Background: Proton pump inhibitors (PPIs) and H2-receptor antagonists (H2RAs) are more frequently used in the treatment of acid-related gastrointestinal diseases. However, the comparative relationships with the severity of dyspepsia and Vitamin B12 status in long-term treatment are not yet clear. Objective: To compare dyspepsia severity and serum vitamin B12 status among patients receiving long-term PPIs and H2RAs. Methods: This analytical cross-sectional study was conducted at Sheikh Zayed Hospital, Lahore, Pakistan from 1st November 2025 to 30th April 2026. It was assumed that a total of 308 patients were included in the study, 154 of whom had taken PPIs and 154 had taken H2RAs. The severity of dyspepsia was determined using the Leeds Dyspepsia Questionnaire, and serum B12 levels were assessed by immunoassay. The statistical analysis was performed by group comparison, correlation analysis, and multivariable regression. Results: PPI users had significantly lower scores for dyspepsia compared to H2RA users (12 versus 15, p = 0.004). PPI users also had lower median vitamin B12 levels (258 pg/mL compared to 322 pg/mL, p<0.001). Vitamin B12 deficiency was more frequent in the PPI group (35.1% versus 20.1%, p=0.003). Adjusted use of PPI was linked with increased odds of vitamin B12 deficiency (adjusted OR = 2.02; 95% CI: 1.20-3.40). Conclusions: There were no significant differences detected in the severity of dyspepsia between the long-term PPI users and H2RA users, but there was a significant difference detected in vitamin B12 status.
Dyspepsia is a frequent gastrointestinal condition (FGC) with abdominal discomfort, pain or burning, and early satiety and postprandial fullness, and is a significant driver of healthcare resource utilization and sub-optimal quality of life.[1] It can occur as a complication of functional gastrointestinal disorders and/or organic disorders of the gastrointestinal tract such as peptic ulcer disease or gastroesophageal reflux disease (GERD).[2] Anti-gastric medications are often used to treat upper gastrointestinal symptoms, such as proton pump inhibitors (PPIs) and histamine-2 receptor antagonists (H2RAs).[3] PPIs typically offer a longer duration of acid suppression than do H2RAs and are used extensively in acid-related
gastrointestinal diseases.[4] Increased duration of therapy, however, poses significant questions concerning duration of dyspeptic symptoms, nutritional implications, and the need for ongoing therapy.[5]
In 2024, a systematic review and meta-analysis of 256,915 patients across 40 countries found the overall global functional dyspepsia prevalence rate to be 8.4% (95% CI: 7.4-9.5%).[6] Its prevalence was slightly higher in developing countries than in developed countries (9.1% versus 8.0%), indicating that it has a significant burden in various healthcare settings.[6] PPIs are still widely used in treating acid-related gastrointestinal disorders, with about 7-15% of people reported to be taking them at a particular time and almost one in four reported to be using them for at least 1 year.[7] Presently, there are indications that PPIs are more effective than H2RAs in healing and maintaining remission of erosive esophagitis.[8] However, if the upper gastrointestinal symptoms do not improve following long-term therapy, it indicates that the mechanism of dyspepsia is not entirely acid suppression.
A key issue related to long-term acid suppression is its effect on B12 absorption.[9] The ability to release protein-bound, acid-labile forms of vitamin B12 from food is helped by gastric acid, and prolonged acid suppression can disrupt this process.[10] In 2023, a systematic review and meta-analysis, which included 25 studies, showed an increased risk for vitamin B12 deficiency in PPI users compared with non-users (OR: 1.42; 95% CI: 1.16-1.73), but this association was weak due to significant heterogeneity.[11] A 2025 meta-analysis, however, did not reveal any significant difference between the serum vitamin B12 or homocysteine levels of chronic PPI users and controls.[12] The conflicting results suggest that the association between long-term acid suppression and vitamin B12 status is not fully understood, especially given the variable duration of acid suppression, vitamin B12 dosage, nutritional intake, and patient demographics.
