaware, Florida,* Georgia, Hawaii, Iowa,* Idaho,* Illinois, Indiana,* Kentucky, Louisiana,* Massachusetts,* Maine,* Michigan,* Minnesota,* Missouri, Montana,* North Carolina,* Nebraska,* New Hampshire, New Jersey, New Mexico,* Nevada,* New York,* Ohio, Oklahoma,* Oregon,* Pennsylvania,* Rhode Island,* Texas,* Utah,* Washington,* Wisconsin,* West Virginia, and Wyoming*; 1999 and 2002-2004: North Dakota*; 2001-2004: South Dakota*; 2003-2004: Mississippi* and Virginia; 2004: Tennessee (asterisks indicate states with at least 1 county designated as a Contract Health Service Delivery Area).

Northern PlainsProximalb4.72.3-8.23.83.6-4.11.230.60-2.161.90.7-3.90.80.7-0.92.450.91-5.10
 Central/distalc6.23.3-10.41.81.7-2.03.39e1.77-5.724.52.7-7.01.11.0-1.24.04e2.37-6.41
 Otherd5.02.7-8.31.51.4-1.73.23e1.73-5.442.51.1-4.60.80.7-0.93.15e1.42-5.94
AlaskaProximalb14.58.4-22.83.02.0-4.44.77e2.43-8.922.10.7-4.90.70.3-1.52.860.69-11.11
 Central/distalc11.86.6-19.12.01.0-3.55.79e2.58-13.5311.36.9-17.10.80.3-1.614.86e5.95-44.94
 Otherd6.52.4-13.11.20.5-2.25.55e1.64-16.712.50.9-5.71.00.4-1.92.570.70-8.89
Southern PlainsProximalb4.02.6-6.02.62.3-3.01.530.96-2.321.20.5-2.20.70.5-0.81.750.72-3.56
 Central/distalc2.51.4-4.11.71.4-2.01.440.77-2.462.81.7-4.30.90.8-1.13.02e1.76-4.87
 Otherd2.51.3-4.21.51.2-1.81.670.84-2.972.01.1-3.30.90.8-1.12.18e1.14-3.79
Pacific CoastProximalb2.31.1-4.33.93.7-4.10.600.28-1.110.60.2-1.50.80.8-0.90.720.18-1.81
 Central/distalc5.03.0-7.71.91.8-2.02.62e1.55-4.071.40.6-2.61.21.1-1.31.170.53-2.18
 Otherd3.01.6-5.11.31.1-1.42.42e1.27-4.111.80.8-3.30.70.7-0.82.50e1.13-4.64
EastProximalb1.80.2-6.04.13.8-4.30.440.05-1.471.20.1-3.91.00.9-1.11.210.15-4.16
 Central/distalc1.40.3-4.42.72.5-2.90.510.10-1.642.80.9-6.51.61.5-1.81.720.53-4.01
 Otherd2.60.4-7.51.81.6-1.91.490.25-4.290.40.0-2.50.80.8-0.90.480.01-2.94
SouthwestProximalb2.21.2-3.53.02.8-3.30.710.40-1.151.10.6-2.00.60.5-0.71.960.96-3.54
 Central/distalc7.55.5-9.91.41.3-1.65.27e3.80-7.133.42.3-4.91.00.8-1.13.51e2.27-5.15
 Otherd2.51.5-3.91.41.3-1.61.80e1.07-2.812.01.2-3.10.80.7-0.92.59e1.49-4.16
TotalProximalb3.72.9-4.63.73.6-3.81.000.79-1.241.20.8-1.60.80.8-0.91.47e1.02-2.04
 Central/distalc5.54.5-6.62.01.9-2.12.75e2.26-3.323.42.8-4.11.21.2-1.32.76e2.24-3.37
 Otherd3.22.5-4.01.51.4-1.52.16e1.67-2.751.91.5-2.50.80.7-0.82.46e1.84-3.21

DISCUSSION

Results from this study document dramatic regional variation in the incidence of gastric cancer among AI/ANs that was not observed among NHWs residing in the same geographic areas. AI/ANs in some regions were diagnosed with a higher proportion of central/distal gastric cancers, which may indicate a disproportionate burden of disease because of H. pylori. It is unlikely that these results were influenced substantially by the misclassification of race for AI/ANs, because this issue was addressed by linking cancer registry data with IHS databases and by restricting the analysis to residents of CHSDA counties.

