Reflux and gastro-oesophageal reflux disease

What reflux is

Gastro-oesophageal reflux disease, or GORD, happens when acid and other stomach contents flow back up into the oesophagus and cause troublesome symptoms or damage. The usual symptoms are heartburn, a burning feeling behind the breastbone, and regurgitation of sour fluid into the throat or mouth. Some people get a chronic cough, hoarseness or a sense of something stuck in the throat instead.

It is common. In a population survey of 672 adults in Western Sydney, 12 per cent had heartburn or acid regurgitation at least once a week. A systematic review of population studies worldwide put the Australian figure at 11.6 per cent on a similar definition, against 18 to 28 per cent in North America and under 8 per cent in East Asia.

A weak valve at the lower end of the oesophagus, a hiatus hernia, weight carried around the abdomen, large or late meals and smoking all make reflux more likely. Most people are well controlled with simple changes and an acid-suppressing medicine, and never need to see a specialist at all.

When reflux needs a specialist

Reflux does not need a gastroscopy in everyone. It is worth being assessed when:

• symptoms persist despite a proper trial of treatment, or come back every time it stops
• food or tablets stick on the way down, or swallowing is painful
• there is weight loss, vomiting, iron deficiency, or black or bloody stools
• symptoms start for the first time later in life
• there has been reflux for many years, particularly with other risk factors for Barrett's oesophagus
• long-term acid suppression is being considered, and it would help to know whether it is really needed

Those first few are often called alarm features, and they are worth acting on. It is worth knowing what they are not. Across 15 studies and 57,363 patients, the sensitivity of alarm features for finding an upper gastrointestinal cancer ranged from 0 to 83 per cent. They are a reason to be seen, not a test that rules anything out, which is why persistent symptoms get investigated even without them.

Trouble swallowing is not the same as heartburn and should always be investigated. It can be caused by a narrowing, a ring, eosinophilic oesophagitis, a problem with how the oesophagus squeezes or occasionally something more serious.

Which tests are used, and what each one answers

Each test answers a different question, and the commonest mistake is expecting one of them to answer all three.

• Gastroscopy answers what the oesophagus looks like. It shows inflammation, ulcers, a narrowing, a hiatus hernia and Barrett's oesophagus, allows biopsies and picks up alternatives such as eosinophilic oesophagitis. What it does not answer is whether reflux is causing the symptoms. A normal gastroscopy is very common in reflux disease and does not rule it out.
• pH or pH-impedance monitoring answers whether acid is actually reaching the oesophagus, how often it happens and whether the symptoms line up in time with the reflux episodes. A thin catheter is passed through the nose and worn for 24 hours, or a small capsule is clipped to the oesophagus and records for 48 hours without a catheter. The capsule is not funded in Australia, so the catheter study is what is used here in practice.
• High resolution manometry answers how well the oesophagus squeezes and how the valve behaves. It is used to exclude achalasia and other motility disorders, and it is done before any anti-reflux operation.

Internationally agreed thresholds give these tests teeth. Severe oesophagitis, a peptic stricture, a long segment of Barrett's or acid exposure above 6 per cent of the day count as conclusive evidence of reflux disease. Evidence against it needs a normal gastroscopy together with acid exposure under 4 per cent and fewer than 40 reflux episodes off treatment. In between is a grey zone, which is where the symptom correlation matters.

How reflux is treated

Treatment is stepped, and the first step is not a tablet. Weight loss where it is relevant, smaller meals, leaving two to three hours between the evening meal and lying down, raising the head of the bed and stopping smoking all reduce reflux itself. What the evidence does and does not support is set out below.

Medicines you can buy

Antacids neutralise acid that is already there and work within minutes, which suits occasional breakthrough symptoms. Sodium alginate preparations, commonly known as Gaviscon, work differently: they form a raft on top of the stomach contents and physically get in the way of the acid pocket. They are safe, and the evidence for them is better than their over-the-counter status suggests. Pooling 14 randomised trials and 2,095 people, alginates resolved reflux symptoms substantially more often than a placebo or a plain antacid. Against a proton pump inhibitor they came out behind, though not by enough to be statistically certain.

Proton pump inhibitors

These are the most effective medicines for reflux. Several are available in Australia and all are PBS subsidised. Two practical points are worth more than the choice between them.

• Take it before food. These drugs only switch off pumps that are actively working, so a dose 30 to 60 minutes before a meal does more than the same dose taken at bedtime or on an empty stomach. This is the commonest fixable reason a proton pump inhibitor appears not to be working.
• Treat properly, then step down. The usual course is a full dose for about eight weeks, then a review, then the lowest dose that holds the symptoms. For reflux without significant oesophagitis, taking it only when needed works well. For severe erosive oesophagitis it does not, and continuous daily treatment is needed to keep the lining healed.

If you have been on one for a long time, do not stop abruptly. Acid production rebounds above where it started for a few weeks, and the symptoms that follow are easily mistaken for proof that the drug was essential. The clearest demonstration of this came from a trial in which 120 healthy volunteers, none of whom had reflux, took either a proton pump inhibitor for eight weeks or a placebo. In the month after stopping, 44 per cent of those who had taken the real medicine developed heartburn, regurgitation or indigestion, against 15 per cent of the placebo group. Stepping down over a few weeks avoids that.

