Caffeinated energy drinks in the Canadian context: health risk assessment with a focus on cardiovascular effects

In healthy adult populations, the effect of caffeine on arterial BP is well-documented. Single-occasion intakes of caffeine higher than 250 mg, up to 400 mg, have been shown to significantly raise systolic blood pressure (SBP) by 3–15 mm Hg and diastolic blood pressure (DBP) by 4–13 mm Hg as reported in several studies and meta-analysis (EFSA 2015; Mesas et al. 2011; Nawrot et al. 2003; Nurminen et al. 1999; VKM 2015a). Despite this well-described response to single-occasion intakes of coffee or caffeine, current evidence does not support an association between habitual coffee consumption (200–300 mg of caffeine) and increased BP or between habitual coffee or caffeine consumption and an increased risk of cardiovascular diseases in hypertensive subjects (Mesas et al. 2011).

It is generally recognized that modest changes in heart rate (HR) following moderate consumption of caffeine has no clinical relevance, except when caffeine exposure is above 400–450 mg/day which may result in tachycardia (Nawrot et al. 2003). Caffeine exposure above 400 mg/day also does not impact ECG parameters such as PR interval, QRS duration, corrected QT interval, RR interval, or corrected QT interval dispersion (Nurminen et al. 1999). In addition, caffeine alone is not responsible for arrhythmia except perhaps at extremely high doses (900–1000 mg caffeine per day) (Voskoboinik et al. 2018).

A dose–response meta-analysis (Ding et al. 2014) and a prospective cohort study (Loomba et al. 2016) found no association between coffee consumption (up to 5 cups per day, equivalent to 500 mg caffeine, per day) and cardiovascular disease-related mortality (i.e., stroke, congestive heart failure and ischemia-related mortality). Several recent publications (Doepker et al. 2018; McGuire 2014; Temple 2019) and reports from food safety authorities (EFSA 2015; VKM 2015a) corroborate HC’s Recommended Maximum Daily Intakes (RMDIs) for caffeine of up to 400 mg/day for healthy adults, 300 mg/day for women who are planning to become pregnant, pregnant women and breast feeding mothers and 2.5 mg/kg bw/day for children (aged 4 to 13 years) (Nawrot et al. 2003). The most recent systematic review on the potential adverse effects of caffeine consumption in healthy populations, which took into consideration cardiovascular, behavioral, reproductive, developmental, calcium/bone and acute adverse effects, supports HC’s RMDIs (Wikoff et al. 2017).

Taurine naturally occurs in some food (notably meat and seafood), with mean dietary intakes estimated to vary between 40 and 400 mg/day (VKM 2015b). Human data on potential acute toxicity of taurine and long-term studies (>1 year) are not available. The available human studies related to taurine, while limited by the low number of participants, observations in non-healthy populations and short duration, did not report any significant adverse effects with a dose of taurine of 3 g per day for up to 7 weeks (Chauncey et al. 2003; Wojcik et al. 2010; Zhang et al. 2004), apart from gastrointestinal disturbances noted in 1 review (Wojcik et al. 2010). Clinical treatment with taurine at higher doses (10 g/day for 8.5 months) also showed no adverse effects (Pearl et al. 2014). In a more recent study (Schwarzer et al. 2018), no significant effects on systemic hemodynamics were reported in patients with portal hypertension treated for 4 weeks with a daily taurine regime of 6 g per day.

In a risk assessment article published in 2008, which considered the available published human clinical trials, Shao and Hathcock (2008), concluded that a dose of 3 g per day, is the most suitable endpoint for a reference dose for the safety of taurine. Taken together with a No-Observed-Adverse-Effect-Level (NOAEL) of 1000 mg/kg bw per day (equivalent to 70 g per day for an adult) based on animal studies (EFSA 2009), it is considered very unlikely that daily intakes of up to 3 g of taurine per day from CEDs (equivalent to 3 typical 250 mL cans, Table 1) would cause adverse health effects in healthy individuals. To be considered eligible for a TMA in Canada, HC has determined that CEDs should provide no more than 3 g taurine per day (Health Canada 2013).

