Prevalence of and risk factors for H. pylori within the Namutumba District
Of the 400 adult participants contacted for the study, 376 (94%) met the inclusion criteria, completed the survey questionnaire, and provided a stool sample for fecal H. pylori antigen testing. The majority of excluded participants declined to complete the survey questionnaire and/or provide a stool sample (Fig. 1). Table 1 demonstrates the baseline characteristics for the study population and a comparison between H. pylori-positive and –negative participants.
Table 1
Participant characteristics.
Characteristic | H. pylori-positive (n = 181) | H. pylori-negative (n = 195) | P value |
Age in years, median (IQR) | 45 (23) | 40 (25) | 0.3461 |
Gender, n (%) Female Male | 124 (68) 57 (32) | 138 (70.8) 57 (29.2) | 0.6338 |
Marital status, n (%) Married or cohabitating Single and never married Widowed Separated | 151 (83.4) 7 (3.9) 6 (3.3) 17 (9.4) | 166 (85.1) 9 (4.6) 7 (3.6) 13 (6.7) | 0.7889 |
Highest level of education, n (%) No education Primary Secondary or above | 43 (23.8) 95 (52.5) 43 (23.7) | 68 (34.9) 93 (47.7) 34 (17.4) | 0.0443 |
Proximity to health services, n (%) Less than 5 km Between 5 km and 10 km More than 10 km | 136 (75.1) 38 (21) 7 (3.9) | 130 (66.7) 56 (28.7) 8 (4.1) | 0.2092 |
Smoking status, n (%) Cigarettes Marijuana Other Don’t know | 3 (1.6) 2 (1.1) 0 (0) 1 (0.6) | 3 (1.5) 0 (0) 1 (0.5) 0 (0) | 0.2208 |
Alcohol use, n (%) 1 day per week or less 2 days per week 3 days per week 4 days per week or more | 20 (11) 5 (2.8) 3 (1.6) 10 (5.5) | 36 (18.5) 3 (1.5) 4 (2.0) 4 (2.0) | 0.1115 |
NSAID use, n (%) | 49 (27.1) | 37 (19) | 0.081 |
Number of persons per household, median (IQR) | 6 (4) | 7 (4.5) | 0.4157 |
Number of children in household, median (IQR) | 4 (3) | 4 (3.5) | 0.2459 |
Households with pets/livestock within living quarters, n (%) | 44 (24.3) | 45 (23.1) | 0.7789 |
IQR: Interquartile range.
H. pylori: Helicobacter pylori
NSAID: Non-steroidal anti-inflammatory drug.
At baseline, there was no significant difference between H. pylori-positive and H. pylori-negative participants in most of the demographic and socioeconomic factors analyzed. Of note, those with higher levels of education attained (secondary education and above) were significantly more likely to be positive for H. pylori compared to those with no formal education or having only achieved a primary education (p = 0.0443). While there was a trend toward higher NSAID use among H. pylori-positive participants, the difference was not statistically significant (p = 0.081).
The point prevalence of H. pylori within this study population was 48% (Fig. 1). The magnitude of H. pylori cases varied regionally within the Namutumba District, with the Namutumba sub-county accounting for the highest number of cases and the Nabweyo sub-county recording the fewest number of cases (Fig. 2A). When controlling for the number of participants sampled from each sub-county, the prevalence of H. pylori ranged from 22–68%, with the Namutumba sub-county again representing the highest density of H. pylori cases within the district (Fig. 2B). No statistically significant differences in H. pylori prevalence were observed between any of the sub-counties (not shown).
Dyspepsia among H. pylori-positive and –negative participants
Given the prevalence of H. pylori within this population and the variable symptomatology associated with H. pylori infection12, 26, we sought to determine whether the presence or severity of dyspepsia correlated with H. pylori positivity. We also aimed to identify potential dyspeptic symptoms that might predict H. pylori positivity. To quantify the degree of dyspepsia within our study population, we assigned each participant a dyspepsia score based on the SFLDQ, a questionnaire to quantitatively categorize dyspepsia severity25 that has been validated among African patients27. Of the 376 participants, 326 (86.7%) reported some degree of dyspepsia (SFLDQ score > 0; Figs. 1 and 3A). Most dyspeptic participants reported mild dyspepsia (SFLDQ score 1–8), with similar proportions reporting moderate (SFLDQ score 9–15) and severe dyspepsia (SFDLQ score > 15). The severity of dyspepsia was similar between H. pylori-positive and –negative participants (Figs. 3B-C). Accordingly, the median dyspepsia scores did not significantly differ based on H. pylori positivity (median SFLDQ score 8 vs median SFLDQ score 7; Supplemental Table 3). In addition, H. pylori-positive participants were not significantly more likely to report indigestion, heartburn, regurgitation, and/or nausea, the four symptom components of the SFLDQ, compared to H. pylori-negative participants.
As per the most recent guidelines14, chronically dyspeptic participants found to be negative for H. pylori completed one month of a daily proton pump inhibitor (i.e., omeprazole). Of participants who reported some degree of dyspepsia and who were negative for H. pylori (n = 168), only those with dyspepsia scores of 8 or above (n = 91) were selected to undergo daily omeprazole therapy (Fig. 1). After one month of omeprazole, a significant majority of dyspeptic participants reported subjective improvement in their dyspepsia (not shown).
Among H. pylori-negative participants, the presence or severity of dyspepsia was not significantly associated with various demographic, socio-economic, or behavioral characteristics that have been previously shown to correlate with dyspepsia, including gender, level of education, alcohol use, or NSAID use28 (Table 2).
Table 2
Dyspepsia severity among Helicobacter pylori-negative participants.
