⚕️ Educational information only — not medical advice. Read the full disclaimer

Medications That Can Cause Joint Pain or Worsen Arthritis

Many people are surprised to learn that some of the most commonly prescribed drugs may, in certain circumstances, contribute to joint or bone problems. This page reviews the medications that worsen arthritis — a theme popular in integrative and functional-medicine circles — against the peer-reviewed evidence: the so-called NSAID paradox, fluoroquinolone antibiotics, long-term corticosteroids, and drug-induced nutrient depletions. Each claim is graded for the strength of its evidence, because some associations are well established while others remain emerging or contested.

Never start, stop, or change any medication on your own. Every drug on this page is prescribed because a physician judged its benefits to outweigh its risks for a specific person. This article is informational only. Stopping a needed antibiotic, steroid, acid blocker, statin, or pain reliever can be dangerous. If anything here concerns you, write it down and discuss it with the doctor who prescribes the medication — the decision is theirs and yours together, never this website's.
Important: This page is general educational information, not medical advice. Arthritis varies by person and type. Always consult a qualified healthcare professional before changing your diet, supplements, exercise, or medication, and never stop a prescribed medication without speaking to the prescribing physician.

Key takeaways

  • Whether long-term NSAID use speeds cartilage loss is Contested: lab studies and two older trials of indomethacin raise concern, but evidence for commonly used NSAIDs is mixed and mostly observational. Guidelines still strongly recommend oral NSAIDs for knee and hip osteoarthritis.
  • Fluoroquinolone antibiotics carry an FDA boxed warning for tendinopathy and tendon rupture Strong; their direct cartilage toxicity in adults is Emerging.
  • Long-term corticosteroids are well documented to cause bone loss and are the leading cause of non-traumatic osteonecrosis Strong.
  • Some drugs affect minerals, vitamins or uric acid: diuretics raise the risk of gout, long-term PPIs are linked to low magnesium and B12, and metformin to low B12 Strong. Whether these depletions worsen arthritis itself is unproven.
  • Decisions about any of these medications belong to your prescribing physician.

Conventional medicine uses pharmaceuticals to manage the symptoms of arthritis and the metabolic conditions that often accompany it. A critical but underreported theme in holistic literature is the iatrogenic (medically induced) contribution to joint damage: the idea that, in some cases, the very drugs used to relieve symptoms may add to the underlying burden on cartilage, bone, and connective tissue. We present these claims honestly and with their evidence grade attached — not as settled fact, and never as a reason to abandon a prescription.

The NSAID Paradox Contested

Nonsteroidal anti-inflammatory drugs (NSAIDs) such as ibuprofen, naproxen, and diclofenac are the primary pharmacological tool for osteoarthritis pain worldwide. The 2019 American College of Rheumatology/Arthritis Foundation guideline strongly recommends oral NSAIDs for osteoarthritis of the hand, hip and knee, and topical NSAIDs for knee osteoarthritis.1 So does do NSAIDs make arthritis worse have any basis? Some integrative and prolotherapy practitioners argue there is a paradox: NSAIDs relieve pain, but long-term use may interfere with the joint's own repair process. The peer-reviewed evidence gives that concern a partial, drug-specific basis, but it does not settle it.

Proposed mechanism (NSAID paradox cartilage hypothesis):

NSAIDs work by inhibiting cyclooxygenase (COX-1 and COX-2) and thereby suppressing prostaglandins such as PGE2. Proponents argue that prostaglandins help drive chondrocyte repair, so blocking them blunts cartilage maintenance. Laboratory work partly supports a drug-specific version of this: in cultured human arthritic cartilage, naproxen, ibuprofen, indomethacin and nimesulide significantly inhibited glycosaminoglycan (GAG) synthesis, piroxicam, tiaprofenic acid and aspirin had no significant effect, and aceclofenac, tenidap and tolmetin stimulated it.2 The simple "PGE2 is a repair signal" story is also challenged: in cartilage from people with advanced osteoarthritis, PGE2 inhibited proteoglycan synthesis and promoted collagen breakdown, and the COX-2 inhibitor celecoxib reversed some of these catabolic effects.3

