50

This 2016 Express article gives a rather disturbing warning that two common pain killers might cause the risk of hearing loss.

PARACETAMOL or ibuprofen could cause long-term deafness, experts have revealed. Scientists have revealed the effects could caused by (sic) taking the painkillers for as frequently as (sic) two days a week for six years.

Compared to women who have taken them for less than a year, women who regularly the two (sic) to treat chronic pain increased the risk of hearing loss by more than a sixth, the study revealed.

(Paracetamol is also commonly called acetaminophen or APAP, and under the brand names of Panadol or Tylenol.)

If true, this sounds like a serious concern, but the article is poorly written and doesn't make it clear if it is from combining the two drugs or also applies to them when taken separately, whether it applies to both men and women, or women alone. It doesn't link to the report. It has also been a few years, so there should have been time for post-production peer review, and perhaps other studies.

What is the real story? What demographics have been tested? What dosages of which drugs were involved? Is it more than merely correlation?

Community
  • 1
Oddthinking
  • 140,378
  • 46
  • 548
  • 638
  • 10
    2 days a week for 6 years is *a lot* of drugs. If you have pain that often, see a doctor. Up there with a chiropractor's worry when I told him that I get occasional headaches (once every few months): "most people say they get occasional headaches, only 'five times a week lasting six hours'." – Draco18s no longer trusts SE Sep 10 '19 at 16:50
  • 31
    What's surprising to me about this claim is that it covers both ibuprofen and paracetamol. AFAIK they're completely different kinds of drugs, and just about the only thing they have in common is that they both block pain — and they even do that via different mechanisms. *A priori*, there seems to be no reason to expect them to be more likely to share a common side effect than any two random drugs. One might even wonder if maybe it's actually *chronic pain* (or some common cause of it) that's somehow linked to hearing loss (not that I have any evidence at all to back up that suspicion, either). – Ilmari Karonen Sep 10 '19 at 17:19
  • 1
    @IlmariKaronen The original paper I linked in my answer contains some speculations on the mechanisms, and they are indeed different. I don't have the impression this is well understood. – Mad Scientist Sep 10 '19 at 20:46
  • I don't understand why you've added "[sic]" to "for as frequently as [sic] two days a week". The other two are justified, although I'd change it to "the effects could caused [sic] by". Either that, or correct the text yourself: "the effects could [be] caused by", "women who regularly [take] the two". I think correcting the quote in square brackets would be better. – CJ Dennis Sep 11 '19 at 03:09
  • 1
    @CJDennis: I've put the "sics" in rather than correcting to passively-aggressively demonstrate that the original article is poorly written - it looks thrown together in a rush - which is part of the reason for my doubt about its claims. "As frequently as" should read "As infrequently as", or it doesn't really make any sense. – Oddthinking Sep 11 '19 at 03:25
  • 5
    @Draco18s: You are making some huge assumptions. That people haven't seen a doctor. That people can afford to see a doctor. That 2 doses of Panadol a week isn't actually very low for someone in chronic pain. – Oddthinking Sep 11 '19 at 03:27
  • 1
    Personally, putting the "[sic]"s in looks "passive aggressive" ("Look at the mistake right here!"), not adding the missing words which looks more like ellipsis or paraphrasing without losing the original meaning. – CJ Dennis Sep 11 '19 at 03:36
  • @CJDennis: I think we are agreeing. – Oddthinking Sep 11 '19 at 04:04
  • 1
    Yes, I read "rather than correcting to passively-aggressively demonstrate" as written without a comma, instead of "rather than correcting, to passively-aggressively demonstrate" or "rather than correcting in order to passively-aggressively demonstrate". – CJ Dennis Sep 11 '19 at 04:07
  • 4
    @Draco18s 2 days a week is not unusual in well-managed diseases. There was a time where 4 doses of paracetamol was a good week, migraine wise, despite being taking four different preventive drugs. (Luckily the newest generation works great for me, so it is not a problem anymore). Rheumatism "in remission" means less than half the joints are inflamed (they still hurt). My point is, those dosages are not uncommon for people with chronic diseases. – Davidmh Sep 11 '19 at 17:23

2 Answers2

77

Yes.
That is: They can cause hearing loss. They are properly classified as ototoxic agents.