An important clinical question, however, that has not been fully answered is whether or not there is any benefit of increased acid suppression in terms of better symptom control with no detriment to nutritional status over the long term. Studies have primarily focused on gastrointestinal symptom management and vitamin B12 deficiency as individual outcomes, and there is limited evidence to directly assess both outcomes in PPI versus H2RA users. Measuring these dimensions together may give a more complete picture of the clinical profile of long-term acid suppression and help to identify patients that might need to be subjected to closer therapeutic reassessment or nutritional evaluation. Thus, the goal of this study was to determine differences in the relationship between long-term proton pump inhibitor (PPI) use and serum vitamin B12 levels or symptoms of dyspepsia as compared to H2-receptor antagonist (H2RA) use in patients with long-term acid-suppressive therapy. The results could help inform more personalized prescribing by elucidating the distinct symptom burden and vitamin B12 status of the two treatment groups.
This was an analytical cross-sectional study conducted at Sheikh Zayed Hospital, Lahore, Pakistan, between November 1st, 2025, and April 30th, 2026. This study aimed to evaluate the severity of dyspepsia and serum vitamin B12 levels in long-term proton pump inhibitor (PPI) users vs. long-term histamine-2 receptor antagonist (H2RA) users. They were classified as either above or below the threshold of predominant acid-suppressive medication use. Patients treated with PPIs were in Group I, and patients treated with H2RAs were in Group II.
The OpenEpi sample size framework was used to calculate the sample size to compare two independent proportions with a 95% confidence level, 80% statistical power, a 2-sided significance level of 5%, and an allocation ratio of 1:1. A previous systematic review by the authors (Vitamin B12 Deficiency and Use of Proton Pump Inhibitors, 2023) identified an increased odds ratio for vitamin B12 deficiency in PPI users (1.42; 95% CI: 1.16-1.73).[13]
The proportions of vitamin B12 deficiency were assumed to be 35% for PPI users and 20% for H2RA users, giving an expected absolute difference of 15 percentage points for the proposed calculation. The minimum sample size for each group determined by the Fleiss formula without continuity correction was approximately 138 participants, for a total of 276 participants. After allowing for a 10% proportion of incomplete or unusable records, the target sample size was increased to 308 participants, with 154 participants allocated to each group.
The patients who attended the outpatient departments of the study institution during the study period were recruited using a non-probability consecutive sampling technique. Subjects who met the inclusion criteria were approached sequentially until the desired number was obtained. The participants were categorized based on the long-term documented PPI or H2RA use.
The study sample comprised patients aged 18 to 70 years, both male and female, who had been on maintenance therapy for at least six months using either a PPI or H2RA. Patients with documented acid-related upper gastrointestinal symptoms (dyspepsia, GERD, epigastric discomfort) who were willing to sign an informed consent and have serum vitamin B12 levels performed were eligible for the study.
Patients using both PPI and H2RA at the same time or who had switched between the two classes of drugs in the previous six months were excluded. Vitamin B12-deficient patients, those with a known diagnosis of pernicious anaemia, atrophic gastritis, inflammatory bowel disease, coeliac disease, previous gastric or ileal surgery, or other known malabsorptive conditions were also excluded. Individuals receiving vitamin B12 supplementation or medications known to interfere with vitamin B12 absorption, including metformin, were excluded to minimise potential confounding. Patients with severe systemic illness, patients with incomplete clinical records, and pregnant or lactating women were not included.
After obtaining institutional ethical approval and informed written consent, eligible subjects were selected and enrolled in this study on a structured data collection form. Demographic data such as age, sex, body mass index (BMI), smoking history, medical history, and dietary information were collected. Acid-suppressive therapy information was gathered during patient interviews and from prescription records, including the class of the medication, specific drug, dose per day, number of times a day the medication was administered, and duration of treatment. The subjects were then randomly allocated to the PPI group or H2RA group based on their medication history.
