The hypoglycemia of little Osvaldo

Hypoglycemia and beta-blockers: new evidence from the real world and the scientific literature

Hypoglycemia is a known adverse reaction to beta-blockers. The hypothesized mechanism includes direct inhibition of hepatic gluconeogenesis and pancreatic glucagon release, probably mediated by antagonism on β-2 adrenergic receptors.^2,3,4,5,6
Infants and children are at higher risk for hypoglycemia because of increased glucose utilization in fasting states and lower glycogen stores.⁷
To date, it remains unclear which drugs in this class are most frequently associated with the induction of hypoglycemia and whether their pharmacological characteristics play a potential role in the induction of this adverse reaction. With the aim of quantifying the association between exposure to different beta-blockers and the risk of hypoglycemia reporting in pediatric patients, we performed a nested case-control analysis of suspected adverse reaction reports notified in the US FAERS spontaneous reporting database.
We, therefore, extracted reports entered from 2000 to 2019 in which at least one beta-blocker was involved. Of the 117,963 cases identified, after manually excluding reports with incomplete information and duplicates, 83,954 reports were selected; among these, 1,465 reported hypoglycemia, of which more than 20% involved the pediatric population.
For each individual beta-blocker, the Reporting Odds Ratio (ROR) and its 95% confidence interval (IC95) were estimated. The ROR, based on what was spontaneously reported during pharmacovigilance activities, measures the risk of experiencing an adverse drug reaction. It is therefore a parameter capable of quantifying the association between the use of a drug in daily clinical practice and the increase (statistically significant or not) of the reporting activity. If the ROR is statistically significant, it means that the number of reports of adverse reactions involving that drug exceeds what we would have expected on the basis of previous knowledge; this concept of disproportionality is an indication of a possible “signal” of safety.
Statistically significant RORs were found for nadolol (ROR: 6.98, IC95 5.40 to 9.03), followed by celiprolol (ROR: 2.35, IC95 1.35 to 4.10), propranolol (ROR: 2.14, IC95 1.87 to 2.46), and bisoprolol (ROR: 1.42, IC95 1.25 to 1.61).
After removal of reports with other suspected drugs besides beta-blockers or concomitant drugs for which the hypoglycemic effect is known, nadolol and propranolol not only maintained statistical significance but the disproportionate signal increased, albeit with wide confidence intervals, 14.95 (IC95 10.62 to 21.06) and 5.67 (IC95 4.53 to 7.11), respectively.
To investigate which pharmacodynamic and pharmacokinetic parameters of beta-blockers are potentially associated with the induction of hypoglycemia, a novel statistical approach was employed by coupling data from pharmacovigilance analyses with the different selectivity for β-1 receptors and the different half-life of each drug belonging to this class.
Because pediatric reports were found to be significantly statistically more associated with hypoglycemia than adults (p=0.0001), only cases and noncases that emerged in children were selected and paired 1:1 for age, sex, and concomitant hypoglycemic drugs.
On this selected cohort, the possible association between the two pharmacological parameters and the risk of hypoglycemia reporting was estimated by logistic regression in the form of an adjusted Odds Ratio (aOR). It was found that a half-life longer than 10 hours was positively associated with the risk of hypoglycemia reporting (aOR: 2.23). In contrast, selectivity for β-1 was found to be associated with a decreased risk of hypoglycemia.
Finally, a systematic review of the literature was conducted to compare evidence from a clinical practice setting with data provided by clinical trials. Because of the high heterogeneity of the studies, a wide variation in the frequency of occurrence of hypoglycemia (0.5% to 10%) was observed, with the incidence rate decreasing as the sample size increased.

Thus, the FAERS analysis suggests that nadolol is associated with a 6-fold increased risk of hypoglycemia reporting compared with other beta-blockers. In this regard, no data exist in the literature: to date, this is the first analysis to demonstrate hypoglycemia induced by this class of drugs in the pediatric population, coupling pharmacovigilance data with characterization of the pharmacokinetic profile of individual beta-blockers.
The results of our study support the hypothesis that nonselective beta-blockers are more likely to induce hypoglycemia,⁸ and the risk of signaling was also found to be increased as the half-life of the drug increased. Despite the inherent limitations of this type of analysis, such as the inability to estimate absolute risk because of the lack of total prescription data, this study provides useful insights to support clinicians in choosing the most appropriate drug.⁹
Further large-scale prospective observational studies are needed to confirm these findings.