Caffeine in the management of patients with headache

Caffeine is the most widely consumed psychoactive agent in the world [Citation removed]. It occurs naturally in more than 60 plant species, including tea, kola nuts, coffee beans, mate leaves, guarana plants, and cocoa nuts [15]. In humans, the most common sources of caffeine are dietary, as shown in Table 1 [16], although various medications also contain caffeine. Much of the caffeine used commercially is a byproduct of the production of decaffeinated coffee [17, 18].

Worldwide, approximately 80% of adults consume a caffeinated product every day [19], although daily mean consumption rates vary considerably by country. China (16 mg) and Kenya (50 mg) are at the low end of caffeine consumption and Brazil (300 mg) and Denmark (390) are at the high end [15]. In the United States, the vast majority of adults are regular caffeine users (90%), and estimates of daily consumption range from 168 to 220 mg [15].

Caffeine, the common name for 1,3,7-trimethylxanthine, is a purine alkaloid with the molecular formula C₈H₁₀N₄O₂, a molecular weight of 194.1906 g/mol, and the molecular structure of a trimethyl xanthine (Fig. 1) [18]. Pure caffeine takes the form of white, prismatic crystals; it is odorless, has a slightly bitter taste, and is slightly acidic, with a pH of 6.9 (1% solution) [18].

Caffeine is completely absorbed by the intestinal tract (ie, its bioavailability is 100%) [20] and it is highly soluble in water as well as a range of non-polar organic solvents [21]. When administered orally, caffeine takes 30–120 min to reach maximum plasma concentration, although food can slow the absorption process [15]. Caffeine crosses both the blood-brain and placental barriers [22] and it is secreted in human milk, saliva, bile, and semen [21]. About 95% of a caffeine dose is metabolized in the liver via cytochrome P450 (CYP1A2 demethylation), which converts it into paraxanthine (85%), theobromine (10%) and theophylline (5%) [23, 24]. Less than 2% of a caffeine dose is eliminated unchanged in urine [23]. In adults, caffeine’s elimination half-life is 3–5 h, but it is doubled in smokers and non-coffee drinkers; in newborns, the half-life of caffeine is estimated to be 100 h [21, 23].

The effects of caffeine on nociception are primarily attributed to its nonselective antagonism of the adenosine A₁, A_2A, and A_2B receptors (as caffeine and adenosine possess a similar molecular structure) and its slightly lower-affinity antagonism for the adenosine A₃ receptor [23, 25]. Through antagonism of the A_2A, and A_2B receptors, adenosine induces vasodilation [26, 27]. To combat increased antagonism of these receptors and vasodilation in chronic caffeine users, adenosine receptors are upregulated resulting in vascoconstrictive effects [28]. Additionally, the mechanism by which the adjuvant effects of caffeine occur is not well established although studies have shown that caffeine does not reduce prostaglandin synthesis, a previously suggested mechanism of action [29].

Caffeine dosing influences its pharmacodynamic effects. Higher doses (75–100 mg kg⁻¹) involve central noradrenergic mechanisms, medium to high doses (10–35 mg kg⁻¹) activate central amine systems, and low doses (5 mg kg⁻¹) interacting with central cholinergic mechanisms [30]. At therapeutic doses (ie, 100 times those typical in dietary consumption), caffeine can also induce phosphodiesterase inhibition, Ca²⁺ release, and GABA_A receptor blockade [23, 30].

Among patients with headache conditions, caffeine is used as an analgesic adjuvant. Analgesic adjuvants do not relieve pain by themselves, but augment the action of a known analgesic [31]. Factorial studies that examined the effect of analgesics alone, caffeine alone, and the combination of an analgesic and caffeine are required to assess its role as an analgesic adjuvant. In a 1984 meta-analysis of 30 such studies with more than 10,000 patients with postpartum uterine cramping, episiotomy pain, postsurgical pain, or headache, Laska et al., compared AAC-130 with APAP, ASA, and APAP + ASA to assess the therapeutic effect of caffeine as an analgesic adjuvant [32]. The estimated relative potency was 1.41 (95% CI; 1.23–1.63), indicating that caffeine significantly enhanced the analgesic effect of ASA, APAP, or a combination of the two by about 40% [32].