What the human studies show. The clearest signal is for indomethacin. In 105 people awaiting hip replacement, hips treated with indomethacin lost joint space faster than the opposite hip, a difference not seen with azapropazone.4 In the knee LINK study, more than twice as many patients on indomethacin as on placebo showed radiographic deterioration (40 of 85 vs 19 of 85), while tiaprofenic acid did not differ from placebo.5 For other NSAIDs the evidence is observational: in the Rotterdam Study, using diclofenac for more than 180 days was linked to about 2.4 times the risk of hip and 3.2 times the risk of knee radiographic progression compared with short-term use, with wide confidence intervals;6 and in the Osteoarthritis Initiative, current NSAID users lost slightly more joint space (about 0.04 mm more) than non-users.7

How to read this honestly. Mainstream rheumatology does not accept "NSAIDs accelerate arthritis" as established. The trial evidence of harm is mostly for indomethacin, a drug rarely used long term for osteoarthritis today. The mechanistic data are from cartilage in a dish, and observational links can reflect that people with worse, faster-progressing arthritis take more NSAIDs (confounding by indication), or that pain relief leads to more loading of a damaged joint.6 The Osteoarthritis Initiative authors concluded that clinical trials are needed.7 That is why we grade this section Contested. NSAIDs remain a standard, guideline-supported option. If you are weighing long-term use, that is exactly the kind of conversation to have with your physician, who may discuss the lowest effective dose, which NSAID, periodic review, and physician-supervised dietary and lifestyle alternatives. Oxidative stress and nutrient status, discussed under our root causes of arthritis, form part of that bigger picture.

Fluoroquinolone Antibiotics Strong (tendon) Emerging (cartilage)

Fluoroquinolones — ciprofloxacin, levofloxacin, moxifloxacin, ofloxacin and the newer delafloxacin — are broad-spectrum antibiotics. Their link to fluoroquinolone tendon damage is one of the better-documented drug-musculoskeletal associations: since 2008 the U.S. FDA has required a boxed warning for tendinitis and tendon rupture, most often of the Achilles tendon.8910 The boxed warning now covers disabling and potentially permanent effects on tendons, peripheral nerves and the central nervous system, plus worsening of myasthenia gravis. The 2018 FDA changes on aortic aneurysm and dissection, low blood sugar and mental-health effects were added to the label's Warnings and Precautions section; they are not part of the boxed warning.11 That tendon and FDA-warning risk is well established, so we badge it Strong. The claim that these drugs are directly toxic to articular cartilage in humans is more preliminary and is graded Emerging. Quinolone cartilage damage is well shown in juvenile animals, including rats,12 which is why these drugs are used cautiously in children. In a systematic review covering 16,184 children given ciprofloxacin, musculoskeletal events occurred in about 1.6%, half of them joint pain, and all cases of arthropathy resolved or improved with management.13 Lasting cartilage harm in adults has not been demonstrated.

Proposed mechanisms (mostly cell and animal studies):

  • MMP upregulation. In human tendon-derived cells, ciprofloxacin and norfloxacin increased expression of the collagenase MMP-1 (while reducing MMP-13), and the same group had earlier reported ciprofloxacin boosting MMP-3 and MMP-1.14 In cultured human tenocytes, ciprofloxacin raised MMP-1, lowered the MMP inhibitor TIMP-1 and reduced an enzyme needed to cross-link collagen.15
  • Direct tenocyte toxicity. Ciprofloxacin reduced type I collagen and the β1-integrin receptor that anchors cells to their matrix, and dexamethasone made the effects worse.16
  • Oxidative stress. In tendon cell lines, fluoroquinolones caused early oxidative stress, with toxicity varying between drugs.17
  • Magnesium chelation. Fluoroquinolones bind metal ions, especially magnesium. In rats, ofloxacin caused degenerative changes in Achilles tenocytes that were considerably worse on a magnesium-deficient diet, with cells detaching from the surrounding matrix.18 Whether everyday magnesium status matters in humans is unknown.

The practical point is not "never take antibiotics" — fluoroquinolones can be the right, even life-saving, choice in serious infections. It is that the risk is real enough that regulators warn about it, and that you and your prescriber should weigh it, especially if you have a history of tendon problems or take corticosteroids concurrently.

Full guide: Fluoroquinolones and tendons covers which antibiotics are affected, who is at highest risk, symptoms to act on, and what to discuss with your prescriber.