There are very few publications that report on potential cardiovascular impacts of exposure to both caffeine and taurine. In 1 study (Bichler et al. 2006), where the main objective was to assess the potential impact of CED consumption on short-term memory, the combination of 100 mg caffeine with 1000 mg taurine caused a significant decline in HR (average decrease of 8.1 beats/min within 45 min following consumption). A significant increase in mean arterial BP (average increase of 2.8 mm Hg following consumption) was observed in subjects only after they took a memory test; the authors suggested a combination effect on BP may occur under stressful circumstances. However, whether the BP effect was a result of a synergistic effect of caffeine and taurine cannot be concluded from this 1 study, since the administration of caffeine alone, combined with psychological stress (e.g., taking a test) has also been shown to increase BP (Bichler et al. 2006).

Most published studies that looked at interactions between taurine and caffeine used CEDs as a comparator (Supplementary File S1¹). For example, 6 randomized double-blind control trials compared CEDs to caffeine alone when assessing effects on BP (Brothers et al. 2017; Doerner et al. 2015; Fletcher et al. 2017; Franks et al. 2012; Miles-Chan et al. 2015). One study reported significant SBP and DBP increase (Franks et al. 2012) while another reported significant SBP only (Fletcher et al. 2017) associated with CED intake. However, the other trials did not observe significant differences between CEDs and caffeine alone groups (Brothers et al. 2017; Doerner et al. 2015; Miles-Chan et al. 2015). A meta analysis summarizing these results suggests these effects are due to caffeine alone (Shah et al. 2016a). Despite limitations such as the small number of studies, limitations in study designs and inconsistency of the results, the available evidence does not support an impact of combined exposure to caffeine and taurine on the etiology of ischemic heart disease via thrombotic events (Higgins 2013; Pommerening et al. 2015; Worthley et al. 2010) or vasospasm (Abebe and Mozaffari 2011; Benjo et al. 2012; Hanan Israelit et al. 2012).

With respect to myocardial performance, several studies that included a caffeine control group (Baum and Weiss 2001; Doerner et al. 2015; Grasser et al. 2014) reported increased myocardial contractility 1–2 hours after consuming 1–2 servings of CED (average amount in 1 serving equivalent to 105 mg caffeine, 1304 mg taurine; 160 mg caffeine, 2000 mg taurine in 2 servings). A double-blind crossover study examined the effect of CEDs on contractile function (stroke volume and fractional shortening) of endurance athletes before and after exercise (Baum and Weiss 2001). In this study, 40 minutes before exercise 13 athletes consumed either the equivalent of 2 typical CEDs (2 × 250 mL, each providing 80 mg caffeine, 1000 mg of taurine), a water beverage containing caffeine (160 mg), a water beverage containing taurine (2000 mg) or sugar water (10.5 g). Contractile function among athletes consuming CEDs was significantly increased with exercise compared with consuming caffeine, taurine or sugar water alone, suggesting a combination effect. In a more recent study (Doerner et al. 2015), myocardial contractility using cardiac magnetic resonance was assessed in 32 healthy volunteers following CED consumption (average 105 mg caffeine, 1304 mg taurine). After a washout period of at least 1 week, 10 participants repeated the same protocol but consumed a beverage containing caffeine only (i.e., 105 mg caffeine, no taurine). The results showed a small, but statistically significant increase in left ventricular contractility, 1 hour after consumption, in volunteers who consumed a CED, but not in individuals who consumed the beverage that contained caffeine only. In summary, while some studies reported a modest increase in myocardial contractility following consumption of CEDs containing both caffeine and taurine supporting a possible synergistic role, the available data are too limited to determine the clinical significance of this observation.