Paracetamol/Acetaminophen is certainly not entirely harmless. In fact it is properly classified as an ototoxic drug. An effect observed for quite some time. According to the individual studies: This is more pronounced when using it in each of these cases: higher doses, over longer time, abuse, drug combinations. But also being a women, being a man, being a kid, being elderly. If that sounds paradoxical, be aware that most studies do not even look at the general population but only subgroups, restricting their outcomes to the subgroup studied.

Of these, analgesic drug combinations seem to be the most risky, followed directly by combination with any other ototoxic drug.

However, this is also not peculiar for paracetamol alone, as aspirin for example can have this effect as well. The models to explain this effect are convincing, but real life outcome measures and population risk quantification remain a bit lacking for now.

Focusing on paracetamol: it is an ototoxic chemical, the possible mechanisms quite clearly demonstrated. Then dosage makes the poison, and neither observational long term studies in sufficient quantity and quality are available, nor are human experiments or prospective studies on that level available.

Over-the-counter painkillers linked to hearing loss. The three most commonly used medications in the U.S. have a significant, and little known, side effect.

Ototoxicity is a trait shared by several classes of drugs, most notably by aminoglycoside antibiotics, cisplatin chemotherapy and by acetaminophen combined with narcotic medication. Animal models help to elucidate the mechanisms of hearing loss from these medications.

Acetaminophen combined with narcotic medications, when taken chronically at acceptable doses or when abused at high doses, causes profound irreversible sensorineural hearing loss. Hearing loss had previously only been reported after the use of the narcotic propoxyphene. For this reason it was initially believed that the narcotic in Lortab! (Whitby Pharmaceuticals, Richmond, VA) or Vicodin! (Abbott Laboratories, Abbott Park, Illinois, USA) was responsible for the ototoxicity. However, research in mouse auditory cell line and cochlea cultures show that of the two active ingredients in these drugs, acetaminophen, and not hydrocodone, is responsible for hair cell toxicity. Curhan et al. recently published a study examining the incidence and hazard ratio of self-reported hearing loss due to regular use of aspirin, other NSAIDs and acetaminophen. They reviewed reports in 26,917 men aged 40–74 years, with baseline hearing at 1986, and found 3,488 incident cases of hearing loss. For NSAIDs and acetaminophen, the risk increased with longer duration of use. The hazard ratio for acetaminophen was 1.22, and in men < 50 years old the risk increased to 1.99, more than the other drugs.

enter image description here
Figure 2. Fluorescent microscopy of the basal turn of the cochlea in a control culture (A) with Math1-GFP fluorescence demonstrating the intact single row of inner hair cells (closed arrow) and three rows of outer hair cells (open arrow). A basal turn of a culture exposed to high-dose acetaminophen (B) with decreased Math1-GFP expression demonstrating damage of inner (closed arrow) and outer hair cells (open arrow) compared to the control culture.

enter image description here
Figure 3. Light microscopy of HEI-OC1 cells in culture after 48 h when exposed to hydromorphone alone (A) compared to hydromorphone and acetaminophen (B).
–– Joshua G Yorgason†, William Luxford & Federico Kalinec: "In vitro and in vivo models of drug ototoxicity: studying the mechanisms of a clinical problem", Expert Opinion on Drug Metabolism & Toxicology, Volume 7, Issue 12, 2011.

The overall data varied, demonstrating a measurable effect on self-reported symptoms from NSAIDs as a class, ibuprofen, and acetaminophen, but there are no audiometric data to confirm or refute this suggested effect. Sulindac was the only specific agents to have been studied with formal audiometry as a primary outcome in a randomized double-blind placebo-controlled trial, and the significant effect seen in the unadjusted analysis dissipated in the adjusted multivariate regression.
–– Kyle ME, Wang JC, Shin JJ.: "Impact of nonaspirin nonsteroidal anti-inflammatory agents and acetaminophen on sensorineural hearing loss: a systematic review", Otolaryngol Head Neck Surg. 2015 Mar;152(3):393-409. doi: 10.1177/0194599814564533. Epub 2015 Jan 5.