Dyspepsia severity was estimated by the validated Leeds Dyspepsia Questionnaire (LDQ) that measured the severity and frequency of the upper gastrointestinal symptoms of epigastric pain, heartburn, regurgitation, nausea, and associated discomfort.[14] The questionnaire scores were computed as per the instructions given in the scoring of the instrument, and higher scores were reflective of the higher symptom burden. In order to be consistent in the assessment of symptoms, the same assessment procedure was used in both groups of treatment and both groups of control.
Venous blood, about 3-5 mL, was collected from each of the participants under sterile conditions for biochemical assessment. The blood was collected in the proper collection tubes, centrifuged, and analysed in the institutional laboratory for vitamin B12 concentration by immunoassay method. The serum vitamin B12 levels were obtained in pg/mL. The parameters for vitamin B12 deficiency were: <200 pg/mL, borderline: 200-300 pg/mL, and normal: >300 pg/mL. These were used to compare the status of vitamin B12 in the two treatment groups. All of the data collected were recorded on a standardised proforma, and participant confidentiality was ensured throughout the study.
All data were entered and analysed in IBM SPSS Statistics version 26.0. Continuous variables were evaluated for normality by the Shapiro-Wilk test and graphical analysis. The mean ± standard deviation was used for normally distributed quantitative variables such as age, serum vitamin B12 concentration, and dyspepsia score. Frequencies and percentages were used to report the categorical variables such as sex, medication category, duration of treatment, and vitamin B12 status. An independent-samples t-test was used to compare two normally distributed quantitative variables when comparing PPI users and H2RA users. Categorical variables such as prevalence of vitamin B12 deficiency were compared between the two groups using the chi-square test. Multivariable linear regression was applied to assess the relationship between medication class and the score of dyspepsia and serum Vitamin B12 levels, adjusting for other confounding factors such as diet, age, sex, and treatment duration. The adjusted odds ratios (AORs) with 95% confidence intervals (CIs) for vitamin B12 deficiency in PPI users vs H2RA users were calculated using multivariable logistic regression. A p-value of < 0.05 was considered statistically significant.
A total of 308 patients were included in the analysis, with 154 participants in each treatment group. The mean age of proton pump inhibitor (PPI) users was 44.8 ± 12.6 years, while that of H2-receptor antagonist (H2RA) users was 43.2 ± 13.1 years. There were no statistically significant differences (p > 0.05) between the two groups in terms of age, sex, BMI, smoking, diet, or gastrointestinal history (Table 1).
PPI users had a significantly longer duration of treatment compared with H2RA users (18 months versus 15 months, p = 0.018). Frequency of once-daily medication administration was higher in the PPI group compared to the control group (83.1% vs. 67.5%, p = 0.001). Omeprazole was the most commonly used PPI (41.6%), whereas famotidine predominated among H2RA users (72.7%) (Table 2).
In addition, PPI users had a significantly lower level of severity of dyspepsia, with a median Leeds Dyspepsia Questionnaire score of 12 (IQR: 8-17), than did H2RA users (15, IQR: 10-20, p = 0.004). Individual symptoms (epigastric pain, heartburn, regurgitation, nausea, postprandial fullness, and early satiety) tended to be less common in the PPI group, but there were no significant differences (Table 3).
The median serum B12 was significantly lower in PPI users than in H2RA users (258 vs 322 pg/mL, p<0.001). A higher rate of vitamin B12 deficiency was seen in PPI users (35.1%) than in H2RA users (20.1%) (p = 0.003). Likewise, there was significantly more combined deficient and borderline vitamin B12 status in the PPI group (66.2% vs 48.1%, p = 0.001) (Table 4).
The serum vitamin B12 level in PPI users was negatively correlated with treatment duration (r = -0.29; p<0.001) by correlation analysis. The duration of treatment was weakly positively correlated with dyspepsia severity in H2RA users (r = 0.16, p = 0.047). There was no significant correlation between serum vitamin B12 level and the severity of dyspepsia in both groups (Table 5).
The odds of vitamin B12 deficiency were about twofold higher in those who used PPI than those who used H2RAs after adjustment for potential confounders (adjusted OR = 2.02, 95% CI: 1.20-3.40, p = 0.008). The use of PPI was also independently negatively associated with vitamin B12 levels (β = -52.4, p < 0.001) and dyspepsia scores (β = -2.10, p = 0.004). Furthermore, there were 18% higher adjusted odds of vitamin B12 deficiency with every 12-month increment in treatment duration (Table 6).