There are no human studies regarding potential toxicity of glucuronolactone, which precludes the setting of a maximum level for CEDs. However, glucuronolactone has been used up to 3000 mg/day for chronic treatment of specific conditions, with no evidence of any adverse effects (Rotstein et al. 2013). At levels present in a typical 250 mL CED (up to 1200 mg, Table 1), it is considered unlikely that glucuronolactone would cause adverse effects, notably cardiovascular effects, in the general population. In recent years, there has been a decline in the use of glucuronolactone in CEDs. No relevant studies were identified on physiological effects of glucuronolactone combined with caffeine.

While there is insufficient information to establish a maximum level of inositol for CEDs, high doses of inositol appear well tolerated; specifically, a Canadian study established a NOAEL of 18 g inositol per day (Lam et al. 2006), and 4 g per day is commonly prescribed as a safe treatment for polycystic ovary syndrome (Carlomagno and Unfer 2011). Based on the range of levels generally present in CEDs (up to 200 mg per serving, Table 1), inositol is not expected to cause any adverse effects, including cardiovascular effects, for the general Canadian population. As with glucuronolactone, the addition of inositol in CEDs has declined in recent years. No relevant studies were identified that addressed potential interactions of inositol with caffeine.

No significant impacts on the cardiovascular system have been reported for B vitamins at the daily maximum levels permitted in CEDs sold in Canada (Rotstein et al. 2013).

While some studies showed that caffeine may contribute to an increase in the concentration of homocysteine, a risk factor for cardiovascular disease, it is also reported that B vitamins can lower homocysteine levels without any major adverse effects on endothelial function (Maruyama et al. 2019). This combination effect requires further investigation; however, there are no published data available that address the consumption of CEDs and effects on plasma concentrations of total homocysteine.

The RCTs that reported effects of CEDs on BP are summarized in the supplemental material (Supplementary File S1¹). Overall, available data for BP are consistent with studies showing a modest effect of CEDs increasing both SBP and DBP. It is difficult to precisely quantify the contribution of caffeine to the BP effect that has been reported following CED consumption. However, a meta-analysis published in 2016 (Shah et al. 2016a) focused on the impact of CED consumption on BP parameters. It included 15 studies with a total of 340 individuals assessed for SBP effects and 14 studies with a total of 322 individuals for DBP effect. The SBP and DBP increased by 4.44 mm Hg (IC_95% = 2.71 to 6.17) and 2.73 mm Hg (IC_95% = 1.52 to 3.95) on average respectively, which was statistically significant. The authors attributed the increase in BP to the caffeine content. There also appeared to be a dose effect because SBP was elevated by 3.7 mm Hg when the caffeine content of the CED was <200 mg (equivalent to 2 or less CEDs) and by 6.4 mm Hg when ≥200 mg (equivalent to 3 or more CEDs). However, no such dose-related change was observed for DBP. As normal BP is less than 120 mm Hg (SBP) and 80 mm Hg (DBP) a temporary increase of BP such as that observed following intakes of less than 200 mg of caffeine is not expected to be clinically significant for healthy adults.

Most of the 19 RCTs reported no significant change in HR after consuming CEDs (refer to Supplementary File S1¹). These results were confirmed by Shah et al.’s meta-analysis (Shah et al. 2016a) where HR data for 340 individuals showed no significant change after CED consumption compared with baseline (0.8 beats/min; IC_95% = –1.26 to 2.87). Changes observed in HR in some of the RCTs are consistent with moderate consumption of caffeine; however, these effects are expected to have no clinical impact, except when caffeine doses are greater than 400–450 mg/day. At these doses, symptoms such as tachycardia or palpitations can occur, which are endpoints most often reported in observational studies.

CEDs have been linked to several adverse effect reports, notably cardiovascular arrhythmias. In a review of published case-reports of adverse cardiovascular events after consumption of CEDs (Goldfarb et al. 2014), 15 cases were identified, including 5 atrial arrhythmias, 5 ventricular arrhythmias, 1 QT prolongation and 4 ST segment elevations; a predisposing cardiac abnormality was not identified in the majority of cases. One study reported QT interval prolongation at a relatively high volume of CED consumption (950 mL consumed over a 45-minute period, providing 320 mg caffeine) (Shah et al. 2014) in healthy volunteers. Other case reports of cardiovascular symptoms, in patients with long QT syndrome (LQTS), were reported but at much higher doses (1500 to 1900 mL of a CED, providing 480 to 640 mg caffeine within 4 hours) (Berger and Alford 2009; Rottlaender et al. 2012).

