Superior Index Go to the next: Chapter 10
Print Files: A4 Size.
In Chapter 4, we discussed adverse reactions that can occur when psychiatric drugs are taken. The present chapter reviews adverse reactions that can develop during withdrawal.
This book is directed to people who have experienced problems while taking or withdrawing from psychiatric drugs that have been prescribed by doctors according to generally accepted guidelines for their clinical use.
On the other hand, you may be experiencing serious difficulties from using stimulants, sedatives, or painkillers in large amounts, or in combination with each other, or with alcohol. You may have started this use with a routine medical prescription but are now taking much larger quantities and feel you cannot control the cumulative effects. Similarly, if you have been obtaining psychiatric drugs illegally or from multiple doctors, you may have a serious drug dependence problem. If you are snorting, injecting, or inhaling these drugs, you are exposing yourself to extreme dangers.
Under any of these circumstances, you may need immediate professional help that may include a detoxification or drug-rehabilitation unit.
By contrast, this book focuses on problems associated with the ordinary use and withdrawal from psychiatric drugs that have been routinely prescribed by doctors.
All psychiatric drugs can produce unpleasant, disturbing reactions upon withdrawal or discontinuation. For some drugs that have been with us for decades, especially the benzodiazepine tranquilizers, withdrawal reactions are well described in the medical literature. Doctors tend to know about them. However, doctors are too often unfamiliar with withdrawal problems associated with many of the other psychiatric drugs they routinely prescribe. Since the first edition of this book appeared, we have observed a definite increase in professional and popular sensitivity about withdrawal reactions of psychiatric drugs, especially antidepressants. However, much more education is needed.
Stimulants such as Ritalin, Adderall, and Dexedrine have been widely used for decades, yet many physicians seem unaware of the marked withdrawal reactions they can cause. The same is true of drugs more recently introduced on the market. All newer antidepressants, such as Lexapro, Cymbalta, Paxil, Prozac, and others, often cause distressing withdrawal, yet many physicians appear to be unaware of this fact. For example, family physicians or primary care practitioners prescribe the bulk of antidepressants. Yet, to our knowledge, the first systematic advice to family physicians concerning the management of antidepressant discontinuation was published in American Family Physician only in late 2006 (Warner et al., 2006 [382]).
What we find most disturbing is that, even when doctors do know about the dangers of withdrawal problems from drugs, they often fail to warn their patients119. Sometimes they feel too pressed for time to inform their patients about these dangers. Sometimes they simply forget. Sometimes they are concerned that patients will complain about any adverse effect that the doctor mentions or that the patient hears about. At other times, doctors are simply afraid to discourage their patients from taking the medication. However, medical ethics and sound practice require that physicians advise patients about withdrawal problems. There is no legitimate excuse for not doing so.
Once you read this book, you may know much more about withdrawal problems than your physician does. If you are planning to stop or reduce the dose of a psychiatric drug, you may want to share this book with your doctor.
Withdrawal reactions can be difficult to recognize. For example, consider the case of George who, two nights earlier, stopped taking the Klonopin (clonazepam) that was prescribed for him at bedtime to help him sleep. It's the first time in several months that he has tried to fall asleep without his "sleeping pill". Since stopping the Klonopin, George has been having more trouble falling asleep than ever before. He has been lying awake for hours worrying about what will happen to him if he can't ever fall asleep again without drugs.
George is most likely having a withdrawal reaction from the Klonopin. For several months, his brain has been fighting the effects of the drug, which George has taken on a daily basis. This process has resulted in reactive or compensatory overstimulation of his brain. Now that the drug has been stopped, George's overstimulated brain has taken over and is keeping George awake. Since his insomnia is worse than it was before he started the Klonopin, his withdrawal reaction can be called a "rebound" - a worsening of his original symptoms.
Alternatively George could be suffering from psychological fear of giving up the drug. If he wasn't so scared about doing without a sleeping pill, he might not be having so much trouble sleeping. Having gotten into the habit of using Klonopin, he may now be afraid to give it up. This withdrawal reaction would be described as psychologically caused, in contrast to the physically caused withdrawal problems we are mostly concerned with. In such circumstances, George would need reassurance from his doctor that he will get used to sleeping without medication.
George could also be suffering from his original insomnia. Indeed, for years, he has tended to worry at night. In the absence of drug-induced sleep, his worrying may have returned. This situation would be described as a "relapse" - a return of the original problems experienced before drug use. Counseling or psychotherapy could be especially helpful in dealing with this aspect of George's insomnia.