Table 1. Demographic and clinical characteristics
|
Variable |
PPI (n=154) |
H2RA (n=154) |
p-value |
|
Age, years, mean ±SD |
44.8 ± 12.6 |
43.2 ± 13.1 |
0.275 |
|
18-39 years, n (%) |
56 (36.4) |
62 (40.3) |
0.483 |
|
40-59 years, n (%) |
75 (48.7) |
72 (46.8) |
|
|
60-70 years, n (%) |
23 (14.9) |
20 (13.0) |
|
|
Male, n (%) |
83 (53.9) |
79 (51.3) |
0.648 |
|
Female, n (%) |
71 (46.1) |
75 (48.7) |
|
|
BMI, kg/m², mean ± SD |
26.4 ± 3.8 |
25.9 ± 3.7 |
0.244 |
|
Current smoking, n (%) |
35 (22.7) |
31 (20.1) |
0.579 |
|
Predominantly vegetarian diet, n (%) |
28 (18.2) |
32 (20.8) |
0.565 |
|
History of GERD, n (%) |
91 (59.1) |
84 (54.5) |
0.418 |
|
History of dyspepsia, n (%) |
117 (76.0) |
121 (78.6) |
0.584 |
Table 2. Acid-suppressive medication exposure
|
Variable |
PPI (n=154) |
H2RA (n=154) |
p-value |
|
Treatment duration, months, median (IQR) |
18 (12-30) |
15 (10-24) |
0.018 |
|
6-12 months, n (%) |
47 (30.5) |
61 (39.6) |
0.166 |
|
>12-24 months, n (%) |
59 (38.3) |
55 (35.7) |
|
|
>24 months, n (%) |
48 (31.2) |
38 (24.7) |
|
|
Once-daily administration, n (%) |
128 (83.1) |
104 (67.5) |
0.001 |
|
Twice-daily administration, n (%) |
26 (16.9) |
50 (32.5) |
|
|
Omeprazole, n (%) |
64 (41.6) |
Not applicable |
|
|
Pantoprazole, n (%) |
57 (37.0) |
Not applicable |
|
|
Esomeprazole, n (%) |
33 (21.4) |
Not applicable |
|
|
Famotidine, n (%) |
Not applicable |
112 (72.7) |
|
|
Other H2RA, n (%) |
Not applicable |
42 (27.3) |
|
Table 3. Dyspepsia symptom profile and severity
|
Variable |
PPI (n=154) |
H2RA (n=154) |
p-value |
|
LDQ total score, median (IQR) |
12 (8-17) |
15 (10-20) |
0.004 |
|
Epigastric pain, n (%) |
92 (59.7) |
106 (68.8) |
0.096 |
|
Heartburn, n (%) |
78 (50.6) |
91 (59.1) |
0.135 |
|
Regurgitation, n (%) |
66 (42.9) |
77 (50.0) |
0.209 |
|
Nausea, n (%) |
49 (31.8) |
58 (37.7) |
0.280 |
|
Postprandial fullness, n (%) |
103 (66.9) |
110 (71.4) |
0.390 |
|
Early satiety, n (%) |
71 (46.1) |
76 (49.4) |
0.568 |
Table 4. Serum vitamin B12 findings
|
Variable |
PPI (n=154) |
H2RA (n=154) |
p-value |
|
Serum vitamin B12, pg/mL, median (IQR) |
258 (190-344) |
322 (246-411) |
<0.001 |
|
Deficient (<200 pg/mL), n (%) |
54 (35.1) |
31 (20.1) |
0.003 |
|
Borderline (200-300 pg/mL), n (%) |
48 (31.2) |
43 (27.9) |
|
|
Normal (>300 pg/mL), n (%) |
52 (33.8) |
80 (51.9) |
|
|
Deficient or borderline, n (%) |
102 (66.2) |
74 (48.1) |
0.001 |
Table 5. Correlations with treatment duration and B12
|
Relationship |
PPI: r (p) |
H2RA: r (p) |
|
Treatment duration vs. serum B12 |
-0.29 (<0.001) |
-0.15 (0.063) |
|
Treatment duration vs. LDQ score |
0.12 (0.139) |
0.16 (0.047) |
|
Serum B12 vs. LDQ score |
-0.10 (0.217) |
-0.08 (0.324) |
Table 6. Adjusted regression models
|
Outcome and comparison |
Adjusted estimate (95% CI) |
p-value |
|
Vitamin B12 deficiency: PPI vs. H2RA |
aOR 2.02 (1.20-3.40) |
0.008 |
|
Serum B12 (pg/mL): PPI vs. H2RA |
β -52.4 (-76.5 to -28.3) |
<0.001 |
|
LDQ score: PPI vs. H2RA |
β -2.10 (-3.52 to -0.68) |
0.004 |
|
Vitamin B12 deficiency: duration, per 12 months |
aOR 1.18 (1.04-1.34) |
0.011 |
