Making sense of the evidence
As someone who has Barrett’s and reflux symptoms, I decided to approach this review as a patient looking online rather than as a doctor. But I found it was harder than I expected to give the answers that patients – like me – might be looking for on proton pump inhibitors (PPIs) such as omeprazole, esomeprazole and lansoprazole.
The need-to-knows
I suspect that everyone who reads this will have been given PPIs and will probably know there are a lot of questions about them.
To try and find some answers, I have picked a number of websites to gather current opinions. In particular, I have looked at NICE, NHS, Patient Info, Welsh medical guidelines and the National Library of Medicine, an official American website.
Because there are so many things to consider, this has become quite a long article. I hope you will have time to read it all. But, just in case you don’t, the main things you should know are:
- PPIs are some of the most prescribed medicines, and we do have a lot of information on these pills and how good they are, as well as about their common side effects.
- They are very effective at reducing acid and symptoms of GORD (gastro-oesophageal reflux disease) and most people have few side effects. It’s more difficult to assess longer-term effects or rarer side effects.
- Some of the evidence is disputed, lots of studies and reviews conclude that the evidence is poor and that we need better studies.
- Some of the points I mention should give you things to discuss with you GP or specialist. It’s not designed to give you individual advice or all the answers.
- I have taken PPIs for a long time and – at the moment- I plan to carry on.
The full facts and figures
The things I cover:
- Proven benefits
- Possible benefits
- Common side effects
- Proven risks
- Less common side effects
- Possible long-term risks
One anecdotal story I want to share is that in 2004 I attended a two-day conference for the launch of the new British Hypertension Society guidelines for treating high blood pressure.
The long-awaited guidelines, meant to be followed by all GPs, were launched to great anticipation and were fine. But then came the talks from different members of the committee, the good and great of high blood pressure care in the UK, who were mainly university professors. They all proceeded to say why they disagreed with the evidence and what the studies showed, but all in different ways. There was no consensus. I found it interesting how different, respected academics could interpret large studies in such different ways; some saying it was a good study and proved something useful, while others said the same study was too poor to be much worth.
This is what I have found while researching PPIs. Bear in mind it is also 45+ years since I did academic reviews of scientific papers and fully understood statistics!
You will see that reputable sources give slightly different perspectives. For instance, NICE suggests an increase in fractures is proven, but other sources are less convinced.
Proven benefits
- PPIs are very good at reducing acid in the stomach and, therefore, good at symptom control.
Unproven benefits
- There is no good evidence that different PPIs have any significant benefits over others, although they do appear to have slightly different side effects. For example, some are more likely to cause diarrhoea than others, and they have different recorded drug interactions.
- There is no good evidence that PPIs reduce progression to cancer or dysplasia (a potential pre-cancerous condition). This includes studies which used aspirin as well.
- There are theoretical reasons why acid suppression should reduce the risk of progression. This is mainly because it reduces inflammation which is a general risk for developing cancer. Also, some studies have shown that regular use is probably better than intermittent use.
- In a review from 2019 by Professor Fitzgerald, Director of Early Cancer Institute, and Heartburn Cancer UK trustee, she commented on a study of 2000 patients which showed a 71% reduction in risk.
- But to put this into context, if the risk of developing cancer with significant Barrett’s is 1 in 1000 per year. Then with 2000 patients taking no treatment, two will develop cancer. If taking a PPI gave a 100% reduction in risk, then in this scenario, only one would get cancer. So statistically, for every 2000 people taking PPIs, one person would be saved from getting cancer each year.
- This shows that when we’re talking about relative risk reductions, if something is very rare, even what appears to be a very significant risk reduction does not actually impact many people.
Common side effects
I have taken information on side effects straight from the BNF (British National Formulary), the national guidelines for all prescribing in the UK and the one all doctors refer to (although they don’t give actual figures on what the phrases ‘common’ or ‘rare’ actually mean).