Long-Term Corticosteroids Strong

Corticosteroids such as prednisone and dexamethasone are powerful and genuinely valuable for suppressing severe autoimmune flares in rheumatoid and psoriatic arthritis. The concern here is specifically long-term use, where the bone toll is well documented — hence the Strong grade. This is about long-term corticosteroids and bone loss, not short courses.

Proposed mechanism:

  • Suppressed bone formation. After oral glucocorticoids are started, bone is lost rapidly and fracture risk rises within months in a dose-dependent way, mainly because bone formation is inhibited; altered intestinal and renal calcium handling, lower sex hormones and reduced IGF-1 also contribute.19
  • Osteoblast apoptosis. They increase apoptosis of osteoblasts (bone- forming cells) and osteocytes, shown in mice and in patients with glucocorticoid-induced osteoporosis,20 while prolonging the lifespan of bone-resorbing osteoclasts in animal models.21
  • Osteonecrosis (AVN). In some cases they compromise the microvascular blood supply to bone, precipitating osteonecrosis (avascular necrosis), which can lead to joint collapse. Glucocorticoids are the most common cause of secondary osteoporosis and the leading cause of non-traumatic osteonecrosis.22

Physicians actively manage these risks — for example with the lowest effective dose, limited duration, and bone-protective monitoring. None of that is something to attempt alone; it is a reason to keep your rheumatology follow-ups.

Drug-Induced Nutrient Depletions

Long-term management of metabolic and gastrointestinal conditions can quietly deplete nutrients that joints and bone depend on — the drug-induced nutrient depletion problem. Evidence varies by drug and nutrient, so the badges below differ.

Proton Pump Inhibitors (PPIs) Strong (magnesium)

PPIs such as omeprazole and pantoprazole suppress stomach acid, and long-term use can reduce the body's uptake of some nutrients. In a meta-analysis of 16 observational studies (131,507 people), 19.4% of PPI users had low magnesium versus 13.5% of non-users (adjusted odds ratio 1.71), with higher odds at higher doses.23 US prescribing information for PPIs lists hypomagnesemia as a warning with prolonged use.24 Two or more years of PPI use was also associated with vitamin B12 deficiency (odds ratio 1.65) in a large US case-control study.25 The PPI–magnesium link is well documented, so we grade it Strong. The further idea, common in integrative circles, that low magnesium removes a buffer against inflammation and worsens arthritis is a hypothesis that has not been tested in arthritis patients.

Metformin Strong (vitamin B12)

Metformin, the first-line drug for type 2 diabetes, lowers vitamin B12 over time. In the Diabetes Prevention Program Outcomes Study, low or borderline-low B12 was more common with metformin than placebo at 5 years (19.1% vs 9.5%) and at 13 years (20.3% vs 15.6%), and each year of metformin use raised the odds of B12 deficiency by about 13%. The authors suggested routine B12 testing in people taking metformin.26 This matters for joints mainly indirectly: B12-related nerve symptoms such as tingling, numbness or burning can be mistaken for joint pain. Do not stop metformin; ask your prescriber whether a B12 check is due.

Statins Contested (CoQ10 link)

Statins lower cholesterol by inhibiting HMG-CoA reductase in the mevalonate pathway — the same pathway that produces Coenzyme Q10 (CoQ10), which mitochondria use to make ATP.27 Statin-associated myopathy (muscle pain and weakness) is a recognized effect that can compromise joint stability and mimic an arthritis flare. However, the specific claim that CoQ10 depletion causes these musculoskeletal symptoms — and that CoQ10 supplementation reliably relieves them — is Contested. A 2018 meta-analysis of small trials reported less muscle pain with CoQ10,28 but a 2020 meta-analysis and a randomized trial in people with confirmed statin myopathy found no benefit over placebo.2930 Blinded trials also show that many muscle symptoms blamed on statins occur just as often on placebo: in a 2022 analysis of 19 double-blind trials, only about 1 in 15 muscle-symptom reports among statin users in the first year was actually due to the statin.31 Do not stop a statin without talking to your prescriber. Full guide: Statins and joint pain.

