GME Research Review
GME Research Review is a monthly newsletter edited by Rajnish Mago, MD, Associate Professor of Psychiatry at Thomas Jefferson University, who selects, summarizes, and provides a clinical commentary on the latest published research in psychiatry. Each summary has been derived from the relevant article’s abstract and the clinical commentary has been provided by Dr. Mago.
Issue 37, May 2015
“I read somewhere that benzodiazepines can cause Alzheimer’s disease. I want to get off this medication!”
Billioti de Gage et al. Benzodiazepine use and risk of Alzheimer’s disease: case-control study. BMJ. 2014;349:g5205. PubMed PMID: 25208536
Why is this study important?
- This study received a LOT of press. Several patients and trainees have asked me about this paper, concerned that it has shown that benzodiazepines are “cause” Alzheimer’s disease
- If so, it would be very important to know that since Alzheimer’s disease is a very serious condition. It would mean that we must immediately make efforts to decrease or even stop the use of benzodiazepines in our patients
- If this is not the case, how should we respond if a patient or someone else asks us about this study?
- The study aimed to evaluate the relation between the risk of Alzheimer’s disease and exposure to benzodiazepines that were started at least five years before the diagnosis of Alzheimer’s disease
- This was a Case-Control study. That is, it compared patients with a diagnosis of Alzheimer’s disease (Cases) to subjects without that diagnosis (Controls)
- It was done not on live patients but on patient data in a health insurance database in Quebec, Canada
- Both Cases and Controls were selected randomly from persons > 66 years old and living in the community
- Cases (n=1796) were subjects who a) had a first diagnosis of Alzheimer’s disease, and b) had been followed up within the insurance database for at least six years (to give enough time for a possible causal effect to have occurred)
- Controls (n=7184) were persons without a diagnosis of Alzheimer’s disease and matched to the Cases on gender, age group, and duration of follow-up
- Benzodiazepine use that was started at least five years prior to the diagnosis of Alzheimer’s disease was identified. This was because more recently started benzodiazepines could be due to the effects of having Alzheimer’s disease rather than a potential cause of the Alzheimer’s disease. As the authors note, conditions in which benzodiazepines are commonly prescribed (anxiety, insomnia, and depressive disorders) increase significantly in the years before a diagnosis of dementia
- Benzodiazepine use was categorized into three categories based on total cumulative days of benzodiazepine use: 1 to 90 days, 91 to180 days, >180 days)
- The authors state that, “Benzodiazepine ever use was associated with an increased risk of Alzheimer’s disease.”
- However, given that this was a Case Control study, in my opinion this statement is misleading. Remember, they started with subjects with or without Alzheimer’s disease, then looked back (retrospectively) into the database to see if they had ever taken a benzodiazepine that was started at least 5 years prior to the diagnosis of the Alzheimer’s disease
- So, the more scientifically accurate statement would have been that, “the diagnosis of Alzheimer’s disease was associated with an increased odds of having taken a benzodiazepine”
- In any case, this study doesn’t show that the benzodiazepine is the cause of the Alzheimer’s disease. It only means that the two were related or “associated”
- How strong was the association? In a Case Control study, this is expressed as an Odds Ratio, which in this case was 1.5
- That is, subjects with a diagnosis of Alzheimer’s disease were 1.5 times more likely to have been on a benzodiazepine
- But could this association have been due to presence of anxiety, depression, and insomnia for which the benzodiazepine was prescribed?
- When the odds ratio was “adjusted” for these symptoms, it changed only a little — to 1.4. (Key point though: because this is a database study, they could only know whether the patient had anxiety, depression, or insomnia if these problems were specifically coded as a diagnosis)
- Next point: If benzodiazepine use is “associated” with diagnosis of Alzheimer’s disease, then the association would be expected to vary depending on how much benzodiazepine was taken, right?