Regular use of aspirin, NSAIDs, or acetaminophen increases the risk of hearing loss in men, and the impact is larger on younger individuals.
–– Curhan SG, Eavey R, Shargorodsky J, Curhan GC.: "Analgesic use and the risk of hearing loss in men", Am J Med. 2010 Mar;123(3):231-7. doi: 10.1016/j.amjmed.2009.08.006.

Conclusion

There are several pathways and a multitude of observational data points that in vitro explain and in vivo record paracetamol as being damaging to hearing in principle. Neither really large and conclusive prospective studies are there to present seemingly exact numbers. The fact that paracetamol is swallowed by millions of people daily indicates that in doses small enough to not exceed recommended values these really long term effects are probably statistically significant, but not exceedingly strong. The risks for individuals or on the population level are to be considered separately.

The current studies looking at humans in large numbers have several weaknesses that need addressing. Self-reported hearing-loss, self-reported dose regimen over a long time span, etc. Hearing gets worse as we age, and some amount of hearing loss is 'natural' and thus unavoidable. Not all confounding factors like noise exposure, inflammation, other toxins can be properly and rigorously controlled for over such a long term.

Nevertheless, hearing loss is a very widespread problem and it looks like longterm painkiller use in moderate amounts seems to be a factor not to be overlooked in future studies. According to the observational studies in humans the current perspective is:

“Although the magnitude of higher risk of hearing loss with analgesic use was modest, given how commonly these medications are used, even a small increase in risk could have important health implications. Assuming causality, this would mean that approximately 16.2 percent of hearing loss occurring in these women could be due to ibuprofen or acetaminophen use,” said Curhan.
–– Haley Bridger: "Longer use of pain relievers tied to hearing loss in women", The Harvard Gazette, December 14, 2016.

And

For acetaminophen, regular users aged less than 50 were 99% more likely, regular users aged 50–59 were 38% more likely, and those aged 60 and older were 16% more likely to have hearing loss than non-regular users of acetaminophen. (Curhan 2010)

To turn this perspective around: The risk for any individual taking paracetamol to actually develop hearing problems from that is not expressed in the above numbers and it cannot be calculated from the numbers given so far.

