Table 2. Modeled intestinal-type NCGA outcomes between 1978 and 2040: age-standardized incidence, percent change in incidence, and relative contribution of alternative risk factor scenarios to the base-case scenario percent decline in incidence.
| Scenariosa , b , c | Age-Standardized Intestinal-Type NCGA Incidence (per 100,000) | Percent Change in Age-Standardized Intestinal-Type NCGA Incidence, Mean (Range)d | Percent Relative Contribution of Alternative Risk Factor Scenario to Base-Case Scenario Percent Decline in Age-Standardized Intestinal-Type NCGA Incidence, Mean (Range)d , e | ||||
| Historical | Projected | Historical | Projected | ||||
| 1978 | 2008 | 2040 | 1978–2008 | 2008–2040 | 1978–2008 | 2008–2040 | |
| Base case (all risk factors) | 11.0 | 4.4 | 2.3 | −60.1 (−55.5 to −64.8) | −47.3 (−34.7 to −59.4) | — | — |
| H. pylori and smoking only | 12.7 | 9.2 | 5.7 | −28.1 (−17.8 to −35.4) | −37.7 (−27.0 to −47.7) | 46.7 (29.8–57.9) | 80.5 (60.7–100.0)f |
| H. pylori only | 10.9 | 8.1 | 6.4 | −25.7 (−21.1 to −29.7) | −21.4 (−16.5 to −25.4) | 42.8 (35.4–48.4) | 46.0 (34.7–65.8) |
| Smoking only | 15.4 | 15.1 | 12.1 | −2.0 (−10.3 to +6.6) | −19.9 (−10.6 to −29.2) | 3.2 (0.0–16.8)g | 41.8 (28.6–66.6) |
| All other causes only | 11.1 | 5.7 | 4.6 | −49.0 (−40.0 to −55.9) | −19.2 (−6.1 to −33.4) | 81.5 (70.3–93.4) | 40.2 (11.9–68.8) |
| +H. pylori | 9.6 | 3.7 | 2.6 | −61.2 (−55.2 to −67.8) | −30.4 (−18.2 to −43.3) | 101.9 (91.9–112.9) | 64.4 (39.9–81.1) |
| +Smoking | 12.9 | 6.5 | 4.0 | −49.6 (−43.9 to −54.5) | −38.3 (−24.9 to −53.6) | 82.6 (74.5–89.7) | 80.5 (62.6–98.2) |
| No tobacco control | 11.4 | 5.0 | 3.5 | −56.0 (−49.8 to −59.8) | −30.4 (−17.0 to −43.0) | 93.1 (82.7–100.0)h | 64.3 (46.1–83.4) |
| Complete tobacco control | 10.1 | 3.6 | 1.6 | −63.9 (−57.7 to −67.8) | −55.0 (−37.2 to −76.6) | 106.3 (100.0–114.2)i | 115.8 (100.0–140.7)i |
For scenarios in which smoking was halted at pre-1900 birth cohort levels, we classified all individuals who initiated smoking after 1925 as never smokers (as smoking rates were relatively stable before then) and assumed that all smokers faced an annual rate of cessation of 1%.
For scenarios in which H. pylori was halted at pre-1900 birth cohort levels, we assumed a prevalence of 73%.
For scenarios in which “all other risk factors” was halted at pre-1900 birth cohort levels, we assumed a negligible rate of decline in the probability of developing atrophy for all birth cohorts.
Among the 50 randomly selected good-fitting natural history parameter sets identified via calibration.
Based on comparisons to the base-case scenario within each parameter set. Calculated as the alternative risk factor scenario percent decline divided by the base-case scenario percent decline, and reported as a percentage.
Range includes estimates for two parameter sets in which the intestinal-type NCGA decline estimated with only H. pylori and smoking trends exceeded the base case, reflecting the underlying dynamics with “all other risk factors.” Assumed the percentage of the base-case scenario explained by “H. pylori and smoking only” was 100% for these parameter sets.
Range includes 14 parameter sets for which intestinal-type NCGA incidence increased as a result of smoking trends. Assumed the percentage of the base-case scenario explained by “smoking alone” was zero for these parameter sets.
Range includes estimates for two parameter sets in which more than 100% of the observed decline was explained by “no tobacco control,” as a result of higher background mortality rates among smokers (e.g., intestinal-type NCGA incidence was lower as higher smoking rates resulted in a greater number of individuals dying from smoking-related competing risks and therefore fewer individuals being at risk to develop intestinal-type NCGA). Assumed the percentage of the base-case scenario explained by “no tobacco control” was zero for these parameter sets.
Range includes estimates for six parameter sets (two for 1978–2008 and four for 2008–2040) in which less than 100% of observed decline was explained by “complete tobacco control,” as a result of lower background mortality rates among non-smokers (e.g., intestinal-type NCGA incidence was higher as lower rates of smoking resulted in more individuals living longer and developing intestinal-type NCGA). Assumed the percentage of the base-case scenario explained by “complete tobacco control” was 100% for these parameter sets.