Because Klonopin and other sleep-inducing drugs commonly cause withdrawal problems, we are probably correct in specifying drug withdrawal as the most likely diagnosis in George's case. If he gradually resumes more normal sleep in a few days or weeks, this diagnosis would be confirmed. As noted, it is often difficult to distinguish drug withdrawal reactions from other problems, such as a return of the original symptoms. Usually however, it's safest to assume that physical withdrawal is playing a role and that a slow, tapered withdrawal will be helpful.
Consider the possibility that you are having a withdrawal reaction if you begin to experience uncomfortable physical or emotional reactions within hours, days, or weeks of reducing or stopping a medication. The same is true if you are experiencing these reactions between drug doses. Short-acting drugs are especially likely to cause withdrawal reactions between doses. These culprits include all of the stimulants, the sleeping pills Ambien and Halcion, and the benzodiazepine tranquilizer Xanax.
You may also be experiencing a withdrawal reaction if the reaction is the opposite of the drug effect. For example, if you get very tired and feel that you are "crashing" a few hours after taking a stimulant - or if you get agitated and even "high" a few hours after taking a tranquilizer - you may be going through withdrawal.
For many reasons, our knowledge of drug withdrawal problems is not as advanced as it should be. To begin with, these problems are very complicated. As in George's case, it can be difficult to separate psychological factors from physical ones. Or it can be difficult to determine whether the problems involve physical withdrawal or a return of the individual's original troubles. In addition, different people vary in terms of their withdrawal reactions to the same drug. For example, George may feel nervous and have trouble falling asleep for a few days after withdrawing from Klonopin, whereas Gina may hear a ringing sound in her ears and find that her physical balance is impaired.
As we have mentioned, doctors are often reluctant to admit that withdrawal is a serious problem. Moreover, when they do witness withdrawal reactions, they rarely report them to the appropriate agencies. For example, probable withdrawal reactions from antidepressants such as Paxil and Zoloft are very common, occurring in 30 percent of patients (Rosenbaum et al., 1998 [323]). Yet the number of reports of all adverse reactions from antidepressants sent to the FDA varies between 2 and 300 per 1 million prescriptions (Young and Currie, 1997 [397]).
Perhaps most important, almost all drug research is paid for and controlled by drug manufacturers, and these profit-driven corporations tend to be very reluctant to identify drug withdrawal problems that might discourage people from using their products. From drug research through medical education and clinical practice, then, many factors interact to inhibit the growth of our knowledge about drug withdrawal. As a result, there is no significant tradition or body of wisdom in psychiatry that focuses on drug withdrawal reactions. All of the emphasis tends to be on how to start and to maintain drugs in patients rather than on how to stop them.
When you talk to your doctor about problems stopping or reducing the dose of your psychiatric drug, keep in mind that your doctor may not know much about the problem or may even be irrationally denying it's existence. For example, the older antidepressants, commonly called tricyclics, have been in use for more than fifty years, and their withdrawal reactions have been repeatedly documented. Yet some doctors seem completely unaware of the existence of these reactions. A survey in 1997 tested the awareness of psychiatrists and general practitioners concerning antidepressant drug withdrawal. The authors concluded that "a sizeable minority of physicians denied being confidently aware of the existence of antidepressant withdrawal symptoms" (Young and Currie, 1997 [397], p. 28). The implication is that, in routine clinical practice, at least a "sizeable minority" of doctors make no attempt to diagnose antidepressant withdrawal reactions or to distinguish them from other symptoms. Another recent survey of psychiatrists showed that withdrawal reactions rated near the bottom of the list of factors helping these physicians decide which antidepressant to prescribe to their patients.
Your doctor may also mistakenly attribute your withdrawal reactions to your "mental illness". Especially if you have unsuccessfully tried to withdraw from the drug previously your doctor may try to convince you that you have a "chronic illness" requiring lifetime drug use. The irony is that the longer you stay on the drug, the more likely you are to suffer something beyond a mild reaction when you attempt to withdraw. Your unsuspecting doctor, and even you, might see this as a sign that you "really need" your drug. In reality what you really need is help in gradually withdrawing.