The present analysis assessed the severity of dyspepsia and vitamin B12 levels in patients on long-term proton pump inhibitors (PPIs) versus H2-receptor antagonists (H2RAs). The severity scores of dyspepsia were significantly lower in PPI users, and there was a significantly lower level of vitamin B12 in the blood of PPI users compared to H2RA users, and a higher prevalence of vitamin B12 deficiency among PPI users compared to H2RA users. These results indicated that there were differences in symptoms and nutritional status between the treatment groups. The cross-sectional design, however, did not demonstrate that these differences were direct consequences of medication exposure, as the associations may have been attributable to other factors related to the patient, including baseline symptom severity, medication choice, diet, and other factors.
The magnitude of dyspepsia severity difference between the two groups was consistent with what was seen in the literature and consistent with the effectiveness of acid-suppressive drugs. In the Japanese Society of Gastroenterology guidelines on FD published in 2022, both PPI and H2RAs were effective treatments for FD, but efficacy on symptom improvement was not consistent.[15] Similarly, Black et al. (2022), in the British Society of Gastroenterology guidelines, reported that PPIs and H2RAs were more effective than placebo, with relative risks of persistent symptoms of 0.88 and 0.79, respectively.[16] Results were consistent with the clinical significance of both classes of drugs, but failed to demonstrate a consistent superiority of PPI over H2RA. Hence, the negative associations between PPI use and dyspepsia scores found in the current analysis should be interpreted with caution, especially due to the lack of pretreatment symptom scores.
In the present analysis, there was a greater prevalence of deficiency in vitamin B12 status in PPI users compared to H2RA users (35.1% vs. 20.1%). In a cohort study of 1225 patients in Islamabad, Mumtaz et al. (2022) also found that long-term PPI use was linked to reduced vitamin B12 levels, and differences were noted between users of omeprazole and pantoprazole.[17] A systematic review published in 2022 found 7 studies with a total of 48,108 participants that demonstrated decreased vitamin B12 levels in the blood of people taking PPIs for a long time.[18]. There was no consistency in the individual study results. The results were consistent with the current findings in general, although some differences in the study subjects, medication exposure, and laboratory measurements made direct comparisons of deficiency prevalence less feasible.
In a systematic review and meta-analysis published in 2023, which included 25 studies, PPI users had higher odds of vitamin B12 deficiency compared to non-users (OR=1.42; 95% CI: 1.16-1.73).[13] This adjusted association seen in the present analysis (OR = 2.02; 95% CI: 1.20-3.40) was somewhat in the same direction, but stronger in magnitude. There was, however, significant heterogeneity, and the available evidence was not conclusive. In addition, the groups were different in that the group of PPIs was directly compared with the group of H2RAs, not with people who were not taking PPIs. For that reason, the estimates should not be considered equivalent on a direct basis. On the other hand, Shahid et al. (2025), in a systematic review of 693 older adults taking multiple drugs, found that PPI use resulted in a mean decrease of 12-18% in serum vitamin B12 levels after 12 months of treatment.[19] The contrasting findings suggested that the association of long-term PPI use with B12 status was unclear. This variation may have been partly due to differences between the age of the participants, nutritional status, duration of exposure, and biochemical definitions.
