Which Pain Medication Is Right for Me?

Pain medication is the one area where a naive comparison does real harm, because the drugs grouped under the label are not alternatives to one another. Ibuprofen and codeine do different things, for different kinds of pain, with different risks. Ranking them against each other on any single axis produces nonsense.

So the sensible way to read this page is in two halves. First, what kind of pain is being treated, which determines the group. Then, within that group, which member fits you best, and that is where genetics has something concrete to add.

This decides the group before anything else. Inflammatory and musculoskeletal pain generally responds to anti-inflammatories. Nerve pain often responds poorly to both anti-inflammatories and opioids, and is usually treated with an entirely different class. Acute severe pain after surgery or injury is where opioids have their clearest role. Chronic non-cancer pain is where opioids perform worst relative to their risks.

Getting this wrong is the most common reason a pain medication does not work, and it has nothing to do with genetics.

Anti-inflammatories carry real constraints. They can raise blood pressure, stress the kidneys, and cause gastrointestinal bleeding, and those risks rise with age, dose and duration. A history of ulcer or of kidney impairment may take the whole group off the table, or require gastric protection alongside. Celecoxib is easier on the stomach but carries its own cardiovascular considerations. These are usually the binding constraints, not efficacy.

Short courses and long-term use are different decisions. Many anti-inflammatories are reasonable for a few days and problematic for a few months. Opioids raise questions about tolerance and dependence that barely arise over three days and dominate over three months. The right answer for a sprain is often the wrong answer for arthritis.

Anti-inflammatories interact with anticoagulants, some blood pressure medications, and each other. Opioids interact with sedatives, and the combination with benzodiazepines is a well-established cause of respiratory depression.

This is also where interactions and genetics collide: several common medications, including some antidepressants, block CYP2D6. If you take one, codeine and tramadol may behave as though you were a poor metabolizer regardless of your genotype, which is a form of phenoconversion.

Within each group, this is where genetics earns a place. For the anti-inflammatories, CYP2C9 governs clearance and slow metabolizers accumulate higher levels at standard doses.^[1] For codeine and tramadol, CYP2D6 does something more dramatic: it is required to convert them into the compound that actually relieves pain, so the enzyme determines whether the drug works at all.^[2]

That difference is worth dwelling on. For most drugs, metabolism affects how much you get. For codeine, it decides whether you get anything.

Several effective anti-inflammatories are inexpensive and available over the counter, which is why they are usually tried first. Opioids require a prescription and, increasingly, a documented rationale. Practical availability shapes this decision more than it does most.

The two genes here map almost exactly onto the two groups, which is why a genetic result tends to point you within a group rather than between them.

The practical consequence is asymmetric. A CYP2C9 result usually adjusts a dose. A CYP2D6 result usually changes the drug, because there is no dose of codeine that works for someone who cannot activate it, and no safe dose for someone who activates it too well.

Notably, several opioids do not depend on CYP2D6 at all, so a poor or ultrarapid result narrows the options without exhausting them.

The clearest case is codeine or tramadol that did nothing at all, which is a classic CYP2D6 poor metabolizer picture and is worth distinguishing from a drug that helped a little but not enough. It is equally relevant in the opposite direction: unusual drowsiness or breathing difficulty on a standard dose warrants urgent attention and is consistent with ultrarapid metabolism.

For the anti-inflammatories the case is softer. A CYP2C9 result is useful mainly if you need an anti-inflammatory long term, or if you have had side effects at doses most people tolerate.

• Migraine Medication Not Working? How Genetics Affects Your Treatment
• Pain Medication Not Working? Why Your Body May Process It Differently
• Celecoxib (Celebrex) Side Effects: The CYP2C9 Connection
• How Is Ibuprofen Metabolized? CYP2C9, Side Effects, and Your Genes
• Meloxicam (Mobic) Side Effects: How CYP2C9 Raises the Risk
• 23andMe Drug Response: What You'll Actually See in a Report From Your Data

The most likely explanation is CYP2D6. Codeine is inactive as taken and has to be converted into morphine by that enzyme. Poor metabolizers convert very little and get correspondingly little relief. A medication you are taking may also be blocking the enzyme, which produces the same result temporarily. Either way, taking more is not the answer.

They are broadly comparable at equivalent doses, and the differences that matter are in side-effect profile and duration rather than raw potency. Celecoxib is gentler on the stomach; some others last longer per dose. Which suits you usually comes down to your kidney, stomach and cardiovascular risk rather than to strength.

No. The genes discussed here affect how you metabolise specific opioids, which influences relief and side effects. They do not predict dependence, which is driven by a much wider set of factors that no current test measures reliably. Be sceptical of anything sold on that basis.

Not directly. They treat different problems and carry different risks, so a ranking that mixes them is misleading. The useful comparison is within a group once the type of pain has determined which group is appropriate.