The risk of depression is elevated in patients with cardiovascular diseases, but several specific antihypertensive therapies are associated with a reduced risk, and none appear to increase the risk, according to a population-based study that evaluated 10 years of data in nearly 4 million subjects.
“As the first study on individual antihypertensives and risk of depression, we found a decreased risk of depression with nine drugs,” reported a collaborative group of investigators from multiple institutions in Denmark where the study was undertaken.
In a study period spanning from 2005 to 2015, risk of a diagnosis of depression was evaluated in patients taking any of 41 antihypertensive therapies in four major categories. These were identified as angiotensin agents (ACE inhibitors or angiotensin II receptor blockers), calcium antagonists, beta-blockers, and diuretics.
Within these groups, agents associated with a reduced risk of depression were: two angiotensin agents, enalapril and ramipril; three calcium antagonists, amlodipine, verapamil, and verapamil combinations; and four beta-blockers, propranolol, atenolol, bisoprolol, and carvedilol. The remaining drugs in these classes and diuretics were not associated with a reduced risk of depression. However, no antihypertensive agent was linked to an increased risk of depression.
All people living in Denmark are assigned a unique personal identification number that permits health information to be tracked across multiple registers. In this study, information was linked for several registries, including the Danish Medical Register on Vital Statistics, the Medicinal Product Statistics, and the Danish Psychiatric Central Register.
Data from a total of 3.75 million patients exposed to antihypertensive therapy during the study period were evaluated. Roughly 1 million of them were exposed to angiotensin drugs and slightly more than a million were exposed to diuretics. For calcium antagonists or beta-blockers, the numbers were approximately 835,000 and 775,000, respectively.
After adjustment for such factors as concomitant somatic diagnoses, sex, age, and employment status, the hazard ratios for depression among drugs associated with protection identified a risk reduction of 10%-25% in most cases when those who had been given 6-10 prescriptions or more than 10 prescriptions were compared with those who received 2 or fewer.
At the level of 10 or more prescriptions, for example, the risk reductions were 17% for ramipril (HR, 0.83; 95% CI, 0.78-0.89), 8% for enalapril (HR, 0.92; 95% CI, 0.88-0.96), 18% for amlodipine (HR, 0.82; 95% CI, 0.79-0.86), 15% for verapamil (HR, 0.85; 95% CI, 0.79-0.83), 28% for propranolol (HR, 0.72; 95% CI, 0.67-0.77), 20% for atenolol (HR, 0.80; 95% CI, 0.74-0.86), 25% for bisoprolol (HR, 0.75; 95% CI, 0.67-0.84), and 16% for carvedilol (HR, 0.84; 95% CI, 0.75-0.95).
For verapamil combinations, the risk reduction was 67% (HR, 0.33; 95% CI, 0.17-0.63), but the investigators cautioned that only 130 individuals were exposed to verapamil combinations, limiting the reliability of this analysis.
Interpreting the findings
A study hypothesis, the observed protective effect against depression, was expected for angiotensin drugs and calcium-channel blockers, but not for beta-blockers, according to the investigators.
“The renin-angiotensin systems is one of the pathways known to modulate inflammation in the central nervous system and seems involved in the regulation of the stress response. Angiotensin agents may also exert anti-inflammatory effects,” the investigators explained. “Dysregulation of intracellular calcium is evident in depression, including receptor-regulated calcium signaling.”
In contrast, beta-blockers have been associated with increased risk of depression in some but not all studies, according to the investigators. They maintained that some clinicians avoid these agents in patients with a history of mood disorders.
In attempting to account for the variability within drug classes regarding protection and lack of protection against depression, the investigators speculated that differences in pharmacologic properties, such as relative lipophilicity or anti-inflammatory effect, might be important.
Despite the large amount of data, William B. White, MD, professor emeritus at the Calhoun Cardiology Center, University of Connecticut, Farmington, is not convinced.
“In observational studies, even those with very large samples sizes, bias and confounding are hard to extricate with controls and propensity-score matching,” Dr. White said. From his perspective, the protective effects of some but not all drugs within a class “give one the impression that the findings are likely random.”
A member of the editorial board of the journal in which this study appeared, Dr. White said he was not involved in the review of the manuscript. Ultimately, he believed that the results are difficult to interpret.
“For example, there is no plausible rationale for why 2 of the 16 ACE inhibitors or angiotensin II receptor blockers or 4 of the 15 beta-blockers or 3 of the 10 calcium-channel blockers would reduce depression while the others in the class would have no effect,” he said.
Despite the investigators’ conclusion that these data should drive drug choice for patients at risk of depression, “I would say the results of this analysis would not lead me to alter clinical practice,” Dr. White added.
According to the principal investigator of the study, Lars Vedel Kessing, MD, DSc, professor of psychiatry at the University of Copenhagen, many variables affect choice of antihypertensive drug. However, the depression risk is elevated in patients with cardiovascular or cerebrovascular disease and hypertension.
When risk of a mood disorder is a concern, use of one of the nine drugs associated with protection from depression should be considered, “especially in patients at increased risk of developing depression, including patients with prior depression or anxiety and patients with a family history of depression,” he and his coinvestigators concluded.
However, Dr. Kessing said in an interview that the data do not help with individual treatment choices. “We do not compare different antihypertensives against each other due to the risk of confounding by indications, so, no, it is not reasonable to consider relative risk among specific agents.”
The authors reported no potential conflicts of interest involving this topic.
SOURCE: Kessing LV et al. Hypertension. 2020 Aug 24. doi: 10.1161/HYPERTENSIONAHA.120.15605.