Caffeine and Pain ⇑ Jana Sawynok

Ò PAIN 152 (2011) 726–729

www.elsevier.com/locate/pain

Topical review Caffeine and pain ⇑ Jana Sawynok

Department of Pharmacology, Dalhousie University, Halifax, NS, Canada B3H 1X5

Sponsorships or competing interests that may be revelant to content are disclosed at the end of this article. article info abstract

1. Introduction 2. Caffeine as an adjuvant analgesic

Caffeine is widely consumed for its central nervous system Caffeine has been added to formulations containing aspirin, stimulant effects such as increased alertness and decreased fatigue. acetaminophen and other non-steroidal anti-inflammatory drugs It is present in a variety of beverages (coffee, tea, energy drinks) (NSAIDs) for some time. It was probably originally added to offset and some foods (chocolate, desserts); it is also available as a drug potential sedative effects of analgesics, but is now regarded as an where it is used as a stimulant, or is added to analgesics in over- adjuvant analgesic. Clinical studies published in the 1960s and the-counter formulations (Table 1). Estimated daily consumption 1970s indicated that caffeine-containing analgesics produced ef- is 168–220 mg/day in North America, and 390–410 mg/day in fects similar to those of the analgesics alone, but in the 1980s, Scandinavian countries [11]. Population-based surveys in the US an analysis of 30 unpublished studies (post-operative pain, head- confirm such intake levels (193 mg/day) and note high consump- ache) derived a relative potency of 1.41 for analgesics (aspirin, tion rates (85–95% of adults) [9]. Pharmacological actions of caf- acetaminophen) containing 65 mg caffeine, and this established feine are attributed primarily to block of adenosine A1,A2A and its reputation as an adjuvant analgesic [21]. More recent meta- A2B receptors; it has a lower affinity for A3 receptors [11]. Caffeine analyses of adjuvant actions of caffeine suggest that caffeine may also can inhibit phosphodiesterase, promote Ca2+ release and block be useful for enhancing relief of headache pain [43], but for post-

GABAA receptors, but concentrations required to elicit these effects surgical pain, caffeine adds little to the analgesic action of acetami- are one hundred times higher and unlikely to be reached by normal nophen and aspirin [41,42], while additive effects with ibuprofen are dietary use of caffeine [11]. inconsistent [27]. The effects of caffeine on pain are of interest for several reasons. The effects of caffeine in combination with acetaminophen, (1) Caffeine is present in analgesic formulations and exhibits adju- aspirin and other NSAID analgesics have been examined in detail vant properties, increasing the analgesic effect of the primary in a series of preclinical studies. An analysis using a single test, constituent. (2) Following chronic caffeine use, withdrawal head- the uric acid functional impairment model, examined several dose aches occur; on the other hand, habitual caffeine over-consumption combinations of caffeine with several analgesics [14]. In the acet- is associated with generation of headaches. (3) Caffeine, in lower aminophen–caffeine study, sixteen dose combinations were exam- doses than implicated in adjuvant analgesia, blocks antinociception ined (100-10, -18, -32, -56; 178-10, -18, -32, -56; 316-10, -18, -32, by several agents in preclinical studies, and dietary caffeine could -56; 562-10, -18, -32, -56 mg/kg) – antinociception was augmented interfere with the analgesic actions of these agents. (4) Adenosine in a mid-dose range particularly at acetaminophen 316 mg/kg and is implicated in antinociception resulting from acupuncture, and caffeine 32 mg/kg. A 3-zone model was proposed, whereby only in caffeine use may interfere with the effectiveness of this modality. the ascending part of a sigmoidal curve did caffeine augment the This topical review will address each of these caffeine pain stories. action of acetaminophen. Subsequent studies indicated caffeine 18–32–56 mg/kg also augmented antinociception by aspirin, several other NSAIDs [14] and ibuprofen [22] in this model. There is an extensive preclinical literature on intrinsic antinoci-

