A Cough That Will Not Go Away: The Eight-Week Line, the Three Causes Behind Most Cases and What to Try
A cough that has outlasted the illness that started it is one of the most common reasons people end up seeing several doctors and getting several antibiotics, none of which help. The reason is usually that nobody has worked through a short list in order.
Three definitions first, because they decide everything that follows:
- 🟢 Acute cough — under 3 weeks. Almost always a viral infection. Needs nothing
- 🟡 Subacute — 3 to 8 weeks. Usually post-infectious, and usually resolves on its own. A cough persisting for a month after a cold is normal, not a complication
- 🔴 Chronic — 8 weeks or more. This is the one that needs investigating rather than waiting out
Note that this is a dry or mostly dry cough we are discussing. A cough bringing up sputum on most days for months is a different condition, covered in our guide to chronic bronchitis.
🚨 Before the list: what has to be excluded
Every adult with a chronic cough needs a chest X-ray, and these features mean assessment rather than a treatment trial.
⚠️ Red flags. Coughing blood, even a trace — always investigated. Unintentional weight loss, night sweats or fever. A new cough, or a clear change in a long-standing cough, in a current or former smoker over 45. Hoarseness lasting more than three weeks. Difficulty swallowing. Breathlessness. Recurrent chest infections. Anything abnormal on examining the chest. Finger clubbing or enlarged lymph nodes.
These are the features that distinguish the cough that is a nuisance from the cough that is the first symptom of something else, and lung cancer presents this way often enough that the X-ray is not optional.
📊 The three causes behind roughly nine in ten cases
In a non-smoker with a normal chest X-ray who is not taking an ACE inhibitor, the cause is almost always one of three things — and frequently more than one at the same time, which is why treating only the most obvious often half-works.
| Cause | Clues | The treatment trial |
|---|---|---|
|
Upper airway cough syndrome post-nasal drip, rhinitis, sinus disease |
Constant throat clearing, a tickle or dripping sensation at the back of the throat, blocked or running nose, worse on waking | Nasal steroid spray, with an antihistamine if there is an allergic pattern. 4 to 8 weeks before judging |
| Asthma and its cough-only forms | Worse at night and early morning, after exercise, in cold air, with laughing, or seasonally. Often no wheeze at all | Inhaled corticosteroid, 4 to 8 weeks. Spirometry with reversibility testing where available |
| Reflux | Worse lying flat, after meals, on talking or singing. Heartburn is frequently absent, which is why it gets missed | Acid suppression plus posture and diet measures, 8 to 12 weeks |
The practical approach is therefore sequential trials rather than one visit and a diagnosis. Each gets a fair length of time, each is stopped if it does nothing, and a partial response means two causes are present.
💊 The cause that is a tablet, and is missed constantly
⚠️ ACE inhibitor cough. Blood pressure tablets ending in pril — ramipril, lisinopril, perindopril, enalapril and the rest — cause a persistent dry tickly cough in up to one in five users.
Two details make it hard to spot. It can begin weeks or even months after starting the tablet, so people do not connect the two. And it does not respond to anything: cough syrups, antibiotics, inhalers, all useless while the tablet continues.
It resolves on stopping, usually within one to four weeks, occasionally taking up to three months. Switching to an angiotensin receptor blocker — a sartan — keeps the blood pressure control and removes the cough, with cough rates no higher than placebo. Do not stop a blood pressure tablet on your own; ask for the switch.
If you take one of these and have an unexplained chronic cough, this is the first thing to check, before any investigation.
Other drug causes worth a glance: sitagliptin and related diabetes tablets occasionally, and some drugs cause cough through other routes. And any inhaled irritant — vaping, occupational dust and fumes, a new workplace — belongs in the history.
🫁 Cough-variant asthma, and why a steroid inhaler is the test
This is the most satisfying cause to find, because treatment usually works well. In cough-variant asthma the cough is the only symptom — no wheeze, no breathlessness, normal examination and often normal spirometry. In non-asthmatic eosinophilic bronchitis there is airway inflammation of the same type without any airflow obstruction at all.
Both respond to an inhaled corticosteroid, and in practice a trial of one is both the treatment and the test.
- 🔹 Pulmicort (Budesonide) or Flovent (Fluticasone) twice daily, as directed, for four to eight weeks
- ⏳ Do not judge it at two weeks. Inhaled steroids reduce inflammation gradually and the cough improves after the inflammation does. Stopping early is the commonest reason this cause gets missed
- 💧 Rinse the mouth after each dose to avoid oral thrush and hoarseness
- 🎯 Technique is half the treatment. Ask for a spacer with any pressurised inhaler and have your technique watched rather than described — a large share of people use these incorrectly
- 💊 Singulair (Montelukast) is an oral alternative or addition, particularly where there is an allergic or exercise-related pattern. Report any mood change or vivid dreams, which it occasionally causes
- 💨 A rescue inhaler such as Proventil (Albuterol) relieves the moment and treats nothing. If it helps, that is a useful clue pointing at asthma — and the steroid is what changes the outcome
The range sits in the asthma category, and the full picture of asthma is in our asthma guide.