- For a cumulative dose of < 91 days of dosing, no association between benzodiazepine use and a diagnosis of Alzheimer’s disease was found. For 91 to 180 days of dosing, the odds ratio was 1.3; for > 180 days of dosing, it was 1.8
- The association between Alzheimer’s disease and benzodiazepines was found both for those patients who were currently using benzodiazepines and those who had discontinued them at least one year earlier. That is, the association was also found in those who were not currently on a benzodiazepine
- The authors concluded that benzodiazepine use is associated with an increased risk of Alzheimer’s disease (as pointed out above, I don’t agree that this is a valid conclusion)
- They reasoned that the stronger association with longer exposure to benzodiazepines supports the possibility that the benzodiazepine use is not only “associated” with but is a direct contributing cause of the Alzheimer’s disease
- They did acknowledge the possibility that the benzodiazepine use may also be only indirectly associated with Alzheimer’s disease. That is, benzodiazepine use may be associated with an unknown factor that contributes to causing Alzheimer’s disease
- The most important thing for us to understand here is that in a retrospective (Case Control) study, what is established is association, not causation. While different analyses were used to support a true causal relationship, they cannot really establish that the benzodiazepines were the cause
- One factor to think about in trying to evaluate a potential causal relationship: By what possible mechanism could benzodiazepines directly contribute to the development of Alzheimer’s disease rather than indirectly due to the effects of depression, anxiety, etc? There is little biological plausibility to the hypothesis. As the authors suggest, in the future, animal studies could evaluate potential biological mechanisms
- Meanwhile, we must assume that it is simply an association. It may be in part due to a) other risk factors present in patients who are prescribed benzodiazepines, and b) for those patients who were on the benzodiazepine at the time of the diagnosis, increased diagnosis of Alzheimer’s disease due to cognitive adverse effects of the medication
- Of course, nothing I have said here should be taken to suggest that that I disagree with the recommendation that benzodiazepine use be minimized in older patients due to their potential adverse effects on cognition while the patient is taking a benzodiazepine
In bipolar depression, can we consider adding armodafinil to a mood stabilizer?
Calabrese et al.; Armodafinil Treatment Trial Study Network. Efficacy and safety of adjunctive armodafinil in adults with major depressive episodes associated with bipolar I disorder: a randomized, double-blind, placebo-controlled, multicenter trial. J Clin Psychiatry. 2014;75(10):1054-61. PubMed PMID: 25099397.
Why is this study important?
- Bipolar depression is notoriously hard to treat, right? Do you know of any medication that consistently works really well for it? I don’t.
- So, any new option for treatment of bipolar depression is of interest to us
- Armodafinil is the R-isomer of modafinil which consists of the R- and L- isomers
- Its only apparent advantage over modafinil is that its effect seems to last for longer through the day
- This study aimed to evaluate whether adding armodafinil to a maintenance medication for bipolar disorder was efficacious for bipolar depression and what the adverse effects may be.
- This was a multicenter, randomized, placebo-controlled clinical trial
- Adults with bipolar I depression were enrolled
- Subjects had to be taking a maintenance medication (lithium, divalproex, second-generation antipsychotic) for at least four weeks
- Armodafinil 150 mg/day was added to the maintenance medication
- The primary efficacy measure was change in the severity of depression as measured by the Inventory of Depressive Symptomatology-Clinician-Rated (IDS-C30) total score
- 400 patients were randomized to addition of either armodafinil 150 mg/day or placebo
- After eight weeks of treatment, there was a statistically significantly greater decrease in IDS-C30 total score in patients on armodafinil vs on placebo
- But was this a clinically useful difference?
- For this, always look right away at the percentage of patients who “responded” in each group
- The proportion of IDS-C30 responders (50% or greater decrease in severity of depression) was statistically significantly higher for the 150 mg/day armodafinil group versus placebo at final visit (46% vs 34%)
- A rule of thumb is that the difference in percentage of subjects achieving an outcome should be at least 10% for the difference to be considered clinically meaningful (a 20% difference would a terrific benefit). So, the difference in percentage of responders here is clinically useful
- Now, we are interested not only the percentage of responders but also in the percentage of remitters (i.e., those who had minimal or no symptoms after treatment)
- The proportion of IDS-C30 remitters was similar in both groups: 21% for armodafinil 150 mg/day versus 17% for placebo
- This does not bother me too much. It is usual for there to not be a significant difference in remission between active treatment and placebo in short term studies. Remission is a higher bar than response
- Next, what about adverse events?