Thiazide & Loop Diuretics Strong (gout)

Thiazide and loop diuretics (such as hydrochlorothiazide and furosemide), used for hypertension and fluid overload, increase renal reabsorption of uric acid while depleting magnesium and potassium.27 The resulting rise in serum uric acid is a well-recognized iatrogenic trigger for gout. In a UK study of 24,768 people with new gout, current diuretic use was linked to 2.36 times the risk among those with hypertension,32 and in a second UK study of 91,530 cases the odds were 2.64 for loop diuretics and 1.70 for thiazides, but not raised for potassium-sparing diuretics.33 This is graded Strong. If you track your labs, see our guide to arthritis inflammation biomarkers, where uric acid and related markers and their optimal targets are explained.

A constructive next step: if any of these depletions concern you, the productive move is to discuss alternatives with your doctor — who can review whether a drug is still needed, monitor the relevant nutrient or marker, and, where appropriate, address depletions under medical supervision. Please do not self-prescribe supplements to "counter" a medication.

Master Reference Table

The table below summarizes the drug classes discussed above, their proposed mechanisms, the reported effect on joints or bone, and our evidence grade. Read every badge in context — and remember this is a reference, not a recommendation to change anything.

Medications associated with worsened arthritis or joint/bone risk (educational reference).
Drug class Examples Proposed mechanism Effect on joints / bone Evidence badge
NSAIDs Ibuprofen, naproxen, diclofenac COX inhibition lowers prostaglandins; some NSAIDs (e.g. naproxen, ibuprofen, indomethacin) reduce cartilage matrix synthesis in vitro, others do not Faster joint-space loss with indomethacin in two older trials; small or uncertain associations for other NSAIDs in observational studies Contested
Fluoroquinolones Ciprofloxacin, levofloxacin, moxifloxacin, ofloxacin, delafloxacin MMP-1/MMP-3 upregulation, ROS/oxidative stress, magnesium chelation impairing matrix adhesion Tendinopathy & rupture (FDA boxed warning); cartilage damage in juvenile animals; usually reversible joint symptoms in children Strong (tendon) Emerging (cartilage)
Corticosteroids (long-term) Prednisone, dexamethasone Osteoblast/osteocyte apoptosis, prolonged osteoclast survival, reduced calcium absorption, impaired bone microvasculature Glucocorticoid-induced osteoporosis; osteonecrosis / avascular necrosis (AVN) Strong
Proton pump inhibitors Omeprazole, pantoprazole Acid suppression; reduced uptake of magnesium and vitamin B12 Hypomagnesemia; B12 deficiency (any effect on arthritis itself is unproven) Strong (magnesium)
Metformin Metformin Reduced vitamin B12 status with long-term use B12 deficiency; nerve symptoms that can be mistaken for joint pain Strong (vitamin B12)
Statins Atorvastatin, simvastatin HMG-CoA reductase inhibition also lowers CoQ10 in the mevalonate pathway Statin-associated myopathy / musculoskeletal pain that can mimic arthritis flares Strong (myopathy) Contested (CoQ10 link)
Thiazide & loop diuretics Hydrochlorothiazide, furosemide Increased renal uric acid reabsorption; magnesium and potassium depletion Elevated uric acid → iatrogenic trigger for acute gout Strong (gout)
🧬

Root Causes of Arthritis

How oxidative stress, nutrient status, and metabolic dysfunction set the stage for joint damage — context for the depletions above.

Explore the root causes →

🩸

Biomarkers to Track

Uric acid, vitamin D, hs-CRP and other markers — what they mean and their optimal targets.

See the biomarkers →

🥗

Anti-Inflammatory Plan

Physician-supervised dietary and lifestyle approaches to discuss as part of a bigger picture.

Discuss alternatives with your doctor →

Frequently asked questions

Can NSAIDs make arthritis worse over time?

The evidence is mixed. Laboratory studies show some NSAIDs reduce cartilage matrix synthesis, and two older trials found faster joint-space loss with indomethacin, but for commonly used NSAIDs the human evidence is mostly observational and open to confounding. Mainstream guidelines still strongly recommend oral NSAIDs for knee and hip osteoarthritis, so this is a Contested topic. This page is educational only; never start or stop a medication without speaking to your prescribing physician.

Which medications are linked to joint or tendon problems?

Well-documented associations include fluoroquinolone antibiotics with tendinopathy and tendon rupture, long-term corticosteroids with osteoporosis and osteonecrosis, thiazide and loop diuretics with uric acid elevation and gout, and long-term PPIs and metformin with low magnesium or vitamin B12. Statins can cause muscle symptoms, though many are not actually caused by the statin. Discuss any concerns with your doctor.

