A decade ago the Federal Drug Administration (FDA) released several warnings that children, adolescents (ages 10–17), and young adults (ages 18-29) taking antidepressants were at increased risk for suicidal ideation and behavior. A recent study found that following these warnings, antidepressant use among adolescents, young adults, and adults dropped, and psychotropic drug poisonings (a validated measure of suicide attempts) increased among adolescents and young adults. Numbers of completed suicides did not change for any age group.
The decision to place the warnings on antidepressant packaging was somewhat controversial because it was based on studies that were not necessarily designed to measure suicide risk. The relationship between depression, medication, and suicide is complicated. Medication can improve mood, but patients may seek out medication because of pre-existing suicidal thoughts, and the medication may not reduce these in young people.
The reduction in antidepressant use that occurred after the warnings was accompanied by a drop in depression diagnoses in children and adults. Studies have suggested that the decreases in antidepressant were not accompanied by increases in other treatments, such as psychotherapy or atypical antipsychotics, among young people. Increased monitoring of patients was called for in the FDA’s box warning, but did not take place.
The study of the aftermath of the FDA warnings, published by Christine Y. Lu et al. in a 2014 article in the journal BMJ, used data from 11 insurance networks throughout the US. The researchers used an interrupted time series study design, which is used to show whether a policy causes an abrupt change in the expected slope of study outcomes. Data covered the pre-warning period (first quarter of 2000 to third quarter of 2003), the warning “phase-in” period (last quarter of 2003 to last quarter of 2004) and the post-warning period (first quarter of 2005 to last quarter of 2010). The study cohorts included around 1.1 million adolescents, 1.4 million young adults, and 5 millions adults per quarter.
Among adolescents, the previously upward trend in antidepressant use declined by 31.0% in the second year after the warnings, and psychotropic drug poisonings increased by 21.7% (a figure that was statistically significant for males). Poisonings by any drug increased by 13.9% in the second year after the warnings. After 2008, the downward trend in antidepressant use reversed, indicating that either the initial effects of the warning had worn off or that modifications to the warnings in May 2007, which encouraged patients and doctors to consider the risk of antidepressants alongside the risk of leaving mood disorders untreated, led to increased use.
Among young adults, the upward trend in antidepressant use declined by 24.3% in the second year after the warnings, and psychotropic drug poisonings increased by 33.7%, a statistically significant change for both male and female patients.
Among adults, to whom the warnings were not directed, antidepressant use decreased by 14.5% in the second year after the warnings.
The study by Yu et al. is the first to show that suicide attempts actually increased after the FDA warnings. The authors suggest that the increase in suicide attempts might be a consequence of undertreating mood disorders, since antidepressant use dropped simultaneously. The warnings and related media attention may have led to these unintended consequences, since media reports can sometimes be oversimplified.
The antidepressant agomelatine (which is available in many countries, but not the US) and the anti-insomnia drug ramelteon (Rozerem) both act as agonists at melatonin M1 and M2 receptors. New research is clarifying the role of these receptors in sleep.
In new research from Stefano Comai et al., mice who were genetically altered to have no M1 receptor (MT1KO knockout mice) showed a decrease in rapid eye movement (REM) sleep, which is linked to dreaming, and an increase in slow wave sleep. Mice who were missing the M2 receptor (MT2KO knockout mice) showed a decrease in slow wave sleep. The effects of knocking out a particular gene like M1 or M2 end up being opposite to the effect of stimulating the corresponding receptor.
The researchers concluded that MT1 receptors are responsible for REM sleep (increasing it while decreasing slow wave sleep), and MT2 receptors are responsible for slow wave non-REM sleep.
The new information about these melatonin receptors may explain why oral melatonin supplements can make a patient fall asleep faster, but do not affect the duration of non-REM sleep. The authors suggest that targeting MT2 receptors could lead to longer sleep by increasing slow wave sleep, potentially helping patients with insomnia.