Of the 19 RCTs identified, 7 investigated the impact of CED intake on ECG corrected QT interval (QTc), to allow comparison of QT values over time at different heart rates and improve detection of patients at increased risk of arrhythmias. Of those 7, 1 study (Gray et al. 2017) investigated a small population of patients (n = 24) with pre-existing genetically confirmed Long QT Syndrome (LQTS) and did not observe any significant differences in the QTc nor in any other ECG parameters in comparing patients who consumed CEDs (250 mL × 2 within 30-min interval; equivalent to 160 mg caffeine, 2 g taurine) versus a placebo. Three individuals either with a severe history of cardiovascular incidents or severe phenotype of LQTS (QTc interval >500 ms) presented with an increased QT interval of at least 50 ms following CED consumption, indicating these sensitive individuals may be at increased risk when consuming 2 typical 250 mL CEDs within 30 min. However, there are several limitations with this study (Gray et al. 2017) that make these conclusions difficult to interpret. Notably the fact that 83% of the patients were being treated with beta-blockers during the study, the small sample size, the absence of plasma concentration-time curves for caffeine and taurine, and only 1 dose of CEDs was provided to patients, thus a dose effect cannot be confirmed. Lastly, the existence of several LQTS pathogenic variants in the selected study population, with possible differences in LQTS severity, greatly limits any extrapolation of these results to the general population. Given the limitations of the data, this genetically confirmed population should likely exercise caution when consuming CEDs.

The other 6 studies looked at young adult populations (average 20–29 years) with no identified underlying disease. Significant changes in QT interval was reported in 2 trials (Fletcher et al. 2017; Shah et al. 2016c), 2 hours after consuming a high volume of CEDs (950 mL ≈ 4 cans of 250 mL, composition not provided). The first study (Fletcher et al. 2017) observed that the change in QTc from baseline in the CED group (950 mL) was significantly higher than the caffeine-alone group but only at the 2-hour time point following consumption (0.44 ± 18.4 ms vs –10.4 ± 14.8 ms; p = 0.02). No difference was observed at any other time period (1, 4, 6 or 24 hours following consumption). In the second study (Shah et al. 2016c), an increase in QTc interval of 3.4 ms, 2 hours after CED consumption (950 mL) was observed compared with a decrease of –3.2 ms in the placebo group (p = 0.03). Similar to the first study, the results were not significant at any other time period (1, 3.5 or 5.5 hours following consumption). For comparison, the U.S. Food and Drug Administration’s guidance for clinical evaluations of QT/QTc interval prolongation associated with drugs indicates additional follow-up is needed only when a change in QTc interval of greater than 10 ms is observed on ECG (FDA 2005, 2017). For the 2 studies cited (Fletcher et al. 2017; Shah et al. 2016c), despite the significant differences reported between each group, the increase in QTc intervals from baseline, 2 hours after consumption of CEDs, was less than 10 ms suggesting that the effect was transient and reversible with no long term clinical impact. The remaining 4 studies that explored QT intervals did not report any significant changes between CEDs and caffeine only (Brothers et al. 2017) nor CEDs and placebo (Basrai et al. 2019; Ragsdale et al. 2010; Shah et al. 2016b). Based on the available data, there is some evidence that higher CED intakes (equivalent to 4 servings of a typical 250 mL CED) in healthy adults may increase QTc intervals; however, the clinical significance of these observations is unclear. Additional studies are required to investigate the potential electrocardiographic effects of CEDs in people with pre-existing cardiac diseases and, to a lesser extent, in healthy adults.