From the BNF:
Side-effects For all proton pump inhibitors
Common or very common
Abdominal pain; constipation; diarrhoea; dizziness; dry mouth; gastrointestinal disorders; headache; insomnia; nausea; skin reactions; vomiting
Uncommon
Arthralgia (joint pain); bone fractures; confusion; depression; drowsiness; malaise; myalgia (muscle aches); paraesthesia (numbness, pins and needles); peripheral oedema (swollen legs, fluid retention); thrombocytopenia and leucopoenia (low blood cells); vertigo; vision disorders
Rare or very rare
Alopecia (hair loss); gynaecomastia (breast enlargement in men); hallucination; hepatic disorders (liver); hyperhidrosis (sweating); hyponatraemia (low sodium in the blood); nephritis tubulointerstitial (kidney problems); pancytopenia and agranulocytosis (blood problems); photosensitivity reaction; severe cutaneous adverse reactions (SCARs)(severe skin rashes some made worse by the sun); stomatitis (sore mouth); taste altered
Frequency not known
Gastrointestinal infection; hypomagnesaemia (low magnesium in the blood more common after one year of treatment, but sometimes after three months of treatment); subacute cutaneous lupus erythematosus (rash)
Monitoring requirements
Measurement of serum-magnesium concentrations should be considered before and during prolonged treatment with a proton pump inhibitor, especially when used with other drugs that cause hypomagnesaemia or with digoxin
As you can see, it’s an extensive and worrying set of issues when all written down in one place, but most people get few side effects.
The BNF also has a list of cautions.
Can increase the risk of fractures (particularly when used at high doses for over a year in the elderly); may increase the risk of gastrointestinal infections (including Clostridium difficile infection); may mask the symptoms of gastric cancer; may reduce absorption of vitamin B12 with long-term treatment; patients at risk of osteoporosis.
Risk of osteoporosis
Patients at risk of osteoporosis should maintain an adequate intake of calcium and vitamin D, and if necessary, receive other preventative therapy.
Gastric cancer
Particular care is required in those presenting with ‘alarm features’. In such cases, gastric malignancy should be ruled out before treatment.
Cancer referrals
What they mean by ‘alarm symptoms’ are ones which would get a 2-week-wait cancer referral, which are over 55 years with weight loss and pain, a palpable lump in the tummy or difficulty swallowing. Other symptoms such as vomiting blood, nausea, new onset reflux, low blood count, raised platelet count or nausea and vomiting are also included as needing proper assessment.
For details, see NICE guidance NG12. These are the guidelines for GPs covering all cancers and worrying symptoms.
As PPIs are so good at reducing symptoms, even if something similar is causing them, you don’t want doctors to treat it without properly ruling out a serious underlying cause or to just make someone feel better when actually they need more tests.
Drug interactions
When it comes to drug interactions, the best source is the BNF. Most of the interacting drugs that are listed are fairly uncommon. Most people will be under a specialist, and the drugs that interact cannot be bought over the counter. But PPIs can, without anyone asking any questions. Potentially the most serious could be methotrexate, a drug that is used to treat conditions such as rheumatoid arthritis, psoriasis and Crohn’s disease.
Another issue that concerns me is that the BNF list does not include an interaction with Metformin, a drug for diabetes, which has also been implicated in causing a low B12. Including this would encourage GPs to be more concerned about vitamin B12 levels in patients who take both metformin and a PPI.
The big silent risks and the evidence
- Low magnesium
- Infection
- Vitamin deficiencies
- Osteoporosis and fractures
- Dementia
- Cancer
- Kidney disease
Low magnesium
I think there is overwhelming consensus that this can happen in rare cases. It is serious and can cause fits and heart problems. Hence, the recommendation to monitor levels. This is in the BNF but may not be something all GPs are aware of. There is one study that shows 25% of those who get low magnesium do not respond to supplements and have to stop PPIs.
Infection
The acid in your stomach serves as a barrier against infections getting in through your stomach. The evidence is that reducing the acid increases the risk of some gut infections including C. difficile, a particularly unpleasant bowel infection causing severe diarrhoea. But it’s mainly caused by antibiotics.
There is also some evidence that there is an increased risk of pneumonia, possibly related to bugs coming up from your stomach and going into your lungs with the reflux. This may be more of a concern for older people and those with existing lung disease.
Vitamin deficiency
Gastric acid is needed to separate B12 in your diet before absorption, so low acid can cause deficiency. Deficiencies in Vitamin B12 appears to be rare. No one appears to recommend regular blood checks, but they should be considered in older people and those on some drugs such as Metformin. Low B12 would also show up if you have a normal blood count checking your haemoglobin.
There have also been reports of causing low iron, but these are rare and of unknown cause. PPIs may also reduce the absorption of calcium.
Osteoporosis and fractures
Many sources quote this as proven (BNF, MHRA review 2012), others point out they are mainly observational studies. Observational studies show links but not cause.