🔥 Reflux cough, usually without the heartburn
Reflux reaching the throat irritates the cough receptors there, and in a substantial proportion of people it produces a cough and no heartburn at all — which is why they reject the suggestion when it is made. The clues are positional and behavioural rather than digestive.
- 🛏️ Worse lying flat, and on first lying down at night
- 🍽️ Worse after meals, especially large or late ones
- 🎤 Worse on talking at length, singing or bending forward
- 🗣️ Hoarseness, a lump-in-the-throat sensation, or a sour taste on waking
What actually matters more than the tablet here is the physical measures, and they get skipped:
- 🛏️ Raise the head of the bed by 10 to 15 cm — blocks under the legs, not extra pillows, which bend the stomach and make it worse
- ⏰ Nothing to eat for three hours before lying down
- 🍽️ Smaller meals, and less alcohol, caffeine, chocolate, fatty food and late-night eating
- ⚖️ Weight reduction and loosening tight waistbands, both of which reduce pressure directly
- 💊 Acid suppression — Prilosec (Omeprazole) 20 to 40 mg daily, or another proton pump inhibitor, 30 to 60 minutes before the first meal of the day, for 8 to 12 weeks. Taken at a random time it works considerably less well
Honest caveat: acid suppression alone frequently fails for cough, especially where there is no heartburn, because non-acid reflux and the mechanical event itself still occur. That is why the posture and timing measures carry so much of the weight, and why failure of a PPI does not rule reflux out.
🌀 When all three have been treated and the cough remains
This is where most articles stop and where a large number of patients actually live. The modern explanation is a specific and useful one.
💡 Cough hypersensitivity syndrome. The cough reflex itself becomes sensitised — the nerves in the airway and the processing in the brainstem are turned up, so stimuli that should cause nothing now trigger coughing.
The recognisable pattern: a tickle or irritation in the throat that demands a cough, triggered by talking, laughing, cold air, aerosols and perfumes, a change in temperature, with bouts that are hard to stop once started. It often begins after a viral infection and then never settles.
It is commoner in women by roughly two to one, which reflects a genuinely more sensitive cough reflex rather than anything else. And it is not psychological, not a habit, and not something to be told to ignore.
What helps, once the treatable causes have been addressed:
- 🗣️ Speech and language therapy for cough control — the single most useful referral and the least known. It teaches recognition of the urge, suppression techniques, breathing patterns and throat care, and it has real trial evidence. Ask for it by name
- 💊 Neuromodulator medication — low-dose gabapentin, pregabalin, amitriptyline, or low-dose slow-release morphine in specialist hands. These act on the sensitised reflex rather than on the airway, and they are prescribed by a cough or respiratory clinic
- 💧 Throat care —> sipping water through bouts, avoiding throat clearing which re-irritates, humidifying dry rooms, and steam-free nasal saline
- 🧪 Newer targeted drugs acting on the airway nerve receptors have reached the market in some countries, which is the first genuinely new class in decades
🚭 The smoker cough
If you smoke, that is the cause until proven otherwise, and the useful fact is the one most people are not told: stopping resolves or substantially reduces the cough in the majority within weeks to a few months.
The catch that sends people back: it often gets worse for the first few weeks, as the cilia recover and start clearing what has accumulated. That phase passes, and knowing it is coming is what gets people through it.
Vaping also causes chronic cough and throat irritation, and switching from smoking to vaping frequently does not resolve it.
🚫 What does not work
| Commonly tried | Reality |
|---|---|
| Antibiotics | No effect on any of the causes above. A chronic cough is not an infection, and repeated courses are how people end up with resistant organisms and no improvement |
| Cough suppressant syrups | Weak and inconsistent evidence. Honey in warm water performs about as well and costs less |
| Antihistamine and decongestant combinations | Useful where there is a genuine allergic nasal cause, close to useless otherwise |
| Steam inhalation over a bowl | No proven benefit, and a real scalding risk |
| Stopping before four weeks on a treatment trial | The commonest reason a cause gets wrongly excluded. Each trial needs its full length |
| Throat clearing | Feels necessary and re-irritates the throat each time, keeping the cycle running |
🧱 What a long cough actually costs
This is worth naming because people apologise for consulting about it, as though a cough were trivial.