Characteristic | None (n = 27) | Mild (n = 89) | Moderate (n = 45) | Severe (n = 34) | P value |
Female, n (%) Male, n (%) | 17 (63)* 10 (37)* | 59 (66)** 30 (34)** | 35 (78)# 10 (22)# | 27 (79)## 7 (21)## | 0.2665 *0.4076 *0.4478 #0.3448 ##0.3000 |
Level of education, n No education Primary Secondary or above | 10 16 1 | 26 43 20 | 16 20 9 | 16 14 4 | 0.1373 |
Alcohol use, n 1 day per week or less 2 days per week 3 days per week 4 days per week or more | 24 1 1 1 | 64 18 2 3 | 35 9 0 1 | 25 8 1 0 | 0.5747 |
NSAID use, n (%) | 4 (15) | 13 (15) | 13 (29) | 7 (20) | 0.2267 |
Use of herbal medicines, n Yes No | 2 24 | 19 68 | 13 31 | 11 22 | 0.0668 |
* Comparison between males and females with no dyspepsia (dyspepsia score of 0).
** Comparison between males and females with mild dyspepsia (dyspepsia score of 1–8).
# Comparison between males and females with moderate dyspepsia (dyspepsia score of 9–15).
## Comparison between males and females with severe dyspepsia (dyspepsia score > 15).
While there appeared to be greater dyspepsia severity among women compared to men, this was not statistically significant. However, the probability of developing severe dyspepsia among all participants increased with age, while the probability of having mild dyspepsia decreased with age (Fig. 4A). Similar findings were seen among H. pylori-negative participants (Fig. 4B). Interestingly, the probability of being free of dyspepsia (i.e., SFLDQ score of 0) increased with age among H. pylori-positive participants, while the probability of having severe dyspepsia was largely unchanged as a function of age (Fig. 4C). Within our study population, therefore, the probability of developing severe dyspepsia or being free of dyspepsia could be modeled as a function of age, based on H. pylori status (Supplemental Table 4; see Methods).
The SFLDQ relies on assessing the frequency of four symptom components, including indigestion, heartburn, regurgitation, and nausea. While no individual symptom was significantly associated with H. pylori positivity (Supplemental Table 3), there was a strong correlation among the frequency of dyspeptic symptoms within the sampled population. Among all participants, those reporting one symptom from the SFLDQ were significantly more likely to report a separate symptom, regardless of the symptom. Similarly, participants reporting the absence of one symptom were significantly more likely to also report the absence of a different symptom. These correlations were modeled and found to be highly statistically significant across all combinations of symptoms (Figs. 5A-F) and were independent of gender (Supplemental Fig. 1). For example, heartburn and indigestion correlated strongly (p = 5.01 × 10− 14; Fig. 5A), such that participants with frequent heartburn (i.e., once a day or more) were also significantly more likely to report frequent indigestion. Similarly, those reporting no heartburn were also significantly more likely to report no indigestion.
Efficacy of triple therapy on H. pylori eradication and dyspepsia severity
To determine the efficacy of triple therapy on H. pylori eradication, all participants who tested positive for H. pylori, regardless of the presence of dyspepsia, underwent 14 days of standard treatment (see Methods). Of the 181 participants who were positive for H. pylori, 171 (94.4%) began triple therapy. Nine participants were lost to follow-up prior to starting treatment, and one participant declined treatment (Fig. 1). Participants undergoing treatment met with study team members three to five times during their 14-day regimen to assess for symptoms, adverse reactions, and medication compliance. Symptoms experienced during the treatment regimen are listed in Supplemental Table 5. No adverse events were reported.
Approximately one month after completing treatment, participants met with study team members to complete a follow-up questionnaire and to submit a stool sample for fecal H. pylori antigen testing. Of the 171 participants who underwent treatment, 165 (96.4%) filled the follow-up questionnaire and submitted a stool sample at the completion of the study. Five participants who completed treatment were lost to follow-up and did not complete a follow-up questionnaire or undergo repeat fecal antigen testing. One participant refused to complete the follow-up questionnaire and provide a stool sample after completing triple therapy (Fig. 1).
Of the 165 H. pylori-positive participants who completed therapy and post-treatment testing, 148 were negative by fecal antigen testing at the completion of the study, for an eradication efficacy of 89.7% (Fig. 6A). 17 participants (10.3%) were still positive, and these participants were provided with quadruple therapy (see Methods). Triple therapy resulted in a significant improvement in dyspepsia severity, decreasing from a mean dyspepsia score of 8.8 before therapy to a mean score of 1.7 after therapy (Fig. 6B; p < 0.0001). Of various risk factors analyzed, only participants who missed four or more doses of their medications during the course of the triple therapy regimen were significantly more likely to fail H. pylori eradication (p = 0.0415; Table 3).
Table 3
Factors associated with Helicobacter pylori eradication.
Characteristic | Fecal antigen negative (n = 148) | Fecal antigen positive (n = 17) | P value |
Female gender, n (%) | 103 (70) | 9 (53) | 0.3981 |
Level of education, n No education Primary Secondary Tertiary | 38 74 34 2 | 3 11 2 1 | 0.4342 |
Alcohol use, n None 1 day per week or less 2 days per week 3 days per week 4 days per week or more | 112 18 5 3 10 | 16 1 | 0.4185 |
NSAID use, n (%) | 42 (28) | 2 (12) | 0.1415 |
Use of other antibiotics, n (%) Yes No Don’t know | 37 (25) 80 (54) 31 (21) | 5 (29) 6 (35) 6 (35) | 0.6090 |
Experienced new symptoms during treatment, n (%) Yes No Don’t know | 84 (57) 63 (42) 1 (1) | 11 (65) 6 (35) | 0.9384 |
Number of missed doses, n (%) 0–4 More than 4 | 146 (99) 2 (1) | 15 (88) 2 (12) | 0.0415 |