The general consensus is to take other risk factors into consideration when prescribing.
Dementia
A Danish study (1) followed two million people aged 65-70 for, on average, ten years and looked at subjects developing dementia and their past or present PPI use. The study ran 2000-2018. This was first published Oct 2023. They conclude that PPI use in midlife has more risk than later life and that the study has limitations and needs more research.
A good Portuguese review (2) in March 2023 (before the Danish study) concluded that there was no statistically significant association in previous studies. As does the 2023 American review (3). There is no mention in the BNF (4), which is updated regularly.
Cancer
There are theoretical reasons to suggest using PPIs could be associated with an increased risk of gastric cancer related to gastrin levels or H. pylori infections. But, again, is this causative? Does PPI use cause this? The British Society of Gastroenterologists 2019 (5) and the American review appear to feel that the link is tenuous.
Kidney disease
In 2017, the European Medicines Agency did a review and suggested there was no evidence that needed a change in guidance for PPIs and kidney failure.
Heart disease
Again, associations have been suggested but the evidence is weak.
My overall thoughts
- There’s poor evidence that PPIs reduce Barrett’s progression at any dose, but there are theoretical reasons to suggest benefits.
- For the general use of PPI, it is recommended that you want good control of symptoms at the lowest possible dose.
- It’s unlikely that there will be good quality studies in near future to prove anything – one way or the other – because these drugs are so highly used it would be hard to design a trial.
- Many of the studies are observational and it’s hard to know if one issue causes the other. You can Google ‘casual v causal’ and get very good articles from CRUK to Wikipedia that explain different trials.
- There’s limited evidence, from large studies looking at people on these drugs for a long period, for concerns such as Alzheimer’s, cancer, heart disease, kidney failure and fractures.
- There is definite evidence for magnesium deficiency and possibly B12 deficiency and infections related to long-term PPI use. Therefore, there’s good evidence for monitoring magnesium and possibly B12 too.
- People should be assessed for osteoporosis and fracture risks and action should be taken where needed.
- Doctors and patients should be mindful of all proven or possible rare side effects, such as severe skin reactions, including SLE, nephritis, low blood cell counts.
- Adding it all up, I am still taking my PPIs.
Sources
I have added details of the articles I have used. There were many I did not use and a few which I felt were of dubious accuracy and balance when it came to their conclusions.
The following have been numbered in the blog:
1) Good USA based site, reviewed the Danish study.
2) Review by a Portuguese team of all recent studies. Published in March 2023, it did not include the Danish study.
3) 2023 USA Government site recommendations. This appears to me to be a good, balanced up-to-date account, link here.
4) NICE links to the BNF, the British National Formulary, which is the place GPs would look for advice, read here.
5) The British Society of Gastroenterology did a statement in 2019, read here.
One of the latest reviews was NICE in 2023, link here.
Review by Prof Fitzgerald and Corley in 2019 (Rebecca Fitzgerald is a trustee of HCUK).
This is taken from Patient, a UK website with good quality medical input.
I found a good review from 2023 which reviewed 11 studies which they regarded as the most relevant from the last 10 years, read here.
An Italian study did a review of stomach cancer studies (narrowed down to 14 studies ) in Dec 2022, read here.
Their conclusions were:
- A clinical practice update of the American Gastroenterological Association suggested recently that “the decision to discontinue PPIs should be based solely on the lack of an indication for use, and not because of concern for PPI-associated adverse effects”.
- Our findings further support this statement. Although the association between PPI use and gastric cancer is one of the most debated concerns of PPI use in the medical literature, in this comprehensive systematic review with meta-analysis we did not find evidence of an association.
- Our findings are reassuring, especially to all those patients who have an indication for long-term PPI use and need persistent and effective acid gastric suppression to prevent serious health consequences.
However, a Chinese study review suggested an increased link.
Their conclusions are:
- The findings of this updated meta-analysis suggest that the risk of gastric cancer is increased in patients treated with PPIs compared with patients not taking PPIs. The findings are from a summary of observational studies, and PPIs have well-established clinical benefits for patients. Therefore, physicians should regularly assess the actual benefits for patients, especially patients with long-term PPI treatment for GERD. Our study findings also highlight the necessity for long-term randomised controlled trials to evaluate the risk of gastric cancer in patients treated with PPIs.
All Wales medicines strategy group, safe use of PPIs 2018, available here.
Cancer guidelines NICE NG12, link here.