If a drug causes unpleasant effects to some users when they stop taking it, then some of these users will inevitably become physically dependent on it. This conclusion is basic to the definition of physical dependence. According to the World Task Force on sedative-hypnotics, "[p]hysical dependence is defined as the appearance of specific withdrawal symptoms when the medication is abruptly discontinued"120. Boston University's online Pharmacology Glossary states that dependence "is characterized by the necessity to continue administration of the drug in order to avoid the appearance of uncomfortable or dangerous (withdrawal) symptoms"121. And a report in the New England Journal of Medicine confirms this emphasis: "Use of the term `physical dependence' implies that an objective withdrawal syndrome will occur after a drug is discontinued" (Shader and Greenblatt, 1993 [339], p. 1402).
On the basis of these definitions, we can further conclude that all psychiatric drugs are drugs of dependence. This includes drugs, such as lithium, that are not commonly recognized as causing withdrawal symptoms. The occurrence of withdrawal symptoms defines dependence inasmuch as users will experience discomfort when stopping the drug and will experience relief when restarting it. As a result of experiencing discomfort, some will "choose" to remain on the drug only in order to avoid prolonged withdrawal distress. As already described, they may also mistake the withdrawal symptoms for "mental illness" and decide that they "need" the medication. Some authors reject the idea that drugs like antidepressants can make people dependent or addicted because once users stop taking them, they do not engage in "drug-seeking" behavior, that is, they do not try to obtain the drugs by any means or visit different physicians to coax prescriptions from them. But such behavior would be completely unnecessary for drugs that can be so easily and legally obtained from a doctor's prescription.
The existence of unpleasant withdrawal reactions reflects negatively on a drug. For economic and political reasons, drug manufacturers, regulatory agencies, researchers, and individual doctors tend to downplay, ignore, or deny information about a popular drug's withdrawal effects. They will sometimes do so even after convincing evidence has become available for all to see. The resistance and denial have been documented in years past - notably, with respect to the benzodiazepines, the barbiturates, the stimulants, and even the opiates122. Yet these very drugs are today considered "classically addictive".
A similar denial of the nature, extent, and implications of withdrawal reactions has occurred with respect to the SSRI antidepressants and others (see next page)123. Only lately have drug manufacturers and other important players begun to acknowledge that withdrawal reactions are a serious problem. Ironically, these drugs were never tested for their dependence potential and became popular partly because they were offered as replacements for older antidepressants and benzodiazepines - drugs known to cause withdrawal problems.
Psychiatry's official diagnostic manual, the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV APA, 1994 [6]), contains a diagnostic category entitled "Substance Withdrawal". described as "the development of a substance-specific maladaptive behavioral change, with physiological and cognitive concomitants, that is due to the cessation of, or reduction in, heavy and prolonged substance use" (DSM-IV, [6], p. 184). On the following page, DSM-IV emphasizes the discomfort of withdrawal, stating that the "syndrome causes significant distress or impairment in social, occupational, or other important areas of functioning" (DSM-IV, [6], p. 185).
This definition appears to be describing withdrawal from illicit or disapproved substances. Yet much prescribed, medically approved psychiatric drug use is also "heavy and prolonged". As we document later in this chapter, withdrawal reactions that can cause severe and long-lasting distress and impairment have been specified for all classes of psychiatric drugs. Two panels of experts have also identified a distinct, complex, and sometimes severe withdrawal syndrome for SSRI antidepressants. However, DSM-IV "recognizes" such reactions for only two currently used classes of psychiatric drug: stimulants and tranquilizers (DSM-IV APA, 1994 [6]).
For years, writers in the medical literature have interchangeably used the terms discontinuation, withdrawal, and abstinence syndrome, increasingly, in discussions of psychiatric drugs, the accepted term is discontinuation. In the eyes of some experts, this word is less stigmatizing - to the drugs. We read in an editorial in the British Medical Journal that "[t]he common lay belief that antidepressants are addictive probably contributes to the significant undertreatment of depressive illness. It is important not to foster this belief inadvertently - one reason that `discontinuation reaction' is a better term than `withdrawal reaction' " (Haddad, Lejoyeux, and Young, 1998 [194], p. 1105).
But this is doublespeak. Rather than provide evidence to address justified fears that antidepressants produce dependence, experts urge that the worrisome phenomena be called by a different name! We, on the other hand, suggest that a sound attempt be made to answer important questions: Do "discontinuation" or "withdrawal" effects drive people to remain on their drugs indefinitely? How frequently are these effects mistaken for relapses? And do these effects mistakenly convince doctors that patients "need" their drugs?