The molecular biology of gastric cancer is complex and varies by gastric site and histology. The incidence of gastric cancer in the stomach cardia, which accounts for 39% of gastric cancers in US males,31 has been increasing and may involve the same elusive risk factors that are driving increasing rates of esophageal adenocarcinoma. By comparison, rates for gastric cancer that affect other stomach sites have fallen sharply over the last half century in the US and elsewhere.5

Adenocarcinoma of the stomach commonly is grouped into 2 primary variants: diffuse and intestinal (well differentiated).32 The incidence of the diffuse type of cancer is similar in most populations, suggesting that the intestinal type may be responsible for regional variation. The diffuse type of gastric cancer is more commonly hereditary, affects younger patients, and often is associated with blood group A. Hereditary diffuse gastric cancer follows an autosomal-dominant pattern and involves mutation in the cell adhesion protein E-cadherin.33 Sporadic, nonhereditary cases involve bialleic inactivation of the cadherin 1 gene CDH1, which encodes for E-caderin.34 By comparison, the intestinal type of gastric cancer is more common in older individuals, involves the stomach body and antrum, and mirrors adenocarcinoma elsewhere in the gastrointestinal tract.35 The molecular sequence of events leading to the intestinal-type variant is not completely understood, but appears to follow a chronic gastritis-atrophy-metaplasia-dysplasia-carcinoma sequence.36

Over the past 2 decades, data have been accumulating that support an increasingly strong causal relation between infection with H. pylori and the diagnosis of noncardia gastric adenocarcinoma.37–39 Prospective studies in high-risk populations have reported H. pylori infection as a definite risk factor for development of gastric cancer.40, 41 Infection with H. pylori results in a chronic, active immune response that, in the absence of antibiotic-induced eradication, persists for the life of the host.

H. pylori is a common infection that causes chronic gastritis and peptic ulcer disease.42, 43 It has been characterized by the International Agency for Research on Cancer as a Class 1 carcinogen.44 Prevalence of H. pylori infection is related chiefly to age and geographic location. Estimates from the current National Health and Nutrition Examination Survey revealed an overall antibody prevalence of 27.1%, with prevalence estimates approximately 20% greater for non-Hispanic black, Mexican-American, and other Hispanic groups.45 Seroprevalence increases with age, and previous studies in the US have documented that seroprevalence among individuals aged ≥60 years is approximately 50%.46–53 Although most data on H. pylori prevalence in AI/AN communities are based on work in Alaska,54–57 where prevalence is approximately 75%, a population-based survey in an American Indian community in Montana also revealed a high prevalence (53%).58 In developing countries, prevalence approaches 90%, with most individuals infected before age 10 years.59, 60

Environmental factors other than H. pylori also influence the risk of developing gastric cancer. Consumption of fresh fruits and vegetables has been associated consistently with lowered risk of the disease.61 Micronutrients, including vitamin E (α-tocopherol), carotenoids, selenium, and especially vitamin C (ascorbic acid), all have been linked with reduced risk, although short-term intervention trials with these nutrients has not demonstrated protective effects.62, 63 High consumption of salt, nitrite, and nitrates has been associated consistently with gastric cancer risk. Refrigeration may have played a role in reducing gastric cancer rates over the last 60 years by decreasing reliance on food-preservation methods such as salt curing, pickling, and meat smoking (the latter 2 are sources of carcinogenic N-nitroso compounds).64, 65 Cigarette smoking also has been associated with an increased risk of gastric cancer; however, the absence of control for confounders such as H. pylori infection and fresh produce consumption have hindered the interpretation of many studies.66–68

Obesity may be an important risk factor for gastric cancer of the cardia.69 Behavioral Risk Factor Surveillance System data indicate that AI/AN populations from all regions have a higher prevalence of obesity than NHW populations.70 Despite this finding, proximal gastric cancer incidence was only significantly higher in AI/ANs than in NHWs among males in Alaska. Family history also confers an elevated risk of gastric cancer. Between 10% and 30% of patients with gastric cancer have a family history of the disease,71–73 and twin studies have suggested that inherited genes contribute approximately 28% of the added risk, with environmental factors making up the remainder.74

In summary, this report on gastric cancer in AI/AN populations builds on previous publications by addressing the misclassification of race and by presenting incidence rates both nationwide and by region. Results from this study suggest a need to better characterize the burden of H. pylori among AI/ANs and NHWs and to clarify whether this factor alone is responsible for the disproportionate burden of gastric cancer in AI/AN populations. Regional differences in gastric cancer incidence in AI/AN populations may provide an opportunity to elucidate risk factors and identify future means of prevention.

Ancillary