A more relevant comparison was given by Jamal et al. (2026), who studied long-term use of PPI and H2RA in Karachi.[20] The authors reported adjusted odds ratios of 2.156 (95% CI: 0.826-5.626) for PPI exposure and 1.264 (95% CI: 0.319-5.015) for H2RA exposure in relation to vitamin B12 deficiency. None of the groups obtained statistical significance. The results did not reveal any significant differences between the classes of medication, despite the estimates indicating a potentially greater relationship with PPI use.[20] The present results were also statistically significant, but study design, choice of reference groups, definition of exposure, and precision of estimates limited direct comparisons.
The negative association between treatment duration and vitamin B12 concentration in PPI users in the current analysis indicated a relationship where longer duration of PPI use was linked to lower vitamin B12 levels. In a 2026 retrospective cohort study from Saudi Arabia, blood levels of vitamin B12 decreased from a median of 312 pg/mL to 297 pg/mL, and the prevalence of deficiency rose from 8.2% to 12.2%.[21] The investigators, however, found no significant link between changes in B12 and length of treatment. Similarly, a retrospective cross-sectional study published in 2026 involving 255 long-term PPI users reported differences in vitamin B12 deficiency according to PPI type, with deficiency observed in 38.6% of omeprazole users and 59.0% of esomeprazole users.[22] These results indicated the possibility that there could be some differences in medications that need to be explored further, but neither study showed that the vitamin B12 deficiency was caused by the length of treatment alone.
There were a few limitations in the present analysis. The cross-sectional design of the study limited generalisability and the ability to examine temporal or causal relationships. Selection bias could have been introduced in consecutive sampling. There were no baseline dyspepsia scores or vitamin B12 levels to assess for changes after treatment was initiated. Multivariable analyses included some selected confounder factors, but dietary intake, underlying gastrointestinal disease, medication adherence, and indications for prescription cannot be ruled out as potential residual confounders. Moreover, the vitamin B12 status was based on serum level only, without also considering the presence of specific biomarkers such as homocysteine or methylmalonic acid. There were a number of limitations, but the combination of evaluating dyspepsia severity and vitamin B12 status gave a clinically useful framework by which the long-term acid-suppressive treatment groups were comparable.
The findings indicated that long-term PPI users had lower scores for dyspepsia but a greater number of biochemical vitamin B12 deficiency than did patients taking H2RAs. These observations were consistent with some published studies, but other studies had found weak or non-significant associations. The results thus led to continued future research with baseline measurements, standardised medication exposure assessment as well as more extensive assessment of vitamin B12 status, but did not provide any evidence of superiority or nutritional safety of either class of drugs in the long term.
LIMITATIONS
There were several limitations in the present study. The study was cross-sectional and carried out in one centre, restricting the generalisability of the results and the ability to make causal conclusions. Selection bias may have occurred due to the use of non-probability consecutive sampling. Baseline scores for dyspepsia and vitamin B12 were not available, and it was not possible to measure the effect of treatment after treatment was started. Potential confounding factors were taken into account, but the possibility of residual confounding due to dietary intake, medication adherence, and underlying gastrointestinal diseases could not be ruled out. In addition, only serum levels of vitamin B12 were measured and not any confirmatory markers like homocysteine or methylmalonic acid (MMA).
Long-term PPI use was associated with lower dyspepsia severity scores but reduced serum vitamin B12 concentrations and a higher prevalence of vitamin B12 deficiency compared with H2RA use. There was also a negative correlation between duration of treatment and vitamin B12 levels, especially among PPI users. This information indicated that there may have been a difference between the two acid-suppression drugs in terms of their clinical and nutritional characteristics. In patients with risk factors for vitamin B12 deficiency, periodic review of treatment need and assessment of vitamin B12 may be warranted.