There is some evidence that caffeine monotherapy is associated with more pain relief than placebo in treating samples that include either migraine alone or persons with migraine or TTH [33, 34]. There is also limited evidence for the therapeutic use of caffeine monotherapy in patients with other primary and secondary headaches. For example, in patients with post-dural puncture headache (PDPH), studies have shown that acute use of 300 mg (oral) and 500 mg (intravenous) caffeine can provide relief [35]. A Cochrane review concluded that these doses of caffeine were superior to placebo at reducing the proportion of participants with PDPH persistence, as well as those needing supplementary interventions [35]. However, lower doses of caffeine (75 mg or 125 mg) combined with APAP 500 mg given every 6 h for 3 days have been shown to be ineffective for prophylaxis of PDPH [35], and a review of controlled and open-label studies investigating prophylaxis and treatment of PDPH with oral or IV caffeine cited methodologically weak trials and a paucity of reliable results in concluding that the use of caffeine as an antinociceptive agent for prevention or treatment of patients with PDPH is insufficiently supported by the available pharmacological and clinical evidence [36]. In patients with hypnic headache, an uncommon condition in elderly patients featuring stereotyped attacks of nocturnal headache for which outcomes from controlled studies are lacking, a review of data from observational studies found that the most effective acute treatment is caffeine (given either as 40–60 mg tablets or as a cup of coffee) [37].

When combined with analgesics, caffeine produces a range of clinical effects in these patients. Caffeine promotes the absorption of analgesics, including APAP, ASA, and IBU [38‐40]. This effect is attributed to rapid lowering of gastric pH. At low doses (≤10 mg kg⁻¹), caffeine has been shown to inhibit the antinociceptive effects of APAP [41], amitriptyline [42], carbamazepine and oxcarbazepine [43]. At higher doses (10–35 mg kg⁻¹), caffeine enhances pain relief with APAP [44, 45] and several different NSAIDs [46, 47]. Meta-analyses of caffeine combined with APAP, ASA, or IBU have found only weak adjuvant effects in patients with postoperative pain [48, 49].

Coadministration of caffeine can have other effects that are relevant for patients taking analgesics, including interfering with the effectiveness of some therapies. For instance, a 200-mg caffeine dose can inhibit the analgesic effects of transcutaneous electrical nerve stimulation [50], and consuming caffeine within 2–4 h of intravenous adenosine therapy for paroxysmal supraventricular tachycardia has been shown to necessitate the use of higher adenosine doses [51].

Caffeine has a wide range of physiologic effects that are unrelated to the direct treatment of headache pain but may influence improvement in headache symptoms. Multiple studies have shown that caffeine can enhance mood, alertness [23, 52‐55]; exercise performance [56, 57]; the speed at which information is processed, awareness, attention, and reaction time [58, 59].

Caffeine also increases gastric motility, as measured by pressure waves and propagated contractions in transverse/descending colon. The magnitude of the effect of caffeinated coffee is similar to eating a meal, 60% stronger than water, and 23% stronger than decaffeinated coffee [60]. Pharmacokinetic studies have shown that the absorption of caffeine is related to gastric emptying [61]. This effect of caffeine has important clinical implications for people with migraines, who may experience reductions in gastric motility both during and between attacks [62]. This reduction in gastric motility slows the absorption of acute medications and reduces their effectiveness [63]. In addition, widely used migraine-specific therapies, including the oral and injectable formulations of sumatriptan, significantly decrease gastric motility in healthy volunteers [64‐68], potentially exacerbating disease-related gastroparesis and, perhaps, decreasing or delaying oral absorption of triptans resulting in lower efficacy in more advanced attacks of migraine.

In the United States, the Food and Drug Administration considers caffeine a substance generally recognized as safe (GRAS) when used in 1) cola-type beverages in accordance with good manufacturing practice and 2) stimulant drug products [69‐71]. The caffeine contained in OTC analgesics is limited to a dose of 64–65 mg per tablet though two tablet doses are common [69‐72]. In clinical trials, caffeinated analgesics were generally well tolerated and caused the types of mild and transient adverse events (eg, upset stomach) that might be expected from single doses of analgesics at OTC doses [33, 34, 73, 74]. For example, one meta-analysis estimated that adding caffeine to analgesics significantly increased the number of patients who experienced nervousness and dizziness (relative risk 1.60) [75].