⇑ Tel.: +1 902 494 2596; fax: +1 902 494 1388. ceptive actions of caffeine, and this has been reviewed recently E-mail address: [email protected] [28]. While the effects of caffeine depend on the nature of the test,

chronic daily headache. Several studies from different countries Table 1 Human caffeine intake levels. report a significant correlation between caffeine consumption and headache [32]. Furthermore, a case study in children with Intake level Total intake (mg/day) Single dose (mg) chronic daily headache who consumed caffeine-containing bever- LOW <100 30–75 ages reported headache resolution when these were withdrawn MODERATE 101–200 76–150 [19]. These observations suggest that reduction in caffeine intake HIGH 201–1000 151–300 may be beneficial in leading to headache resolution in certain Caffeine content: coffee 100–200 mg, specialty coffees up to 500 mg, tea 25–50 mg, instances. soft drinks 30–50 mg, energy drinks 100–300 mg, frozen desserts 50–80 mg, drugs 65–200 mg (www.cspinet.org/new/cafchart.htm)[18]. 4. Caffeine and other analgesic drugs dose of caffeine, and species tested, several studies report intrinsic In the past decade, an increasing number of preclinical studies antinociception in different models generally at doses of 25– have reported that doses of caffeine lower than those that exhibit 100 mg/kg. Antinociception is attributed to block of adenosine adjuvant analgesic effects inhibit antinociception by several agents A2A and A2B receptors, and such actions may also contribute to (Table 2). These results are interpreted to reflect involvement of adjuvant analgesia [28]. Additional non-adenosine based mecha- adenosine A1 receptors because: (a) A1 receptor agonists produce nisms, such as changes in the activity and synthesis of cyclo-oxy- antinociception in nociceptive, inflammatory and neuropathic genase enzymes at certain sites also are implicated in adjuvant tests, and (b) selective A1 receptor antagonists mimic the blocking analgesia [10]. effect of caffeine. Low doses of caffeine inhibit antinociception by amitriptyline, venlafaxine, carbamazepine and oxcarbazepine, all 3. Caffeine and headaches of which are currently used, or are being explored, as analgesics for neuropathic pain in humans; they also inhibit actions of allopu- Acute administration of caffeine has been reported to produce rinol which is used to treat gout, cizolirtine which is an experimen- intrinsic analgesic actions in some headache studies [5,36].At tal analgesic, and even antinociception by acetaminophen in some higher doses (300–500 mg) caffeine relieved post-dural puncture models. headaches [4,40], but repeated administration of lower doses did The ability of low doses of caffeine to inhibit antinociception in not produce such an effect [6]. Vasoconstrictor actions of caffeine preclinical studies raises the possibility that dietary caffeine intake (secondary to adenosine receptor blockade) are implicated in relief could interfere with analgesic efficacy in a clinical setting. Ideally, a of headache [32]. clinical trial would evaluate, in a prospective manner, the effects of Following chronic caffeine ingestion, cessation of intake leads to caffeine on analgesic agents. However, given the widespread use of a withdrawal syndrome in which headache and fatigue are prom- caffeine, potential adaptive changes that occur following long-term inent; symptoms have an onset of 12–24 h, peak at 24–48 h, and use of caffeine, and challenges in recruiting trial participants who last about 1 week [16]. Headaches were originally recognized in would agree to give up caffeine consumption in order to partici- the context of withdrawal from high intake levels (>600 mg/day), pate in blinded trials, this approach may not be optimal for explor- but even moderate intake levels (100 mg/day) lead to withdrawal ing the issue. The experience with another drug in which headaches [15]. Caffeine withdrawal has been recognized as a fac- adenosine is implicated in its mechanism of action may be instruc- tor in peri-operative headaches, as prior to surgery, ‘‘nothing by tive for considering trial design. mouth after midnight” has been routine practice. Retrospective Methotrexate, in low doses, is an effective therapy for rheuma- studies reported a relationship between caffeine consumption toid arthritis. In a rat model of arthritis, it relieves the severity of and headaches in the perisurgical interval [8,37]. A subsequent symptoms by an action that involves adenosine A2A receptors on prospective study determined the effect of restoring caffeine con- inflammatory cells – caffeine completely suppresses this activity sumption on the day of surgery, and reported elimination of peri- [24,25]. In clinical studies, the effects of caffeine intake on metho- surgical headaches [17]. A further study investigated the effect of trexate actions in rheumatoid arthritis were initially examined by caffeinated beverages versus intravenous caffeine on the incidence monitoring caffeine consumption and dividing into low (89 mg/ of post-operative headaches and found both procedures reduced day), medium (151 mg/day) and high (259 mg/day) intake levels headache [38]. Collectively, these studies suggest that attention – subjects in the high intake group had less improvement than to caffeine intake, and restoration of intake on the day of surgery, those in the low intake group [26]. This study involved newly re- can be helpful in alleviating post-operative headaches. cruited subjects (N = 39) and a methotrexate dose of 7.5 mg/day. In contrast to the above scenarios, in which caffeine is benefi- A subsequent analysis of a larger number of subjects (N = 264) on cial in resolving headaches, habitual caffeine consumption is asso- established doses of methotrexate (16 mg/day) found no differ- ciated with development of headaches such as migraine and ences in outcomes between low (39 mg/day), medium (165 mg/