- Adverse events observed in > 5% of either the armodafinil 150 mg/day or placebo groups and more frequently with 150 mg/day armodafinil were diarrhea (9% vs 7%), and nausea (6% vs 5%), respectively. These differences are so similar that I don’t think we can say that armodafinil differs from placebo in this regard
- There was one case of hepatic failure leading to death and one case of acute hepatitis
- Addition of armodafinil 150 mg/day to a mood stabilizer was efficacious for bipolar I depression and was well tolerated
- If you look deeper into the paper, you find additional information that may be relevant
- Anxiety occurred in 4% of patients on 150 mg/day of armodafinil and 0% of those on placebo
- Psychosis occurred in two patients on armodafinil and none on placebo. While this can be a coincidence, I mention it here because I had anecdotally heard about psychosis emerging while on armodafinil. Let’s keep this possibility at the backs of our minds
- Regarding the one case each of hepatic failure and of acute hepatitis, whether these events were causally related to armodafinil is not clear but cannot be ruled out. It should be noted that modafinil and armodafinil have been associated with elevations of GGT and alkaline phosphatase
- FYI, two other phase III studies done around the same time did not show statistically significant benefit from addition of armodafinil to maintenance medication in patients with bipolar depression. Therefore, the manufacturer is not going to continue to try to obtain an FDA approval for armodafinil as a treatment for bipolar depression
- Overall, in my opinion, armodafinil or modafinil cannot be considered proven treatments for bipolar depression, but could be kept as options. In particular, they may be considered earlier in treatment algorithms if fatigue and daytime sleepiness are present. Also, it is not uncommon to see patients with comorbid bipolar depression and sleep apnea
Is using escitalopram OK for depression occurring right after a patient has a myocardial infarction or unstable angina?
Kim et al. Escitalopram treatment for depressive disorder following acute coronary syndrome: a 24-week double-blind, placebo-controlled trial. J Clin Psychiatry. 2015;76:62-8. PubMed PMID: 25375836.
Why is this study important?
- Major depressive disorder and coronary artery disease often occur together. Eight per cent of men and five per cent of women with major depressive disorder also have coronary artery disease. That’s a lot of people!
- Recently, cardiologists have even started to say that depression should be formally called a risk factor for adverse medical outcomes in patients with acute coronary syndrome
- By the way, what does “acute coronary syndrome” mean exactly? It includes both myocardial infarction and unstable angina (angina at rest or during light activity), both caused by a sudden decrease in blood flow to the heart
- So, there are a lot of patients with major depressive disorder and coronary artery disease, some of whom would be appropriate to treat with antidepressants
- But, we also have to check to see if the antidepressant can have any adverse effects on the heart, especially in such acutely ill patients
- This study evaluated both the efficacy and safety of escitalopram in treating depressive disorders 2 to 14 weeks after an episode of acute coronary syndrome
- A total of 217 patients with a depressive disorder (121 patients with major depressive disorder and 96 patients with minor depressive disorder) and acute coronary syndrome were enrolled
- The patients were randomized to receive either escitalopram (flexible dose 5 to 20 mg/day) or placebo for 24 weeks
- The primary outcome measure was the Hamilton Depression Rating Scale but other rating scales were used as well
- For assessing cardiovascular safety, echocardiography, ECG, laboratory tests, body weight, and blood pressure were done
- There was a greater reduction in severity of depression in patients treated with escitalopram than in patients given placebo
- The mean difference in Hamilton Depression Rating Scale scores between patients on escitalopram or placebo was 2.3. (A mean difference of two points or greater is usually considered meaningful.)
- The “effect size” was 0.38 indicating a modest effect overall. This is usual in antidepressant studies; the difference between drug and placebo is not huge
- No adverse effects on cardiovascular safety measures
- The only adverse event reported statistically significantly more often than on placebo was dizziness
- Escitalopram was found to be efficacious for treating depressive disorders in patients with acute coronary syndromes
- No evidence was found of reduced cardiovascular safety due to treatment with escitalopram
- Some previous studies like the well-known SADHARDT study (Glassman et al., 2002), did not find antidepressant superior to placebo for depressive disorder occurring immediately after an episode of acute coronary syndrome. However, in the SADHARDT study, sertraline was superior to placebo in the subgroup of patients with more severe depression and those with prior episodes of depression. Personally, my interpretation of this has been that not everyone who is sad/depressed after a bad thing happens has a pharmacologically treatable depression
- In this study, it would seem that the medication worked overall. However, inside the paper we find that only in major depressive disorder was drug statistically significantly better than placebo. In minor depressive disorder, drug was not statistically significantly better than placebo. Thus, higher severity of depression is a predictor of potential benefit from medication vs placebo
- Studies like this, evaluating the treatment of mental disorders in patients with significant medical comorbidity, are important. Medical comorbidity is very common in our patients yet in the initial clinical trials done when the medication are introduced, patients with significant medical comorbidity are excluded
- When a medication is found to be generally safe in patients who have just had a myocardial infarction or unstable angina a few days ago, this is reassuring to us when we treat patients whose coronary artery disease is not that acute
- However, 28% of patients dropped out of the study after randomization but before the first visit. Therefore, their data could not be included. There was some data to indicate that these patients were sicker than those who continued. My conclusion from this is that the results of this trial may not apply to the sickest patients
- I have another important problem with how the data about cardiovascular safety was reported in this study. As we have argued elsewhere (Mago et al. Medically serious adverse effects of newer antidepressants. Current Psychiatry Reports. 2008;10(3):249-257), even when the average values of parameters like blood pressure, QTc interval, etc are not elevated, the possibility remains that a few patients may have had clinically significantly abnormal values. I think it is important to report the percentage of patients who developed an undesirable change, e.g., prolonged QTc or a predefined amount of increase in the QTc. (Of course, it is hard to assess the statistical significance of such changes in a relatively small sample size.)