Studies also suggest that caffeine stimulates gastric acid production and relaxes the lower esophageal sphincter [76], increasing the risk of gastric and duodenal ulcers [77, 78] as well as gastroesophageal reflux disease [79]. However, a meta-analysis of studies enrolling more than 8000 healthy subjects detected no significant association between coffee intake and acid-related upper gastroduodenal ulcer diseases, including gastric and duodenal ulcers, reflux esophagitis, and non-erosive reflux disease [80]. The authors speculated that the known protective effects of coffee (relaxing effect, antioxidant effect, phytochemical effects) might outweigh the risks of increased gastric acid secretion [80].

When used in patients with renal failure, caffeine requires no dose adjustment [12]. However, as a matter of good clinical judgment, it is prudent to consider all analgesics potentially nephrotoxic and to counsel against excessive, protracted use [46, 81]. Excessive use of phenacetin-containing analgesics can cause renal papillary necrosis and interstitial nephritis, but there is no convincing evidence that nonphenacetin-containing analgesics, including acetaminophen (APAP), aspirin (ASA), and mixtures of these two compounds, or NSAIDs, are associated with pathologically or clinically defined renal disease [82].

The role of caffeine in the treatment of migraine is complex. Caffeine combination products are useful in treating migraine and other headache types [33‐35]. However, an association between caffeine intake and acute migraines has been observed [83]. Caffeine withdrawal may precipitate headache in the short term (NEJM) while cessation of caffeine may be beneficial. This suggests that complete abstinence of caffeine, like abrupt discontinuation of analgesics for those with MOH headaches [84], may be beneficial to those suffering from migraines [83]. Furthermore, overuse of APAP/caffeine/ASA has been shown to affect regional brain glucose metabolism in patients with chronic migraine [85].

Frequent use of analgesics is an important health problem [84], and medication overuse by patients with episodic headache conditions is associated with the development of chronic headache conditions [86‐89]. Avoiding medication overuse is an important aspect of headache care [3]; once established, medication overuse may be difficult to treat, and recidivism is common [90]. Some research has identified dietary and medicinal caffeine consumption, specifically OTC caffeine combinations, as a modest risk factor for progression from episodic to chronic headache [91, 92]. After adjustment for demographic factors, primary headache type, and number of medications taken, the association of caffeine-containing combinations, and opioid-containing combinations as a risk factor for migraine progression was attenuated [92]. Additional support for the lack of positive association between risk of migraine chronicity and the appropriate use of caffeine-analgesic combinations came from the American Migraine Prevalence and Prevention study, which reported that AAC-130 had no association with development of chronic migraine. After adjusting for headache frequency and medication use, AAC-130 was not associated with an increased risk of chronic migraine onset (OR = 0.87, 95% CI 0.64,1.19) [84]. Acute treatment guidelines recommend that in all cases and with all acute medications, migraineurs should limit medication use to 2–3 headache days per week [8]—advice that may apply equally to patients with TTH [93].

Chronic intake of caffeine can cause a dependence syndrome [94, 95] that has also been positively associated with a history of cigarette smoking and problematic alcohol use [96]. However, compulsive drug-seeking behavior has never been reported among caffeine-dependent persons, and reports of deliberate caffeine abuse are uncommon [97]. It has been suggested that the prevalence of caffeine dependence may increase as a result of the promotion of caffeinated energy drinks to adolescents [98], and cases of intoxication—characterized by nausea/vomiting, tachycardia, hypertension, agitation, and dizziness—have been reported [99]. Caffeinated energy drinks have also been associated with seizures [100], stroke [101], and, rarely, death [102]. These drinks are also frequently combined with alcohol, a combination that has been found to increase alcohol consumption in laboratory animals [103]. Perhaps as a consequence, consumption of alcohol mixed with caffeine has been found to be associated with high-risk behaviors such as binge drinking [103].

Tolerance to the effects of caffeine on blood pressure, heart rate, diuresis, adrenaline and noradrenaline plasma levels, renin activity, and sleep patterns generally occurs within a few days [22], and regular daily doses of 750–1200 mg can lead to tolerance to its subjective, pressor, and neuroendocrine effects [104‐106]. In addition, a withdrawal syndrome characterized by headache and sometimes other symptoms (ie, decreased cognitive performance, nausea/vomiting) may occur 12 to 24 h after stopping chronic consumption of caffeine in about 50% of patients [22, 31].