Table 2 Caffeine blocks antinociception by several different agents in preclinical models.

Agent Caffeine Test, species Reference Amitriptyline 1.5–7.5 mg/kg Spinal nerve ligation, rats [7] Amitriptyline 5 mg/kg Streptozotocin, rats [35] Amitriptyline 10 mg/kg Formalin, mice [29] Venlafaxine 5 mg/kg Hot plate, mice [39] Carbamazepine 5–20 mg/kg Inflammatory hyperalgesia, rats [34] Oxcarbazepine 5–20 mg/kg Inflammatory hyperalgesia, rats [34] Oxcarbazepine 10 mg/kg formalin, mice [30] Allopurinol 10, 30 mg/kg Intradermal capsaicin, mice [31] Cizolirtine 5 mg/kg Streptozotocin, rats [1] Acetaminophen 10 mg/kg Hot plate, tail flick, mice [12] Ò 728 J. Sawynok / PAIN 152 (2011) 726–729 kg) and high (422 mg/day) caffeine intake groups [2]. Similarly, can lead to headaches in some situations, and levels of caffeine in- analysis of subjects with psoriatic arthritis (N = 150) treated with take warrant direct clinical attention. These considerations expand methotrexate exhibited no difference in methotrexate dosing attention to effects of caffeine on pain well beyond its well-known requirements between low (<120 mg/day), moderate (120– role as an adjuvant analgesic. 180 mg/day) and high (>180 mg/day) caffeine intake levels [33]. The latter study further noted a wide range of doses of methotrexate to treat this condition, and no correlation between methotrex- Conflict of interest statement ate and caffeine intake levels. The pragmatic methodology used in these trials provides a means by which the potential influence of There are no conflicts of interest pertaining to this manuscript. caffeine intake on analgesic and other effects of established drug modalities could be examined. Acknowledgement 5. Caffeine and non-pharmacological modalities Studies performed in the author’s laboratory that are cited in this manuscript were supported by the Canadian Institutes of Acupuncture is used widely to manage chronic pain. The Health Research. mechanistic basis of experimental acupuncture, particularly elec- troacupuncture, involves multiple endogenous mediators (e.g. endorphins, monoamines, peptides, hormones) [44]. Acupuncture References in traditional Chinese medicine (manual acupuncture) involves insertion and twirling of needles and may involve further [1] Aubel B, Kayser V, Farré A, Hamon M, Bourgoin S. Evidence for adenosine- and peripheral mechanisms and mediators. Tissue analysis of the serotonin-mediated antihyperalgesic effects of cizolirtine in rats suffering from diabetic neuropathy. Neuropharmacol 2007;52:487–96. consequences of needle insertion and manipulation reveal [2] Benito-Garcia E, Heller JE, Chibnik LB, Maher NE, Matthews HM, Bilics JA, engagement of collagen fibres and fibroblasts and changes in their Weinblatt ME, Shadick NA. Dietary caffeine intake does not affect tension kinetics within connective tissue [20]. Recently, it was methotrexate efficacy in patients with rheumatoid