- While the authors say that dizziness was the only adverse event that occurred statistically significantly more often than on placebo, the problem with that is that most adverse events are not specifically looked for. That is, patients are not asked if they have dizziness, dry mouth, fatigue, etc. Another way of looking at adverse events that is frequently used is to identify adverse events that occurred at least twice as often on drug than on placebo. By this approach, other adverse events on escitalopram were sexual dysfunction (7%), insomnia (5%), and fatigue (3%).
When even clozapine doesn’t work for schizophrenia, is there yet another treatment strategy that can still be tried?
Petrides et al. Electroconvulsive therapy augmentation in clozapine-resistant schizophrenia: a prospective, randomized study. Am J Psychiatry. 2015;172(1):52-8. PubMed PMID: 25157964.
Why is this study important?
- Anyone with significant experience in treating schizophrenia must have been humbled by the failure of our antipsychotic medications in about a third of patients
- Certainly, clozapine is efficacious in a proportion of such difficult-to-treat schizophrenia. However, in 45 to 70% of these patients the illness does not respond significantly to clozapine
- One strategy has been to add another antipsychotic to the clozapine. This has been found to have only limited benefit
- In such situations, desperate patients and clinicians need to have other options that may work
- While some prior data exists, this is the first randomized controlled clinical trial to evaluate the addition of electroconvulsive therapy (ECT) to clozapine in patients with “treatment-refractory” schizophrenia
- This was a randomized, single-blind, eight-week clinical trial
- Patients with schizophrenia were enrolled who had failed to improve significantly with other antipsychotic medication and also with clozapine
- They had to have failed at least two trials of an adequate dose of antipsychotic for at least four weeks. In addition, they had to have persistent psychotic symptoms after a trial of clozapine for at least 12 weeks, at a blood level of at least 350 ng/mL
- Patients with clozapine-resistant schizophrenia were assigned to either treatment as usual (clozapine group) or a course of clozapine plus bilateral ECT (ECT plus clozapine group)
- After the eight weeks of the main part of the study, patients in the clozapine only group who did not respond to clozapine were given an eight-week open label trial of ECT (open label means that neither the patient nor the treatment team were blinded about the treatment)
- The ECT was given three times per week for the first four weeks and then twice per week for the last four weeks
- In advance, “Response” was defined as a 40% or greater reduction in symptoms based on the psychotic symptom subscale of the Brief Psychiatric Rating Scale, along with a clinician rating of no more than “borderline” illness. This is a fairly high bar to meet
- Masked raters (those who did not know which treatment group the patient was in) were used to assess the response to treatment
- Nineteen patients received clozapine only and 20 patients received clozapine plus ECT
- Fifty percent of the patients who received clozapine plus ECT were Responders (as defined above) while none of the patients in the clozapine only group responded
- Then, all 19 patients in the clozapine-only group were given open label ECT in addition to continuing the clozapine. This time, 47% of these patients were Responders
- Global cognition was no different in patients who received either treatment. The authors noted that these patients were severely ill and had low cognitive performance at baseline. Some aspects of cognitive functioning may have improved due to improvement in symptoms of schizophrenia, which may have counterbalanced the transient memory impairment often seen with ECT
- There are some theoretical concerns about using ECT in patients on clozapine, especially the possibility of prolonged or spontaneous seizures. However, this was not found in this study
- Addition of ECT to clozapine in patients with clozapine-resistant schizophrenia was efficacious in about 50% of patients and appeared to be safe
- Of course, further research will be needed to find out if the benefit persists
- This study adds the small amount of previous data supporting the addition of ECT to clozapine when other antipsychotics as well as clozapine fail to produce adequate improvement
- The percentage of patients who responded and the degree of improvement noted is impressive given how refractory the illness in these patients was
- If further research does not find the benefit to persist, maintenance ECT may be needed
- This study cannot answer the question whether ECT alone (without clozapine) may have been just as efficacious as the combination
What can we do about sleep problems in children with ADHD and will this help?