LangLаngС
  • 44,005
  • 14
  • 173
  • 172
  • 8
    I blame myself for not making this clearer in the question, but this answer spends a lot of time on (short-term) *acute* overdoses of paracetamol. That is, in my mind, undoubtably, very dangerous. But I understood (from my GP) that (long-term) *chronic* use of recommended levels of paracetamol to be very safe, which is why the original article was shocking to me. – Oddthinking Sep 10 '19 at 12:11
  • @Oddthinking The short-time high-dose exposure shows a very large effect, the long term *regular* use of ø2 caps a week over 1–6 years yields a 5% incidence in one study linked to paracetamol alone. That is shocking for this vastly overprescribed and eaten like candy 'harmless' pill. I gather you want a complete risk profile for recommended (ie usual & maximum) listed as 'safe' dose but only regarding ototoxicity? – LangLаngС Sep 10 '19 at 12:27
  • 4
    Let me give you the straight answer and then explain why you should ignore my desires. I used to take paracetamol at far higher doses than 2 doses a week. I didn't eat it like candy, because I avoided ever exceeding the recommended daily maximum, but I took several doses a day for months. I wasn't prescribed paracetamol, because it is freely available in supermarkets in Australia, but my GP knew and approved, and assured me it was safe long-term. So the idea that it might affect my hearing is shocking news to me. (I still would have made the same decision, but it wasn't an informed decision.) – Oddthinking Sep 10 '19 at 13:12
  • But Skeptics.SE is not a personal advice site. The answers you post shouldn't be tailored specifically to my personal situation, but be useful to others in the future. "Incidences" needs careful explanation, because I imagine there are lots of hearing loss suffering the control groups too. – Oddthinking Sep 10 '19 at 13:14
  • 4
    @Oddthinking The "no personal advice" is the reason I included the other studies linking paracetamol to hearing damage. From in vitro to large acute doses and combined use/abuse are the current state of knowledge. The 'news' part being the problem here as only post-hoc data on population level is available. Of course there are confounding factors, not least simple age and poly-drug use, or using painkiller for ear-ache *because* of noise. Curhan states that as "*perhaps* 16% of the hearing loss observed is attributable to OTC painkiller use, paracetamol seeming the worst of them". – LangLаngС Sep 10 '19 at 13:25
  • 7
    How one can be neither man nor woman nor a kid nor an eldery? – Anixx Sep 10 '19 at 15:56
  • 1
    @Anixx That's tough, indeed. The list illustrates that most studies just focused on one of these subgroups, like data derived from "Nurse's Health Study": very few kids in that cohort, therefore study doesn't lend itself to be applicable to 'dangers for kids'. Real problem here is that a truckload of weak or limited evidence/studies are not a real ersatz for large high quality ones. It's tempting to just add up all these, but it is usually not so simple… – LangLаngС Sep 10 '19 at 16:01
  • You write that the effect is "more pronounced" when the drug is used by men, and also by women, and also by kids. What does that mean? Is this an instance of [Simpson's paradox](https://en.wikipedia.org/wiki/Simpson%27s_paradox), where the correlation is stronger in each of those three categories, and weaker in the overall population? – Tanner Swett Sep 10 '19 at 16:57
  • @TannerSwett Thx. It may be indeed sth like that, but the overall picture is just emerging and what I meant is just that the individual studies are limited to subgroups without the studies being *directly* comparable or combinable in findings. I tried to clarify that above. – LangLаngС Sep 10 '19 at 17:04
  • 2
    @Oddthinking Entire "Lifetime use of a drug" causing 1/6th of people's hearing loss (already a rare thing) does mean that each week has a tiny chance of causing a problem (ie, causes no detectable harm) – Yakk Sep 11 '19 at 15:35
  • 4
    @Anixx you can be ages 20-65 (which is a vast swath of the population). – RonJohn Sep 11 '19 at 21:33
  • 2
    @Anixx Intersex and middle-aged. – ApproachingDarknessFish Sep 13 '19 at 09:14
  • (*otcomes* → *outcomes*) – Peter Mortensen Sep 13 '19 at 12:21
21

The referenced study is likely "Analgesic Use and the Risk of Hearing Loss in Women" in the American Journal of Epidemiology. The conclusion states

In conclusion, this prospective study showed that use of ibuprofen or acetaminophen 2 or more days per week is associated with an increased risk of hearing loss in women and that the magnitude of the risk tends to be greater with increasing frequency of use. There was no relation observed between aspirin use and the risk of hearing loss.

I found an analysis on the NHS site of a Daily Mail article about the same study. It does call the article unnecessarily alarmist, but acknowledges that regular use of these drugs could have health consequences over time:

Many people use medicines like paracetamol or ibuprofen for aches and pains. There's no suggestion from this study that occasional use to manage a headache or muscle strain is harmful.

But the study is a reminder that regular use – defined by the researchers as two days or more a week – could have health consequences over time.

[...]

Also, the increased risk was small – as there's no information about some factors that can also affect hearing, such as exposure to loud noise, it's possible unmeasured confounding factors might have explained the link.

[...]

Overall, however, this study adds to previous evidence that the regular use of paracetamol and NSAID medicines could be a factor in some cases of hearing loss. It makes sense to limit your use of these drugs to when they're needed.

The study does indicate that there could be a health risk, but it is small and connected to regular use, not occasional use.

Community
  • 1
Mad Scientist
  • 43,643
  • 20
  • 173
  • 192