arthritis. J Rheumatol demonstrated that needle insertion and manipulation leads to 2006;33:1275–81. [3] Cabalag MS, Taylor DM, Knott JC, Buntine P, Smit DV, Meyer A. Recent caffeine increased tissue levels of ATP, ADP, AMP and adenosine [13]. ingestion reduces adenosine efficacy in the treatment of paroxysmal Adenosine mediates acupuncture analgesia via A1 receptor activa- supraventricular tachycardia. Acad Emerg Med 2010;17:44–9. tion because inhibition of adenosine metabolism (using deoxyco- [4] Camann WR, Murray RS, Mushlin PS, Lambert DH. Effects of oral caffeine on postdural puncture headache. A double-blind, placebo-controlled trial. Anesth formycin to inhibit adenosine deaminase) augmented the Analg 1990;70:181–4. magnitude and duration of analgesia, and analgesia was elimi- [5] Diamond S, Balm TK, Freitag FG. Ibuprofen plus caffeine in the treatment of tension-type headache. Clin Pharmacol Ther 2000;68:312–9. nated by deletion of the A1 receptor gene [13]. Caffeine can block [6] Esmaoglu A, Akpinar H, Ugur F. Oral multidose caffeine-paracetamol adenosine A1 receptors in modest doses, and could inhibit acu- combination is not effective for prophylaxis of postdural puncture headache. puncture analgesia [45]. J Clin Anesth 2005;17:58–61. The implication of peripheral adenosine in acupuncture analge- [7] Esser MJ, Chase T, Allen GV, Sawynok J. Chronic administration of amitriptyline and caffeine in a rat model of neuropathic pain: multiple interactions. Eur J sia necessitates some new perspectives in evaluating the efficacy of Pharmacol 2001;430:211–8. acupuncture in clinical trials. Future studies should determine [8] Fenelly M, Galletly DC, Purdie GI. Is caffeine withdrawal the mechanism of whether chronic caffeine intake alters the effectiveness of acu- postoperative headache. Anesth Analg 1991;72:449–53. [9] Frary CD, Johnson RK, Wang MQ. Food sources and intakes of caffeine in puncture. While one might consider abstinence from caffeine for the diets of persons in the United States. J Am Diet Assoc 2005;105: an interval prior to an acupuncture session (e.g. for 4 h), the effects 110–3. of acupuncture on chronic pain are usually assessed after a series [10] Fiebich BL, Lieb K, Hüll M, Aicher B, van Ryn J, Pairet M, Engelhardt G. Effects of of sessions (e.g. twice weekly for 6 weeks), and effects on pain caffeine and paracetamol alone or in combination with acetylsalicylic acid on prostaglandin E2 synthesis in rat microglial cells. Neuropharmacol 2000;39: determined post-intervention and at several follow up intervals. 