Not long ago, similar rationalizations deluded doctors into believing that long-term tranquilizer treatment was "effective". From the start of benzodiazepine use in 1963, the existence of physical dependence in patients taking clinical doses of benzodiazepines was firmly established (Lader, 1991 [239], p. 56). Yet, no fewer than fifteen years later, a prominent expert concluded that "[t]he dependence risk with benzodiazepines is ... probably less than one case per 50 million months in therapeutic use" (Marks, 1978 [266], p. 2). As a result of such denial, millions of people became physically dependent on these drugs.
In the remainder of this chapter, we provide a review of medical information about withdrawal reactions to commonly used psychiatric drugs, including the newer ones on the market. We also discuss lesser-known reactions and withdrawal syndromes. We have reviewed the medical literature of the past several decades, looking for published reports concerning discontinuation. Even if your doctor is unsympathetic to our critical stance about medication use, it would be an excellent idea for him or her to read the remainder of this chapter, as it presents what is probably the most comprehensive summary of psychiatric drug withdrawal reactions available anywhere.
Withdrawal reactions from benzodiazepines - drugs that make up most of the class of minor tranquilizers and ordinarily prescribed sleeping pills - are extremely well documented. For most people who take low doses, the main symptoms of withdrawal consist of increases in tension and anxiety as well as motor and perceptual disturbances. However, the withdrawal reaction can become much more severe and even life-threatening (see below).
Tranquilizers can produce withdrawal reactions after only a few weeks of use. Studies of Xanax, for example, have indicated increased anxiety and panic upon withdrawal after only eight weeks. The longer you take a tranquilizer, the higher the doses, and the more abrupt the withdrawal - the more serious your withdrawal reactions are likely to be. Severe reactions tend to be more frequent with short half-life drugs (APA Task Force, 1990 [4]), Halcion (triazolam, used only as a sleeping pill) and Xanax (alprazolam). Intermediate half-life drugs include Ativan (lorazepam), Klonopin (clonazepam), Lectopam (bromazepam), Restoril (temazepam), and Serax (oxazepam). Longer half-life drugs include Dalmane (flurazepam), Valium (diazepam), Tranxene (clorazepate), and Librium (chlordiazepoxide). Benzodiazepines can reverse withdrawal reactions from alcohol and barbiturates. Consistent with this fact, they produce a withdrawal syndrome similar to that associated with these substances124.
A task force report of the American Psychiatric Association (1990) [4] divides the benzodiazepine "discontinuance syndrome" into three familiar categories:
As just noted, benzodiazepine withdrawal (like alcohol or barbiturate withdrawal) may involve seizures. Adults and older persons appear to be equally at risk, but seizures are most likely in cases where high doses have been used, short half-life drugs have been involved, and withdrawal is abrupt (Fialip et al., 1987 [150]).
Today, many doctors routinely encourage long-term users of benzodiazepines to withdraw from the drugs. However, many of these patients are unwilling to do so because they have found the withdrawal symptoms to be intolerable. They remain on the drugs, thanks to doctors who renew their prescriptions. This is particularly true of older patients, especially women, who were started on benzodiazepines before doctors became more aware of their potential for dependence (Rickels et al., 1991 [317]).
The doctor's involvement before, during, and after withdrawal is one of the most important ingredients for success in difficult withdrawal processes (Ashton, 1994 [20]). Patients who withdraw from benzodiazepines require psychological support. This may range from simple but continued encouragement to more formal anxiety-management techniques such as breathing exercises, mental exercises, meditation, and group support; it can also include psychotherapy.
Your doctor can be more effective by using a clear withdrawal protocol rather than an unstructured, casual taper to "test the waters". It may help if you are simultaneously provided with written instructions on each of the steps involved. Although some doctors prescribe Inderal or Tegretol to minimize benzodiazepine withdrawal reactions, these drugs have not been shown to be of definite help, except perhaps in urgent cases when the benzodiazepines must be withdrawn rapidly125.
The vast majority of patients are successfully withdrawn as outpatients. In medical textbooks, recommendations for benzodiazepine withdrawal sometimes mention that inpatients might be switched to a longer-acting drug such as Valium, because withdrawal symptoms are less acute and the relatively larger equivalent dose may be divided more easily. Other doctors believe, to the contrary; that it is best to work with the drug that the patient is accustomed to.
Some medical sources suggest a "10 percent per day" method of tapering benzodiazepines, but we believe that this poses an undue risk to patients. In cases where doctors feel they must taper relatively rapidly without hospitalizing the patient, a "10 percent per week" schedule is more reasonable. This is also the conclusion reached in a systematic review of withdrawal trials of benzodiazepines: "Progressive withdrawal (over 10 weeks) appeared preferable if compared to abrupt since the number of drop-outs was less important and the procedure judged more favorable by the participants" (Denis et al., 2006). Clinical experience with benzodiazepines also suggests that extension of the withdrawal period is not harmful, especially after the initial dose has been decreased by 50 percent. In ordinary circumstances, however, patients should be allowed to share in controlling the process, especially in regard to slowing it down.