Should I stop my arthritis medication?

No. This site does not give medical advice. Medication decisions belong to the prescribing physician, and you should never stop a prescribed medication without speaking to them first.

References

  1. Kolasinski SL, Neogi T, Hochberg MC, et al. 2019 American College of Rheumatology/Arthritis Foundation guideline for the management of osteoarthritis of the hand, hip, and knee. Arthritis Rheumatol. 2020;72(2):220–233. PubMed 31908163
  2. Dingle JT. The effects of NSAID on the matrix of human articular cartilages. Z Rheumatol. 1999;58(3):125–129. PubMed 10441838
  3. Attur M, Al-Mussawir HE, Patel J, et al. Prostaglandin E2 exerts catabolic effects in osteoarthritis cartilage: evidence for signaling via the EP4 receptor. J Immunol. 2008;181(7):5082–5088. PubMed 18802112
  4. Rashad S, Revell P, Hemingway A, Low F, Rainsford K, Walker F. Effect of non-steroidal anti-inflammatory drugs on the course of osteoarthritis. Lancet. 1989;2(8662):519–522. PubMed 2570233
  5. Huskisson EC, Berry H, Gishen P, Jubb RW, Whitehead J. Effects of antiinflammatory drugs on the progression of osteoarthritis of the knee. LINK Study Group. J Rheumatol. 1995;22(10):1941–1946. PubMed 8991995
  6. Reijman M, Bierma-Zeinstra SM, Pols HA, Koes BW, Stricker BH, Hazes JM. Is there an association between the use of different types of nonsteroidal antiinflammatory drugs and radiologic progression of osteoarthritis? The Rotterdam Study. Arthritis Rheum. 2005;52(10):3137–3142. PubMed 16200593
  7. Perry TA, Wang X, Nevitt M, Abdelshaheed C, Arden N, Hunter DJ. Association between current medication use and progression of radiographic knee osteoarthritis: data from the Osteoarthritis Initiative. Rheumatology (Oxford). 2021;60(10):4624–4632. PubMed 33502488
  8. Tanne JH. FDA adds "black box" warning label to fluoroquinolone antibiotics. BMJ. 2008;337:a816. PubMed 18632714
  9. Stephenson AL, Wu W, Cortes D, Rochon PA. Tendon injury and fluoroquinolone use: a systematic review. Drug Saf. 2013;36(9):709–721. PubMed 23888427
  10. Lewis T, Cook J. Fluoroquinolones and tendinopathy: a guide for athletes and sports clinicians and a systematic review of the literature. J Athl Train. 2014;49(3):422–427. PubMed 24762232
  11. US prescribing information (FDA-approved labeling) for fluoroquinolones, e.g. levofloxacin, moxifloxacin and delafloxacin: Boxed Warning; Warnings and Precautions (aortic aneurysm and dissection, blood glucose disturbances, psychiatric adverse reactions). U.S. National Library of Medicine, DailyMed. dailymed.nlm.nih.gov
  12. Pfister K, Mazur D, Vormann J, Stahlmann R. Diminished ciprofloxacin-induced chondrotoxicity by supplementation with magnesium and vitamin E in immature rats. Antimicrob Agents Chemother. 2007;51(3):1022–1027. PubMed 17210779
  13. Adefurin A, Sammons H, Jacqz-Aigrain E, Choonara I. Ciprofloxacin safety in paediatrics: a systematic review. Arch Dis Child. 2011;96(9):874–880. PubMed 21785119
  14. Corps AN, Harrall RL, Curry VA, Hazleman BL, Riley GP. Contrasting effects of fluoroquinolone antibiotics on the expression of the collagenases, matrix metalloproteinases (MMP)-1 and -13, in human tendon-derived cells. Rheumatology (Oxford). 2005;44(12):1514–1517. PubMed 16148020
  15. Menon A, Pettinari L, Martinelli C, et al. New insights in extracellular matrix remodeling and collagen turnover related pathways in cultured human tenocytes after ciprofloxacin administration. Muscles Ligaments Tendons J. 2013;3(3):122–131. PubMed 24367771