2205–13. In this setting, chronic intake patterns are likely of greater rele- [11] Fredholm BB, Bättig K, Holmen J, Nehlig A, Avartau EE. Actions of caffeine in vance. It is important to note that caffeine (200 mg) has been re- the brain with special reference to factors that contribute to its widespread use. Pharmacol Rev 1999;51:83–133. ported to inhibit the efficacy of transcutaneous electrical nerve [12] Godfrey L, Yan L, Clarke GD, Ledent C, Kitchen I, Hourani SMO. Modulation of stimulation, TENS, when given prior to the stimulation in a small paracetamol antinociception by caffeine and by selective adenosine A2 experimental pain trial (N = 17 subjects) in humans [23]. The po- receptor antagonists in mice. Eur J Pharmacol 2006;531:80–6. [13] Goldman N, Chen M, Fujita T, Peng W, Liu W, Jensen TK, Pei Y, Wang F, Han X, tential influence of chronic caffeine intake on TENS analgesia for Chen JF, Schnermann J, Takano T, Beker L, Tieu K, Nedergaard M. Adenosine A1 chronic pain indications therefore also warrants attention. In an- receptors mediate local anti-nociceptive effects of acupuncture. Nature other setting, prior caffeine intake (within 2–4 h) increased dose Neuroscience 2010;13:883–9. [14] Granados-Soto V, Casteñeda-Hernández G. A review of the pharmacokinetic requirements for i.v. adenosine used to treat supraventricular and pharmacodynamic factors in the potentiation of the antinociceptive effect tachycardias [3]. of nonsteroidal anti-inflammatory drugs by caffeine. J Pharmacol Toxicol 1999;42:67–72. [15] Griffiths RR, Evans SM, Heishman SJ, Preston KL, Sannerud CA, Wolf B, 6. Conclusions Woodson PP. Low-dose caffeine physical dependence in humans. J Pharmacol Exp Ther 1990;255:1123–32. [16] Griffiths RR, Woodson PP. Caffeine physical dependence. A review of human Caffeine is the most widely consumed psychoactive agent in the and laboratory animal studies. Psychopharmacology 1988;94:437–51. world. Given its widespread use, and the ability of low doses of caf- [17] Hampl KF, Schneider MC, Rüttiman U, Ummenhofer W, Drewe J. Perioperative administration of caffeine tablets for prevention of post-operative headaches. feine to interfere with antinociception in preclinical studies, it is Can J Anesth 1995;42:789–92. necessary to consider the effects of acute and chronic caffeine in- [18] Harland BF. Caffeine and nutrition. Nutrition 2000;16:522–6. take in several clinical settings. There are challenges to conducting [19] Hering-Hanit R, Gadoth N. Caffeine-induced headache in children and adolescents. Cephalgia 2003;23:332–5. prospective trials, but useful information can be obtained by sim- [20] Langevin HM, Bouffard NA, Badger GJ, Chruchill DL, Howe AK. Subcutaneous ply monitoring intake levels in a structured manner. Adenosine is tissue fibroblast cytoskeletal remodeling induced by acupuncture: evidence now implicated in analgesia by acupuncture and TENS, and the ef- for a mechanotransduction-based mechanism. J Cell Physiol 2006;207: 767–74. fect of caffeine intake on these non-pharmacological modalities [21] Laska EM, Sunshine A, Mueller F, Elvers WB, Siegel C, Rubin A. Caffeine as an also needs to be considered. Both caffeine intake and withdrawal analgesic adjuvant. JAMA 1984;251:1711–8. Ò J. Sawynok / PAIN 152 (2011) 726–729 729