H2 blockers

These reduce acid by a different route. They work faster but less completely. Availability in Australia varies by product and strength, so your pharmacist or GP is the one to ask. They have a place as an add-on for night-time symptoms and for people who cannot take a proton pump inhibitor.

Are reflux tablets safe long term

This is the single commonest question in a reflux consultation, and the internet has not been helpful on it.

Dozens of observational studies have linked long-term proton pump inhibitor use to pneumonia, fractures, kidney disease, dementia, heart attacks and cancer. The problem with all of them is that people who end up on these drugs for years are systematically different from people who do not. They are older, heavier, on more medications, more likely to smoke and more likely to have several other illnesses, and every one of those things independently raises the risk of the outcomes being measured. No amount of statistical adjustment fully removes that.

The question has been settled as well as it currently can be, by randomising it. A trial gave 17,598 people either a proton pump inhibitor or a placebo and followed them for a median of three years, looking specifically for the harms the observational studies had reported. There was no significant difference in pneumonia, fractures, gastric atrophy, kidney disease, diabetes, chronic lung disease, dementia, cardiovascular disease, cancer, hospitalisation or death. The one signal was a modest increase in gut infections, 1.4 per cent against 1.0 per cent.

Three years is not twenty, the people studied were older and had cardiovascular disease, and the trial cannot exclude a small excess of something rare. But it is far stronger evidence than the database studies it was designed to test, and it points the same way as everyday clinical experience.

Two practical consequences. First, people taking a proton pump inhibitor long term do not need routine monitoring of bone density, kidney function, magnesium or vitamin B12, and do not need calcium or probiotic supplements for that reason, which is the explicit advice of the American Gastroenterological Association. Second, none of this is an argument for staying on a dose you do not need. The right dose is the lowest one that controls symptoms, reviewed from time to time, and for uncomplicated reflux it is reasonable to try coming off it.

What actually helps, and what does not

Reflux advice tends to arrive as a long list of forbidden foods. Most of that list has never been shown to change symptoms or acid exposure, even where it has a real effect on the valve in a laboratory. Here is what the evidence supports.

• Losing weight is the strongest single measure. In a study of 10,545 women, the odds of frequent reflux symptoms rose steadily with body mass index, and a gain of more than 3.5 points even within the normal range was associated with nearly three times the odds of frequent symptoms. In a weight loss programme of 332 overweight adults, reflux prevalence fell from 37 per cent to 15 per cent in six months, and two thirds of those with symptoms lost them altogether.
• Raise the head of the bed for night symptoms. In a small randomised crossover trial of 39 people, raising the head of the bed by 20 cm improved reflux symptom scores in 69 per cent against 33 per cent sleeping flat, although quality of life did not change and more than half the participants found sleeping on a slope uncomfortable in some way. The authors felt that balance needed more study before the measure could be recommended outright. It costs nothing and is easily undone, so it is still worth trying. Raise the bed itself, with blocks or a wedge under the mattress. Stacking pillows is not what was tested and tends to bend you at the waist rather than tilt you.
• Finish eating earlier. Late evening meals measurably increase the time acid spends in the oesophagus overnight. Two to three hours before lying down is a reasonable target.
• Stopping smoking reduced reflux symptoms in people of normal weight in one large population cohort. Worth doing regardless of what it does for the heartburn.
• Alcohol, coffee, chocolate, mint, citrus and spicy food all relax the valve in laboratory studies, and yet there is no published evidence that avoiding them improves symptoms or acid exposure. Treat them as personal triggers worth testing one at a time, not as a list of rules. If coffee does not bother you, it does not bother you.

When the tablets are not working

Up to 40 per cent of people treated for reflux remain symptomatic on a proton pump inhibitor. The instinct is to double the dose. The evidence says the better move is usually to find out what is actually happening.

In a study of 168 people still symptomatic on twice-daily acid suppression, monitoring showed that only 11 per cent had symptoms attributable to acid reflux. Another 37 per cent had symptoms linked to non-acid reflux, and about half had symptoms with no time relationship to any reflux at all. More acid suppression was never going to help most of that group.

Two diagnoses account for much of it, and both are real conditions rather than a verdict that nothing is wrong.

• Reflux hypersensitivity: a normal amount of acid, but the symptoms genuinely do coincide with the reflux episodes. A normal amount of reflux is being felt as painful.
• Functional heartburn: normal acid exposure and no relationship between the symptoms and reflux events. The oesophagus is structurally and chemically normal, and the symptom comes from how the signals are processed.

In a series of 200 people with heartburn and a normal gastroscopy, 41 per cent had abnormal acid exposure, 32 per cent had reflux hypersensitivity and 27 per cent had functional heartburn. So the majority of endoscopy-negative heartburn is not an acid problem, which is the main reason for testing rather than escalating the dose.