In Canada, post-market surveillance data on reports of suspected adverse reactions to CEDs has been collected via 2 different regulatory frameworks. From 2004 until 2012, when CEDs were regulated as NHPs, adverse reactions were captured under the Canada Vigilance Program, which has been in place since 1965. Since their transition to the food regulatory framework in 2012, manufacturers and distributors of CEDs who have been issued TMAs are required to annually report all consumption incidents associated with their products to HC.

Between approximately 2004 (the date of licensing of the first CED as an NHP) and 2012, 74 adverse reaction reports were reported in the Canadian Vigilance Program (Table 2). Cardiac disorders were reported in 35% of the cases (n = 47).

Since their transition to the food framework in 2012, 113 adverse reactions (referred to as consumption incidents for food products) related to CEDs were reported to HC under the TMA process (Table 3). Cardiac symptoms were reported in 9 cases (7.9% of total cases) and 3 were reported as serious incidents. Two were fatal due to cardiac arrhythmia and acute coronary artery thrombosis, and involved use of alcohol and recreational drugs. The third serious case was a pericarditis. This latter case reported consuming only 1 can (250 mL) of CED with alcohol. The symptoms reported in the other 6 non-serious cardiac cases, included irregular heart rhythm (1 case), chest pain (4 cases), and increased heart rhythm (1 case).

Regardless of the reporting system (Canadian Vigilance Program or TMAs), adverse event reports following CED consumption present many limitations. The reports often lack information on demographics, quantity or details of CEDs consumed, chronic versus acute exposure to CEDs, involvement in strenuous activities, other substances consumed with CEDs, and pre-existing medical conditions. Although the reporting of adverse reactions associated with CEDs is mandatory for TMA holders, reporting is voluntary for consumers and health professionals. Most of the adverse reactions are reported directly by consumers and are in most cases not validated by a physician. Therefore, this data is not sufficient to support the establishment of a cause and effect relationship between CEDs and the reported adverse reaction.

From 2012 to 2018, 852 calls related to multiple exposure (i.e., not only CEDs) and 281 calls related to single exposure involving CEDs, were received by Canadian poison information centres. For both types of exposure, the number of calls per year was relatively consistent between 2012 and 2016, with an increase in the number of calls in 2017 and 2018 (data not shown). Among the 281 single exposure calls, 1.1% reported a major effect, 7.8% a moderate effect, 14.2% a minor effect, and 77% reported minimal clinical effects. However, it was not possible to obtain data on specific clinical outcomes or symptoms. No fatal case was reported for single exposures. Among the 852 multiple exposure reports, similar clinical outcome numbers are reported, and there was 1 fatal case caused by an intentional overdose resulting in cardiac arrest where the caffeinated beverage was considered a co-exposure (CED was used to ingest the overdose drugs).

The reason for exposure was reported to be unintentional in 80.4% and 51.4% of the calls, for single and multiple exposures, respectively. Most of the calls were reported in children younger than 5 years of age, and represent 49.1% and 30.8% of total calls, for both types of exposure. This high proportion of cases involving unintentional exposures by children aged less than 5 years has also been reported in the US (Seifert et al. 2013).

Several other international studies (Borron et al. 2018, Gunja and Brown 2012; Markon et al. 2019, Seifert et al. 2013) provide analysis of adverse reactions related to CED exposures based on calls to poison information centres. A study based on the Texas poison center network (TPCN) reported data from single substance exposures to CED (multiple product types or CEDs in combination with other substances such as alcohol were excluded) from 2010 to 2014 (Borron et al. 2018). Over this 5-year period, 855 documented CED exposures were reported out of a total of 752 279 calls to the TPCN, or 0.11%. An Australian study analysed data from calls to the New South Wales (NSW) poisons information centres, a mean frequency of 0.04% of CED related-calls were reported to the NSW poisons information centre (297 incidents for 770 000 calls) between 2004 and 2010 (Gunja and Brown 2012).

