4.1 Discussion
This study examined the use of proton pump inhibitors (PPIs) in CHS patients aged 50–75 years in the Haifa and Western Galilee districts between 2005 and 2020. Compared with the control group, the study group was older, had lower GFR values, and had a greater incidence of comorbidities such as diabetes, hypertension, and cardiovascular disease. The most common PPI was omeprazole, and the prevalence of dementia and mortality were highest in the study group. The unadjusted Cox model also showed an increased risk of dementia for the first control (occasional users) but not for the second control (one-time users), and after adjusting for baseline diagnoses, the risk of dementia remained significant for the first control but not for the second control. Additional analysis showed that when standardized for each variable separately, the direction of the association remained consistent except for age at diagnosis. A subgroup analysis of relatively healthy patients showed similar results.
Our results are similar to those published recently by Friesen et al., who suggested PPI use as a marker for poor health status19. We similarly assumed that the effect of PPIs was cumulative and nonreversible, as dementia is a neurodegenerative condition. Interestingly, a recent large meta-analysis provided no clear evidence of an association between PPI intake and the risk of dementia, mainly due to discrepancies in sensitivity analyses. However, the risk of dementia due to PPI use cannot be ruled out20.
The higher mortality in control 2 than in control 1 was an unexpected finding. Rare PPI use may be a proxy for poor baseline health or a sign of neglect since individuals who do not see physicians frequently are less likely to take any medication, as evidenced by concurrent lower antidepressant and benzodiazepine use.
This study has a significant advantage in terms of its long follow-up period, with a median of nearly ten years and a maximum of 15 years. The population in this study was a large sample of HMO patients. Every citizen in Israel is insured by one of the four HMOs, and the population of CHS represents the general population in Israel. Additionally, this retrospective study used comprehensive data extracted from a high-quality database that included data on medication dispensation. The study population therefore represents real-world, high-quality data that represent the general population and is likely generalizable.
Limitations of the study:
A retrospective follow-up study can include many biases and differences in risk factors between the study and control groups. While our study population was drawn from a single health maintenance organization, its demographic characteristics were similar to those of other Israeli populations, suggesting that our findings may be generalizable to the broader Israeli population. Incomplete information, such as the undercoding of dementia diagnoses by doctors and incomplete data on medication purchases, is probable. For example, some patients may purchase their medication outside the HMO system due to a preference for original drugs that are sometimes unavailable in HMO pharmacies, which also do not provide low-dose omeprazole (10 mg), which is available as an OTC drug. Additional missing data on educational status, exercise, ethnicity, and apolipoprotein E status may have also affected the results. However, the E value, which is defined as the minimum strength of association, on the risk ratio scale indicates that an unmeasured confounder would need to have with both the treatment and the outcome to fully explain a specific treatment–outcome association, conditional on the measured covariates, was very high (1.93).
We were unable to control for potential confounding factors such as diet, physical activity, or access to healthcare, which may have influenced the association between PPI use and dementia risk.
This study examines the dispensing of prescribed PPIs and does not include OTC dispensing of PPIs (omeprazole is available in small OTC packages). In addition, medication purchases do not always reflect actual compliance. We were also unable to distinguish between AD and non-AD dementia; AD accounted for 50–70% of all dementia cases, followed by vascular dementia at 15%.
Notably, the mortality rate in the study group was greater than that in the control 1 group (lower quartile), probably because of the clinical background. However, mortality in control 2 (one-time users) was also greater than that in control 1.
The number of patients who took non-omeprazole PPIs was minimal, and it was not possible to compare the different types of PPIs.
Future studies should employ prospective designs, include more diverse Israeli populations, and collect data on potential confounding factors such as lifestyle and socioeconomic status. Additionally, exploring the biological mechanisms underlying the association between PPI use and dementia risk could help to clarify the nature of this relationship.
Implications of our findings for health policy and clinical practice in Israel:
The association between long-term PPI use and increased dementia risk observed in our study has important implications for health policy and clinical practice in Israel. Our findings highlight the need for careful consideration and monitoring of long-term PPI prescribing, especially in older adults. Prescribing guidelines may need to be updated to reflect the potential risks associated with prolonged PPI use, and healthcare providers should be educated about these risks to ensure appropriate prescribing practices21. Health policymakers should consider implementing strategies to monitor and optimize PPI use, such as using PPI long-term prescriptions as a proxy to locate at-risk individuals and offer a tailored approach to reduce the risk of dementia and mortality. By addressing the potential risks of long-term PPI use at both the individual and population levels, we can work towards improving the safety and quality of healthcare for older adults in Israel.
4.2 Conclusion
In conclusion, our large, long-term retrospective cohort study revealed a significant association between long-term PPI use and increased dementia risk in an Israeli population. These findings have important implications for health policy and clinical practice in Israel and highlight the need for careful consideration and monitoring of long-term PPI use, especially in older adults. Our study adds to the growing body of evidence on the potential risks of prolonged PPI use and underscores the importance of further research to better understand this relationship. We suggest that healthcare providers and policymakers consider using long-term PPI prescriptions as a proxy to identify individuals at higher risk of dementia and mortality, enabling the development of tailored interventions to mitigate these risks. By addressing the potential consequences of prolonged PPI use at both the individual and population levels, we can work towards improving the safety and quality of healthcare for older adults in Israel.