Treatment for the latter two aims at how pain signals are processed rather than at acid, in the same family as the approaches used for irritable bowel syndrome and functional dyspepsia. It is worth being straight about the evidence: the one randomised trial to test a low-dose neuromodulator in exactly these two groups found no better symptom relief than a placebo, although quality of life did improve among those who completed it. Reviews still list this approach among the options for reflux that has not responded, and it is a reasonable thing to try, but it is not a reliable answer.

Cough, hoarseness and throat symptoms

Reflux reaching the throat is often blamed for a chronic cough, a hoarse voice, constant throat clearing or a feeling of a lump in the throat. Sometimes it is the cause. More often, in people who have these symptoms without much heartburn, it is not.

This has been tested. A randomised trial of high-dose acid suppression for 16 weeks in chronic posterior laryngitis found the primary symptom resolved in 14.7 per cent on treatment against 16.0 per cent on placebo. A randomised trial in chronic cough in people with little or no heartburn found no difference either.

The practical conclusion is not that these symptoms are imaginary or never caused by reflux. It is that an open-ended trial of acid suppression for a cough or a hoarse voice has a low chance of working, and that measuring the acid and looking for the other common causes is usually the better next step. A hoarse voice that does not settle should always be examined.

Hiatus hernia

A hiatus hernia means part of the stomach has slid up through the diaphragm into the chest. It disrupts the anti-reflux barrier and lets acid pool where it can reflux more easily, so it makes reflux more likely and is found more often in people with Barrett's oesophagus.

It is not the same thing as reflux disease, and this is where a lot of unnecessary worry starts. Small hiatus hernias are common and frequently cause nothing at all. On its own, a hernia on a report is supporting evidence rather than a diagnosis, and most need no treatment beyond managing the reflux.

Size does matter in one practical way. It influences which procedures are sensible, because a large hernia points towards surgical repair rather than anything smaller.

Surgery, and when it is worth considering

For confirmed reflux that will not settle, or where someone does not want to take medication for decades, a laparoscopic fundoplication wraps the top of the stomach around the lower oesophagus to rebuild the valve. It is done through keyhole incisions by an upper gastrointestinal surgeon, and it is funded under Medicare in Australia.

The trade-off is best described with numbers rather than adjectives. A European trial randomised 554 people to surgery or to a proton pump inhibitor and followed them for five years.

• Remission at five years was 92 per cent on medication and 85 per cent after surgery, a smaller gap than most people expect in either direction
• Surgery was clearly better for regurgitation: 2 per cent against 13 per cent
• Surgery was clearly worse for side effects: dysphagia 11 per cent against 5 per cent, bloating 40 per cent against 28 per cent, and flatulence 57 per cent against 40 per cent

A UK trial following people for five years found that 26 per cent of those who actually had a fundoplication were back on a proton pump inhibitor, and 4 per cent had a further reflux-related operation. So surgery is a real option with a real success rate, and it is not a guarantee of life without tablets.

That same UK trial is worth reading alongside the European one rather than after it, because on two points they disagree. Reflux symptom scores favoured surgery at every year of follow-up, it was good value for money by the usual health economic measures, and swallowing difficulty, wind and inability to vomit were no commoner after surgery than on medication. The side effect figures above come from the European trial. How much weight they deserve is therefore not settled, which is an argument for discussing them with the surgeon rather than deciding from a list.

Two things should happen before it is considered. The reflux should be objectively confirmed, with pH monitoring and manometry, because operating on symptoms that turn out not to be acid reliably disappoints. And the decision should be made with the surgeon who would do it.

Patients often ask about the device-based and endoscopic alternatives they have read about overseas, such as magnetic sphincter augmentation or endoscopic fundoplication. None of them has a Medicare item number in Australia and none is part of standard practice here, so this page does not describe them as options.

Reflux and Barrett's oesophagus

In some people with long-standing reflux, the lining at the bottom of the oesophagus is replaced by a lining more like the intestine's. This is Barrett's oesophagus. It causes no symptoms of its own, is found at gastroscopy and carries a small increase in the risk of oesophageal cancer, which is why it is checked periodically.

It has its own page, because the questions it raises are different from the ones reflux raises: Barrett's oesophagus, surveillance and treatment.

What else affects the oesophagus

• Eosinophilic oesophagitis is an allergic-type inflammation that causes food to stick, most often in young and middle-aged men. It is diagnosed on biopsies and treated with medication, diet or both.
• Strictures, rings and webs are narrowings that cause food to stick. Most are treated by stretching them with a balloon during gastroscopy.
• Motility disorders such as achalasia can cause food to stick or chest discomfort and are sometimes mistaken for reflux. Manometry is the test that sorts them out.

Where reflux is assessed in Sydney

A/Prof Viraj Kariyawasam is a consultant gastroenterologist and hepatologist. Reflux that has not settled on treatment, and reflux needing investigation, are assessed at consultation in Norwest, Rouse Hill, Pennant Hills and Macquarie Park.

The gastroscopy is performed at Norwest Private Hospital, Macquarie University Hospital and Pennant Hills Day Surgery. A referral from your GP is needed for a consultation.