Multiple factors, including heritability [107] may contribute to an individual’s susceptibility to caffeine withdrawal [108]. The amount of caffeine consumption required to trigger withdrawal varies; while formal diagnostic criteria for caffeine withdrawal headache require daily consumption of ≥200 mg for at least 2 weeks [1], symptoms have been observed after sudden termination of caffeine exposures of 300 mg for 3 days and 100 mg for 7 days [109].

Some evidence suggests that caffeine withdrawal may play a role in triggering migraine attacks [110], especially weekend attacks [111]. Data also suggests caffeine withdrawal plays a role in dialysis headaches, a frequent complication of hemodialysis [112]. In addition, postoperative headaches may be partially due to caffeine withdrawal triggered by the abstinence typically required before surgical procedures [113]. Caffeine withdrawal headache can be minimized by a staged reduction and elimination of caffeine intake [97].

In 4 randomized, double-blind, 2-period crossover studies involving 1900 patients with TTH, Migliardi et al. compared AAC-130 with APAP alone and placebo [74]. Efficacy endpoints included pain intensity difference (PID), pain relief (PAR), and total pain relief (TOTPAR). AAC-130 provided significantly greater PID and PAR than APAP alone (P < .001), and both active treatments were superior to placebo (P < .001). As shown in Table 2, the magnitude of the adjuvant effect from caffeine for the summed pain intensity difference (SPID), %SPID, and TOTPAR was between 76% and 97%. For measures of peak analgesia and duration of analgesia, the effects were 63–85% greater than the net analgesic effect of APAP monotherapy [74]. Compared with APAP 1000 mg plus placebo, AAC provided significantly superior total pain relief at 4-h postdose (P < .001 vs. APAP and placebo) [74].

Because Migliardi et al. only restricted consumption of dietary caffeine (ie, coffee, tea, soda, chocolate) during the 4-h study period and found significant adjuvancy at the high and low ends of consumption, they concluded that the adjuvant effect was not influenced by patients’ usual caffeine consumption. Since the magnitude of caffeine’s adjuvant effect was larger for headache than in other forms of pain (ie, episiotomy, postsurgical, etc.) [32], caffeine-containing analgesic combinations may be particularly effective for headache pain [74].

In a randomized, double-blind, parallel-group, multicenter, single-dose, 4-arm, placebo- and active-controlled study, 301 patients with TTH treated a single attack with a 2-tablet dose of IBU 400 mg + caffeine 200 mg (IBC-100), IBU 400 mg alone, caffeine 200 mg alone, or placebo [34]. IBC-100 demonstrated significantly shorter times to meaningful improvement in headache relief than IBU alone or placebo; significantly greater total analgesia than IBU alone, caffeine alone, or placebo; and significantly greater peak relief than IBU alone, caffeine alone, or placebo (P < .05). Significantly more subjects obtained meaningful headache relief with IBC-100 than with IBU alone or placebo (P < .05). IBC-100 provided greater analgesic effectiveness than either component alone [51].

The efficacy of AAC-130 in migraine was established by Lipton et al. in 3 double-blind, placebo-controlled, parallel-group studies that randomized a total of 1220 migraineurs, excluding subjects who vomited more than 20% of the time or who usually required bed rest [73]. At 2-h postdose, pain-free rates were 21% for AAC-130 and 7% for placebo, and headache response (pain reduced from moderate or severe to mild or none) was 59% for AAC-130 and 33% for placebo. At 6-h postdose, 51% of patients treated with AAC-130 were pain-free compared with 24% of placebo-treated patients, and 79% of patients treated with AAC had a headache response versus 52% of those treated with placebo (P < .001 for all comparisons). The significant advantages of AAC-130 over placebo were also seen at 2- and 6-h postdose for relief of pain, associated symptoms, and disability (Table 3) [73]. Vomiting was less frequent in those treated with AAC-130 (0.2%) than placebo-treated patients (1.6%; P = .01) [73]. Although this study confirmed that a fixed combination containing caffeine could effectively treat a select population of migraineurs, the adjuvant effect of caffeine could not be assessed because a factorial design was not used.