[22] López JRM, Domínguez-Ramírez AM, Cook HJ, Bravo G, Díaz-Revel MI, Déciga- [33] Swanson DL, Barnes SA, Mengden Koon SJ, el-Azhary RA. Caffeine consumption Campos M, López-Munoz FJ. Enhancement of antinociception by co- and methotrexate dosing requirement in psoriasis and psoriatic arthritis. Int J administration of ibuprofen and caffeine in arthritic rats. Eur J Pharmacol Dermatol 2007;46:157–9. 2006;544:31–8. [34] Tomic´ MA, Vucˇkovic SM, Stepanovic´-Petrovic´ RM, Ugrešic N, Prostran MS, [23] Marchand S, Li J, Charest J. Effects of caffeine on analgesia from transcutaneous Boškovic´ B. The anti-hyperalgesic effects of carbamazepine and oxcarbazepine electrical nerve stimulation. New Eng J Med 1995;333:325–6. are attenuated by treatment with adenosine receptor antagonists. Pain [24] Montesinos MC, Desai A, Cronstein BN. Suppression of inﬂammation by low- 2004;111:253–60.

dose methotrexate is mediated by adenosine A2A receptor but not A3 receptor [35] Ulugol A, Karadag HC, Tamer M, Firat Z, Aslantas A, Dokmeci I. Involvement of activation in thioglycollate-induced peritonitis. Arth Res Ther 2006: 8: adenosine in the anti-allodynic effect of amitriptyline in streptozotocin- R53(doi: 10.1186/ar1914). induced diabetic rats. Neurosci Lett 2002;328:129–32. [25] Montesinos MC, Yap JS, Desai A, Posadas I, McCrary CT, Cronstein BN. Reversal [36] Ward N, Whitney C, Avery D, Dunner D. The analgesic effects of caffeine in of the anti-inflammatory effects of methotrexate by the nonselective headache. Pain 1991;44:151–5. adenosine receptor antagonists theophylline and caffeine. Arth Rheum [37] Weber JG, Ereth MH, Danielson DR. Perioperative ingestion of caffeine and 2000;43:656–63. postoperative headache. Mayo Clin Proc 1993;68:842–5. [26] Nesher G, Mates M, Zevin S. Effect of caffeine consumption on efficacy of [38] Weber JG, Klindworth JT, Arnold JJ, Danielson DR, Ereth MH. Prophylactic methotrexate in rheumatoid arthritis. Arth Rheum 2003;48:571–2. intravenous administration of caffeine and recovery after ambulatory surgical [27] Po ALW, Zhang WY. Analgesic efficacy of ibuprofen alone and in combination procedures. Mayo Clin Proc 1997;72:621–6. with codeine or caffeine in post-surgical pain: a meta-analysis. Eur J Clin [39] Yaba G, Sezer Z, Tekol Y. Interaction between venlafaxine and caffeine on Pharmacol 1998;53:303–11. antinociception in mice. Pharmazie 2006;61:60–2. [28] Sawynok J. Methylxanthines and Pain. In: Methylxanthines. Handbook of [40] Yucel A, Ozyalcin S, Talu GK. Intravenous administration of caffeine sodium Experimental Pharmacology, vol. 200; 2010. p. 311–29. benzoate for postdural puncture headache. Reg Anesth Pain Med [29] Sawynok J, Reid AR, Fredholm BB. Caffeine reverses antinociception in wild 1999;24:51–4.

type mice but not in those lacking adenosine A1 receptors. Neurosci Lett [41] Zhang WY, Po ALW. Analgesic efﬁcacy of paracetamol and its combination 2008;440:181–4. with codeine and caffeine in surgical pain – a meta-analysis. J Clin Pharm Ther [30] Sawynok J, Reid AR, Fredholm BB. Caffeine reverses antinociception by 1996;21:261–82.

oxcarbazepine by inhibition of adenosine A1 receptors: insights using [42] Zhang WY, Po ALW. Do codeine and caffeine enhance the analgesic effect of knockout mice. Neurosci Lett 2010;473:178–81. aspirin? A systematic review. J Clin Pharm Ther 1997;22:79–97. [31] Schmidt AP, Böhmer AE, Antunes C, Schallenberger C, Porciúncula LO, [43] Zhang WY. A beneﬁt-risk assessment of caffeine as an analgesic adjuvant. Elisabetsky E, Lara DR, Souza DO. Anti-nociceptive effects of the xanthine Drugs Safety 2001;24:1127–42. oxidase inhibitor allopurinol in mice: role of A1 receptors. Brit J Pharmacol [44] Zhao ZQ. Neural mechanism underlying acupuncture analgesia. Prog 2009;156:163–72. Neurobiol 2008;85:355–75. [32] Shapiro RE. Caffeine and headaches. Current Pain Headache Reports [45] Zylka MJ. Needling adenosine receptors for pain relief. Nature Neuroscience 2008;12:311–5. 2010;13:783–4.