Hiscock et al. Impact of a behavioural sleep intervention on symptoms and sleep in children with attention deficit hyperactivity disorder, and parental mental health: randomised controlled trial. BMJ. 2015;350:h68. PubMed PMID: 25646809
Why is this study important?
- While medications are usually quite efficacious and invaluable in the treatment of attention deficit hyperactivity disorder (ADHD), various problems tend to remain and addition of other, non-medication, approaches can be helpful
- Sleep problems are very common in children with ADHD and also affect other symptoms of the disorder
- Now, it may be very easy to say, “Follow sleep hygiene!” but can the children/families really implement behavioral interventions for sleep to a significant extent? Will this really work for sleep problems in children with ADHD? Could improving sleep in these patients help with the core ADHD problems? If the answer is “Yes!” to all these questions, this would be an extremely helpful strategy that we would use all the time
- This study aimed to evaluate whether behavioral strategies for improving sleep problems in children with ADHD could also improve the symptoms, behavior, daily functioning, and working memory of children with ADHD, and — don’t miss this — the mental health of their parents. (The authors note that parents of children who have ADHD and sleep problems experience poorer mental health and are more likely to be late for work.)
- This was a randomized, controlled trial done in 21 general pediatric practices in Australia
- 244 children with ADHD aged 5 to 12 years were enrolled. Most of the children had other comorbid problems including autism spectrum disorders
- Sleep hygiene practices and standardized behavioral strategies were delivered by psychologists or trainee pediatricians who were trained in this intervention during two sessions of three hours each. That’s it.
- Families were provided the intervention in two consultations two weeks apart and one follow-up telephone call two weeks after that. Again, that’s it.
- Children in the control group received usual clinical care and nothing was added
- Outcomes were measured at three and six months as follows:
- Severity of ADHD symptoms (parent and teacher versions of the ADHD rating scale IV)
- Sleep problems (parent reported severity, children's sleep habits questionnaire)
- Behavior (strengths and difficulties questionnaire)
- Quality of life (pediatric quality of life inventory 4.0)
- Daily functioning (daily parent rating of evening and morning behavior)
- Working memory (working memory test battery for children, done at six months only)
- Parent mental health (depression anxiety stress scales)
- In children from families that received the intervention, there was a statistically significantly greater decrease in ADHD symptoms at both three and six months with small and small-to-moderate effect sizes, respectively
- Children in families that received the intervention had fewer moderate-severe sleep problems at three months (56% v 30%) and at six months (46% v 34%)
- At three and six months, about half and about one thirds, respectively, of the beneficial effects of the intervention on ADHD symptoms was determined to be mediated through improved sleep
- Families that received the intervention reported greater improvements in all other child and family outcomes except parental mental health. However, the parents did report increased work attendance. At six months, the children also had fewer days late for school
- At both three months and six months, teachers’ reports indicated improved behavior of the children who received the intervention
- Working memory (backwards digit recall) was higher in the intervention children compared with control children at six months
- A brief behavioral sleep intervention modestly improved the severity of ADHD symptoms in a community sample of children with ADHD, most of whom were taking stimulant medications
- The intervention also improved the children's sleep, behavior, quality of life, and functioning, with most benefits sustained at six months after the intervention
- The study has an important methodological limitation that parents knew that they were receiving the intervention and therefore may be more likely to report benefit. However, this is not always the case. I can imagine some parents being somewhat demoralized after dealing with the child’s problems for many years; such parents could be less likely to notice improvement
- Also, the control group did not get a control intervention corresponding to the study intervention. It is common in non-pharmacological studies to provide the control group with a control intervention that provides the same amount of attention, etc, but which lacks the therapeutic ingredient that is being tested. For example, the control group could have been given the same number of sessions but covering general education about ADHD rather than tackling sleep problems. Of course, patients in the control group did continue to get usual care including medication
- Nevertheless, the intervention is simple and easy to implement and, in my opinion, there should be no reason for not providing a sleep-related behavioral intervention to all children with ADHD who have sleep problems (the majority)
- I also think that for children who show limited improvement or have complicated problems, it is possible that a greater number of sessions could produce additional improvement
If you found this discussion of clinically useful research helpful, please make sure you share it with others on Facebook, LinkedIn, Twitter, etc by clicking one of the icons below. Thank you!