Type 2 diabetes can damage the brain, particularly by reducing volume of the hippocampus, and frequently occurs in patients with bipolar disorder. A recent study of patients with bipolar disorder and abnormal glucose metabolism showed that patients with bipolar disorder who also had insulin resistance, glucose intolerance, or type 2 diabetes had smaller hippocampi than both patients with bipolar disorder and normal glucose function and normal control participants without a psychiatric disorder. In those with bipolar disorder and glucose abnormalities, age was associated with lower hippocampal volume to a greater extent than in bipolar patients with normal glucose function.
In the study, published by Tomas Hajek et al. in the journal Neuropsychopharmacology, not only did diabetes or prediabetes reduce the size of the hippocampus, but also reduced gray matter in the cerebral cortex, including the insula.
The researchers hope that treating diabetes, or possibly even its initial symptoms, more effectively may prevent these gray matter losses and slow brain aging in patients with bipolar disorder.
In a 2013 article in the journal European Psychiatry, in which researcher Valery V. Gafarov examined depression’s influence on cardiovascular health in Russia, an astonishing 55.2% of women aged 25–64 years in the study were diagnosed with depression. The study, in which 870 women in the city of Novosibirsk were surveyed over 16 years from 1995 to 2010, was part of a World Health Organization program called “MONICA-psychosocial.”
The researchers collected information on the incidence of myocardial infarction (heart attack), arterial hypertension, and stroke among the women. Over the 16 years of the study, 2.2% of the women had heart attacks and 5.1% had strokes. Women with depression were 2.53 times more likely to have a heart attack and 4.63 times more likely to have a stroke than women without depression.
Among women with average education levels, married women with depression were more likely to have heart attacks, hypertension, and strokes. Hypertension was more likely among women who worked as managers or light manual laborers.
At the 2014 meeting of the International College of Neuropsychopharmacology, researcher N. Miyake described the effects of the nutritional supplement n-acetylcysteine (NAC) on clinical symptoms in subjects with subthreshold symptoms of psychosis.
N-acetylcysteine, a glutathione precursor, has neuroprotective effects. In this case series, four patients with subthreshold psychosis were given 2000mg/day of NAC for 12 weeks. The patients’ symptoms improved to the point that three of the four were no longer considered at risk for psychosis.
Editor’s Note: These promising anecdotal observations deserve careful follow up using a control group. Omega-3 fatty acids have been show to slow conversion to full psychosis and performed better than placebo in a controlled study. Both n-acetylcysteine and omega-3 fatty acids should definitely be studied for those with emerging symptoms of bipolar disorder.
C-reactive protein, or CRP, is a protein found in blood plasma, the levels of which rise in response to inflammation. In a recent study, levels of CRP were able to predict which of two antidepressants a patient was more likely to respond to.
The 2014 article by Rudolph Uher et al. in the American Journal of Psychiatry reported that low levels of CRP (<1 mg/L) predicted a good response to the selective serotonin reuptake inhibitor (SSRI) escitalopram (Lexapro) while higher levels of CRP predicted a good response to the tricyclic antidepressant nortriptyline, a blocker of norepinephrine reuptake.
The research was part of the Genome-Based Therapeutic Drugs for Depression (GENDEP) study, a multicenter open-label randomized clinical trial. CRP was measured in the blood of 241 adult men and women with major depressive disorder. In the article the researchers say that CRP and its interaction with medication explained more than 10% of the individual variance in response to the two antidepressants.
If these findings can be replicated with these and similarly acting drugs, it would be a very large step in the direction of personalized medicine and the ability to predict individual response to medications.
Joanna Soczynska in Roger McIntyre’s lab at the University of Toronto presented a poster at the 2014 meeting of the International College of Neuropsychopharmacology (CINP) on the anti-inflammatory and neuroprotective antibiotic minocycline.
Twenty-seven patients with a major depression received minocycline in addition to the medications they were already being prescribed. Dosage was 100mg twice a day. Treatment with adjunctive minocycline was associated with significant improvement on several scales that measure depression severity.
Editor’s Note: What was particularly interesting was that a subset of patients achieved complete remission, raising the question whether these patients might have markers of inflammation that would predict this excellent response. The authors concluded that the “results provide a rationale for testing minocycline’s efficacy in a larger randomized, placebo-controlled trial.”