In a 3-arm, double-blind, parallel group study of migraine, Goldstein and coworkers (2006) treated 1555 patients with a fixed combination of AAC (acetaminophen 500 mg, acetylsalicylic acid 500 mg, and caffeine 130 mg), IBU 400 mg or placebo. In comparison with placebo-treated participants, both active treatment arms demonstrated better relief of pain and associated symptoms. AAC was superior to IBU for a number of endpoints, including TOTPAR at 2 h, time to onset of meaningful PAR, pain intensity reduction, headache response, and pain free. The mean TOTPAR2 scores were 2.7 for AAC, 2.4 for IBU, and 2.0 for placebo (AAC vs IBU, P < 0.03). The median time to meaningful PAR for AAC was 20 min earlier than that of IBU (P < 0.036). The influence of caffeine once again could not be assessed as this was not a factorial study [114].

Two studies evaluated the efficacy of analgesics and caffeine in study populations that included patients with TTH and migraine [33, 115]. In the first, Diener et al. studied AAC-100 (APAP 200 mg, ASA 250 mg, and caffeine 50 mg), which is available OTC in Germany, in patients with TTH and migraine using a 6-arm, randomized, double-blind, parallel-group factorial design. The fixed combination AAC-100 was compared with APAP 200 mg and ASA 250 mg, APAP alone, ASA alone, caffeine alone and with placebo in 1743 subjects who typically treated their headaches with OTC medications [33]. Pain intensity was recorded on a 100-mm visual analogue scale. Among enrolled subjects, 84% had migraine, 13% had episodic TTH, and 3% could not be classified. The primary endpoint was time to 50% pain relief (PAR).

Time to 50% PAR, AAC-100 was significantly better than the APAP 200 mg-ASA 250 mg combination (P = .02), ASA 250 mg (P = .04), APAP 200 mg (P = .002), caffeine 50 mg (P < .0001), and placebo (P < .0001). All active treatments, except caffeine 50 mg, were significantly superior to PLA (P < .0001) [33]. The median time to 50% PAR was 65 min with AAC-100, 73 min with APAP 200 mg-ASA 250 mg, 79 min with ASA 250 mg, 81 min with APAP 200 mg, 107 min with caffeine 50 mg, and 133 min with placebo. Additional efficacy results comparing AAC-100 with the analgesic combination without caffeine and its constituents are shown in Table 4 [33]. It was concluded that AAC-100 significantly outperformed all the other treatments for reductions in pain intensity, and that the clinically meaningful improvements compared with APAP or ASA monotherapy confirmed the existence of an adjuvant effect of caffeine in migraineurs. Diener et al. also pointed out that, because their patient population typically treated their headache attacks with OTCs, findings can be generalized to the larger population of migraineurs and patients with TTH [33].

A double-blind, crossover pilot study comparing the efficacy of an IBU-caffeine combination with an IBU-placebo combination in 12 children (7 girls, 5 boys) aged 7 to 15 years (mean = 11.9 years) who had TTH (25%) or migraine (75%) was conducted [116]. Study treatments, adjusted to subjects’ weight, included 100 mg IBU + 50 mg caffeine; 200 mg IBU + 50 mg caffeine; 300 mg IBU + 100 mg caffeine; or 400 mg IBU + 100 mg of caffeine. Baseline pain intensity, which was assessed on a 4-point scale (4 = severe, 3 = moderate, 2 = mild, 1 = no pain), was 3.8 in the IBU-caffeine treatment group and 4.0 in the IBU-placebo group [116].

A total of 58% of children obtained faster relief with the IBU-caffeine combination, while 33% appeared to achieve earlier relief with IBU and placebo, and 8% exhibited no difference in the response [116]. Time to meaningful relief (pain intensity ≤2) was about twice as fast (60 min) for attacks treated with IBU-caffeine as it was for those treated with IBU-placebo (≈120 min). At 30 min postdose, PID in the IBU-caffeine group was 1.0 compared with 0.2 in the IBU-placebo group; at 60 min postdose, PID scores were 1.5 for IBU-caffeine and 0.9 for IBU-placebo. Although the IBU-caffeine combination failed to separate significantly on any efficacy endpoint in this small pilot study, it was numerically superior to the IBU-placebo combination, and almost 60% of patients realized clinically important benefits [116].