  16. Sendzik J, Shakibaei M, Schäfer-Korting M, Lode H, Stahlmann R. Synergistic effects of dexamethasone and quinolones on human-derived tendon cells. Int J Antimicrob Agents. 2010;35(4):366–374. PubMed 20034766
  17. Pouzaud F, Bernard-Beaubois K, Thevenin M, Warnet JM, Hayem G, Rat P. In vitro discrimination of fluoroquinolones toxicity on tendon cells: involvement of oxidative stress. J Pharmacol Exp Ther. 2004;308(1):394–402. PubMed 14569066
  18. Shakibaei M, Pfister K, Schwabe R, Vormann J, Stahlmann R. Ultrastructure of Achilles tendons of rats treated with ofloxacin and fed a normal or magnesium-deficient diet. Antimicrob Agents Chemother. 2000;44(2):261–266. PubMed 10639347
  19. Compston J. Glucocorticoid-induced osteoporosis: an update. Endocrine. 2018;61(1):7–16. PubMed 29691807
  20. Weinstein RS, Jilka RL, Parfitt AM, Manolagas SC. Inhibition of osteoblastogenesis and promotion of apoptosis of osteoblasts and osteocytes by glucocorticoids: potential mechanisms of their deleterious effects on bone. J Clin Invest. 1998;102(2):274–282. PubMed 9664068
  21. Weinstein RS, Chen JR, Powers CC, et al. Promotion of osteoclast survival and antagonism of bisphosphonate-induced osteoclast apoptosis by glucocorticoids. J Clin Invest. 2002;109(8):1041–1048. PubMed 11956241
  22. Weinstein RS. Glucocorticoid-induced osteoporosis and osteonecrosis. Endocrinol Metab Clin North Am. 2012;41(3):595–611. PubMed 22877431
  23. Srinutta T, Chewcharat A, Takkavatakarn K, et al. Proton pump inhibitors and hypomagnesemia: a meta-analysis of observational studies. Medicine (Baltimore). 2019;98(44):e17788. PubMed 31689852
  24. US prescribing information (FDA-approved labeling) for proton pump inhibitors, e.g. omeprazole: Warnings and Precautions (hypomagnesemia and mineral metabolism; cyanocobalamin [vitamin B12] deficiency). U.S. National Library of Medicine, DailyMed. dailymed.nlm.nih.gov
  25. Lam JR, Schneider JL, Zhao W, Corley DA. Proton pump inhibitor and histamine 2 receptor antagonist use and vitamin B12 deficiency. JAMA. 2013;310(22):2435–2442. PubMed 24327038
  26. Aroda VR, Edelstein SL, Goldberg RB, et al. Long-term metformin use and vitamin B12 deficiency in the Diabetes Prevention Program Outcomes Study. J Clin Endocrinol Metab. 2016;101(4):1754–1761. PubMed 26900641
  27. Mohn ES, Kern HJ, Saltzman E, Mitmesser SH, McKay DL. Evidence of drug–nutrient interactions with chronic use of commonly prescribed medications: an update. Pharmaceutics. 2018;10(1):36. PubMed 29558445
  28. Qu H, Guo M, Chai H, et al. Effects of coenzyme Q10 on statin-induced myopathy: an updated meta-analysis of randomized controlled trials. J Am Heart Assoc. 2018;7:e009835. PubMed 30371340
  29. Kennedy C, Köller Y, Surkova E. Effect of coenzyme Q10 on statin-associated myalgia and adherence to statin therapy: a systematic review and meta-analysis. Atherosclerosis. 2020;299:1–8. PubMed 32179207
  30. Taylor BA, Lorson L, White CM, Thompson PD. A randomized trial of coenzyme Q10 in patients with confirmed statin myopathy. Atherosclerosis. 2015;238:329–335. PubMed 25545331
  31. Cholesterol Treatment Trialists' Collaboration. Effect of statin therapy on muscle symptoms: an individual participant data meta-analysis of large-scale, randomised, double-blind trials. Lancet. 2022;400:832–845. PubMed 36049498
  32. Choi HK, Soriano LC, Zhang Y, Rodríguez LA. Antihypertensive drugs and risk of incident gout among patients with hypertension: population based case-control study. BMJ. 2012;344:d8190. PubMed 22240117
  33. Bruderer S, Bodmer M, Jick SS, Meier CR. Use of diuretics and risk of incident gout: a population-based case-control study. Arthritis Rheumatol. 2014;66(1):185–196. PubMed 24449584