Exactly this type of study was proposed a year ago by researcher Andy Nierenberg and given the best marks by a National Institute of Mental Health review committee but was turned down for funding because the National Institute of Mental Health has implemented a new initiative, Research Domain Criteria (RDoC), that lays out new criteria for research, limiting funding to those studies that focus on a molecular target that spans several diagnoses.)
Autism spectrum disorders are associated with developmental abnormalities at excitatory synapses. Dendrites, the branched projections of neurons where electrical signals are passed from one cell to the next, are covered in hundreds to thousands of spines that facilitate the synaptic connections with other neurons. These spines are created and also pruned as part of normal learning and development.
Post-mortem examination of the brains of patients with autism spectrum disorders shows increased density of dendritic spines and less pruning in certain neurons in the temporal lobe. These examinations also show impaired mTOR autophagy. MTOR is a protein that plays a role in cell growth and survival. Autophagy is the normal process by which some components of cells are broken down.
A 2014 study by Guomei Tang et al. in the journal Neuron showed that mice that are genetically altered to have overactive mTOR also have reduced dendritic spine pruning, blockade of autophagy, and increased autism-like behaviors. An immunosuppressant drug called rapamycin inhibits mTOR, and treating the mice with this drug corrected the problems with spine pruning and the autism-like behaviors. (This was not true for mice who had been altered to have another type of autophagy.) Normal spine formation was not affected by the restored pruning ability.
Tang et al. concluded that mTOR autophagy plays an important role in dendritic spine pruning, and that restoring neuronal autophagy can correct synaptic abnormalities and restore normative social behavior in mice with hyperactive mTOR.
At the 2014 meeting of the International College of Neuropsychopharmacology, researcher Scott Russo described characteristics of rodents who showed depression-like behavior after 10 days of exposure to a larger, more aggressive animal (a phenomenon known as defeat stress). These animals exhibited many behaviors that resembled human depression, including anxiety-like behaviors while navigating a maze; activation of the hypothalamic-pituitary-adrenal axis; circadian rhythm abnormalities; metabolic changes such as glucose intolerance; susceptibility to addiction; anhedonia, a lack of interest in sucrose, sex or intracranial self-stimulation; and profound and permanent social avoidance.
In susceptible animals, Russo found anatomical changes in the GABAergic neurons of the nucleus accumbens (also known as the ventral striatum), including increased numbers of synapses and a greater number of stubby spines on dendrites (the branched projections of neurons where electrical signals are passed from one cell to the next), as well as greater excitability of glutamatergic input, observed as excitatory post-synaptic potentials.
Russo’s attempt to identify these key neurons among the billions of neurons and the 100 to 500 trillion synapses in the brain was like the search for a needle in a haystack, but thinks he found it. The medium spiny neurons of the nucleus accumbens contain GABA and receive synapses from the prefrontal cortex, amygdala, and intralaminar nucleus of the thalamus (ILT), in addition to dopamine inputs from the VTA, and cholinergic, somatostatin, and orexin inputs. Russo found that it was the ILT inputs that conveyed susceptibility to defeat stress, and their presynaptic endings showed increased levels of glutamate transporters (VGLUT-2). Driving the ILT was sufficient to cause the rodents to display the depression-like behaviors, and silencing the ILT during defeat stress prevented the susceptible behaviors (like social avoidance) and promoted resilience.
Low cortisol after a trauma is a risk factor for developing chronic post-traumatic stress disorder (PTSD). Researcher Joseph Zohar studied has been researching the effects of steroids on the development of PTSD and presented some findings at the 2014 meeting of the International College of Neuropsychopharmacology.
Twenty-five patients who experienced a traumatic event and showed acute stress symptoms were given either a single high-dose injection of hydrocortisone (100–140 mg) or a placebo within six hours of the trauma. Follow-up evaluation took place after two weeks, one month, and three months. Those who received this single high dose of hydrocortisone had lowered stress symptoms and less subsequent PTSD compared to those who received placebo.