Observational studies identified in the literature search were obtained from emergency department visits and cross-sectional studies focusing on cardiac-related events following CED consumption in specific populations. The data collected relies on adverse reactions reported by patients who also reported consuming CEDs (Bashir et al. 2016; Busuttil and Willoughby 2016; Kamijo et al. 2018; Nordt et al. 2012, 2017). In these retrospective studies, subjects were questioned about previous consumption of CEDs, but the adverse effects reported as related to CEDs were not necessarily the reason for their visit to the emergency department. In 2 separate studies on emergency visits for possible cardiac-related issues related to CED consumption, palpitations were self-reported in 12%–16% of the participants who also reported having a history of CED consumption, chest pain in 3%–5% and dyspnea in 4% (Nordt et al. 2012, 2017). As for all retrospective investigations, these studies presented several biases, namely recall bias, since these symptoms may not be causally related to CED consumption only. In addition, these symptoms were self-reported by patients and not validated by physicians, which makes this type of survey at higher risk of over-interpretation. Lastly, other potential sources of caffeine intakes were not queried in some of these studies.

To avoid these biases, 1 study (Busuttil and Willoughby 2016) investigated 60 patients presenting in an emergency department with a complaint of palpitations (defined as “awareness of a fast or erratic heartbeat”) to determine the impact of CED clinical presentations in healthy adolescents and adults (13–40 years). Results showed that DBP was slightly higher in CED consumers while SBP and HR were not significantly different. In contrast, the percentage of abnormal ECG findings in CED consumers was lower than non-consumers, but this difference was not significant.

Ad hoc cross-sectional studies are often limited by small sample size (Kamijo et al. 2018; Pensa et al. 2016). To mitigate this issue, other studies (Attipoe et al. 2018; Hammond et al. 2018; Wiggers et al. 2019) have used online surveys with relatively large sample sizes, including a Canadian study (Hammond et al. 2018), which assessed potential adverse effects associated with CED consumption among youth (12–17 years) and young adults (18–24 years). This study found that 41.5% of the respondents reported experiencing at least 1 adverse reaction after consuming CEDs, while 30.6% reported experiencing at least 1 adverse reaction after drinking coffee. Of the possible adverse effects captured in the survey, the only cardiac-related effect, “fast heartbeat”, was reported in 24.7% for CED consumers versus 10.5% for coffee consumers. The difference between self-reported adverse effects among CED and coffee consumers was considered significant. In an Australian study (Costa et al. 2016), 399 adolescents (12–18 years) from 4 secondary schools completed a self-report electronic survey of consumption patterns and symptoms related to CEDs. Here the participants were presented with 8 short-term symptoms related to cardiovascular effects and possibly associated with CED consumption, including “racing heart” and “palpitations”. Overall, the response rate was 68% with 56% reporting lifetime CED consumption (n = 224) and, among them, 53.2% of adolescent consumers reported experiencing at least 1 of these symptoms following CED consumption. “Racing heart” was the primary symptom reported (41%), with 16% reporting “heart palpitations”. However, use of online survey tools are sometimes based on convenience samples, which maybe subjected to ascertainment bias and sampling errors (lack of quality random sampling). Further, low response rates observed (e.g., 5% in the Hammond study (Hammond et al. 2018) and 7% in the Attipoe study (Attipoe et al. 2018)) limit the generalizability of the data.

Overall, palpitations and tachycardia were generally the most frequently reported categories of cardiac symptoms reported by CED consumers. Although the results of the observational studies indicate that the proportion of participants who experience adverse effects following consumption of CEDs is frequent and/or common, the findings are not necessarily representative of the general population or of the specifically targeted populations such as young adults and adolescents due to the limitations of the observational studies. Furthermore, while the number of reported adverse reactions and the average annual frequency of CED-related calls to poison information centres are underestimated and insufficient to support a cause effect relationship; a rough comparison with the number of CED units sold each year in Canada (320 million units sold in 2019; source: HC internal document, 2019) provides a contrasting perspective on the frequency of occurrence of adverse effects following consumption of CEDs. More comprehensive data are needed to more fully understand the occurrence of adverse effects are following consumption of CEDs.