- 🦴 Rib fractures from coughing, which happen and are extremely painful
- 🚽 Urinary leakage on coughing — very common in women, rarely volunteered, and it improves when the cough is treated. It is not a separate problem to be endured
- 😵 Fainting during a coughing bout, which is a reason to be seen
- 😴 Broken sleep, exhaustion and headache
- 👥 Avoiding meetings, cinemas, travel and company, and the embarrassment of being assumed infectious
- 🧠 Low mood and anxiety, which are measurably common in refractory chronic cough and part of what treatment improves
📋 How to make the appointment productive
- 📅 When it started, and what you were doing or ill with at the time
- 💧 Dry or productive, and if productive, how much and what colour
- 🕒 When it is worst — night, morning, lying down, after food, on talking, in cold air
- 💊 Every medicine you take, with the ACE inhibitor question answered explicitly
- 🚭 Smoking and vaping history, and any dust or fume exposure at work
- 🏡 Nasal symptoms, heartburn, wheeze — even if mild, because they point at the three causes
- 📝 What has already been tried, at what dose and for how long. This is the detail that prevents repeating a failed trial
📞 When to seek help
Urgently:
- Coughing blood, in any amount
- Breathlessness at rest, or chest pain
- Fainting with coughing
- Difficulty swallowing, or food going down the wrong way repeatedly
Promptly:
- Cough lasting 8 weeks or more — and ask for a chest X-ray if you have not had one
- A new or changed cough in a current or former smoker over 45
- Weight loss, night sweats or fever alongside
- Hoarseness beyond three weeks
- Taking a blood pressure tablet ending in pril — ask about switching before anything else is investigated
- Each treatment trial run properly and the cough unchanged — ask for a respiratory or specialist cough clinic referral and for speech and language therapy
- Leaking urine when you cough, which is treatable and worth mentioning
- Repeated antibiotic courses with no lasting benefit — the diagnosis is wrong, not the antibiotic
❓ Frequently asked questions
How long is too long for a cough?
Under three weeks is an ordinary viral cough. Three to eight weeks is usually post-infectious and settles on its own. Eight weeks or more is chronic and needs investigating, starting with a chest X-ray.
What are the commonest causes of a chronic cough?
In a non-smoker with a normal X-ray, three causes account for about nine in ten: upper airway problems such as post-nasal drip, asthma including its cough-only forms, and reflux. Often more than one is present at once.
Can my blood pressure tablet cause a cough?
Yes. ACE inhibitors, the ones ending in pril, cause a dry persistent cough in up to one in five users, sometimes starting months after beginning the tablet. Switching to a sartan resolves it. Ask for the switch rather than stopping it yourself.
Can you have asthma with only a cough?
Yes, and it is called cough-variant asthma — no wheeze, no breathlessness, often normal spirometry. A trial of an inhaled corticosteroid for four to eight weeks is both the treatment and the test.
Why has my steroid inhaler not helped after two weeks?
Because inhaled steroids reduce inflammation gradually and the cough only improves after the inflammation does. Four to eight weeks is the proper trial length, and stopping early is the commonest reason this cause gets wrongly excluded.
Can reflux cause a cough without heartburn?
Frequently. Reflux reaching the throat irritates cough receptors there with no heartburn at all. The clues are positional: worse lying flat, after meals, and on talking or singing.
How should I take a proton pump inhibitor for cough?
30 to 60 minutes before the first meal of the day, for 8 to 12 weeks, alongside raising the head of the bed and not eating for three hours before lying down. Taken at a random time it works considerably less well.
What if everything has been treated and I still cough?
That is usually cough hypersensitivity — the cough reflex itself has become sensitised, triggered by talking, laughing, cold air or perfumes. Speech and language therapy for cough control, and neuromodulator medication from a specialist clinic, are the treatments.
Will antibiotics help a cough that has lasted months?
No. None of the common causes of chronic cough is an infection, so antibiotics do nothing except add side effects and resistance. Repeated courses with no lasting benefit mean the diagnosis is wrong.
Will my cough stop if I quit smoking?
In the majority it resolves or substantially improves within weeks to a few months. It often gets worse for the first few weeks as the airway clearance system recovers, and that phase passes.
📑 Sources and editorial
- Respiratory society guidance on the assessment and management of chronic cough in adults, including the eight-week definition and the sequential treatment trial approach
- Guidance on red flag features requiring chest imaging and urgent referral in chronic cough
- Reviews of the prevalence of upper airway cough syndrome, cough-variant asthma, non-asthmatic eosinophilic bronchitis and gastro-oesophageal reflux as causes of chronic cough
- Reviews of ACE inhibitor-induced cough incidence, time to onset, time to resolution and the switch to angiotensin receptor blockers
- Evidence for inhaled corticosteroid trials in cough-variant asthma and eosinophilic bronchitis, including recommended trial duration
- Evidence on proton pump inhibitor therapy in chronic cough, including the limited benefit in the absence of reflux symptoms, and on non-pharmacological anti-reflux measures
- Consensus statements defining cough hypersensitivity syndrome and refractory chronic cough, including sex distribution
- Randomised evidence for speech and language therapy cough-control interventions, and for gabapentin, pregabalin, amitriptyline and low-dose morphine in refractory cough
- Cochrane reviews of over-the-counter cough preparations, honey, antihistamine-decongestant combinations and steam inhalation
- Studies of complications and quality of life impact in chronic cough, including cough-induced rib fracture, urinary incontinence and syncope
- Related reading: chronic bronchitis, asthma, upper respiratory infections
- Related products: Pulmicort (Budesonide), Flovent (Fluticasone), Singulair (Montelukast), Prilosec (Omeprazole), asthma category
- RXshop Editorial Team — reviewed by Emily Chen, MD, MPH — Internal Medicine Specialist
Medical Disclaimer: The information in this article is for educational and informational purposes only and is not a substitute for professional medical advice, diagnosis, or treatment. Always seek guidance from a qualified healthcare provider with any questions you may have regarding a medical condition, and before starting, stopping or changing any medication.