In Chapter 4, we divided antidepressants into four classes: tricyclics, monoamine oxidase inhibitors (MAOIs), Prozac-like drugs that stimulate serotonin, and atypicals. Predictable symptoms of withdrawal from most of these drugs are now well documented, although many psychiatrists and even more general practitioners remain unaware of them.
Stopping the use of tricyclic antidepressants produces a "bewildering assortment" (Wolfe, 1997 [392]) of withdrawal symptoms, affecting between 20 and 100 percent of users and lasting up to three months. You are at highest risk of experiencing withdrawal reactions from tricyclics if you have been treated with high doses for many years and are rapidly withdrawn (Garner, Kelly and Thompson, 1993 [168]).
So many different physical symptoms can result from withdrawing from tricyclic antidepressants that it is helpful to divide them into categories. The following withdrawal symptoms may occur alone or in combination, and to varying degrees.
We have seen cases in which people, after many years on these drugs, have been withdrawn without ever fully recovering from withdrawal symptoms. Some, for example, continue to suffer indefinitely from nausea and memory problems.
As noted earlier in the book, most of these symptoms - including mania and other mental disturbances - can also occur during treatment from direct antidepressant toxicity.
Children, too, suffer from tricyclic antidepressant withdrawal (Law, Petty and Kazdin, 1981 [245]). In one case involving an eight-year-old boy the abrupt withdrawal of a tricyclic provoked such severe nausea, vomiting, and abdominal cramps that the boy had to be hospitalized because of dehydration (Gualtieri and Staye, 1979 [192]). "Mental irritability" and heart irregularities have also occurred in infants born to mothers taking tricyclic antidepressants during pregnancy (Webster, 1973 [383]).
Problems related to discontinuation from monoamine oxidase inhibitors (drugs such as Nardil, Parnate, Eldepryl, and, in Canada, Manerix) are less documented than for other classes of antidepressants, probably because these drugs are less frequently used. The proportion of patients who experience withdrawal reactions is unknown because we lack systematic studies (Dilsaver, 1990; Lawrence, 1985 [246]). What we do know, however, is that if you stop an MAOI, withdrawal reactions can last from "days to weeks" (Wolfe, 1997 [392]).
Some clinicians believe that "... MAOI withdrawal is usually not a serious problem" (Frenkel et al., 1992, p. 111), whereas others maintain that it "produces syndromes of much greater severity than those precipitated by the withdrawal of [tricyclics]" (Dilsaver, 1994 [132], p. 107). One review mentions "serious cognitive impairment and catatonia that may lead to hospitalization, which can appear when [MAOIs] are discontinued" (Schatzberg et al., 1997b [331], p. 5).
Other withdrawal reactions include anxiety, agitation, paranoia, pressured speech, headaches, lowered blood pressure upon standing, muscle weakness, shivering, and tingling, burning sensations under the skin (Dilsaver, 1990; Lawrence, 1985 [246]). Mania can also occur.
Here are two accounts of grave withdrawal reactions from MAOIs.
"A thirty-four-year-old woman was started on a gradual taper from Nardil. She felt burning sensations, and abruptly quit the drug. She became `hostile, loud, aggressive, disoriented in time and place ...'. She `lunged at imaginary objects. ... She was delirious for 3 days after receiving the last dose of phenelzine. ... Though hospitalized for only 72 hours, she did not fully recover for 6 weeks' " (Liskin et al., 1985 [258], p. 46).
"In a case involving a twenty-one-year-old woman treated for bulimia, Nardil was abruptly removed because the woman experienced severe dizziness upon standing. As a result, `[s]he was sleepy, spent most of her time in bed, and was confused for 3 days and disoriented for 4 days following the last dose. ... She did not recognize her mother or her room. She entered a state of catatonic stupor and was hospitalized' " (Liskin et al., 1985 [258], p. 47).
This group of drugs includes the selective serotonin reuptake inhibitors (SSRIs) such as Prozac, Zoloft, Celexa, Paxil, and Luvox. Effexor, Wellbutrin, Remeron, and Cymbalta can cause similar withdrawal reactions to these drugs.
In 1996, and again in 2004, a panel of psychiatric experts sponsored by antidepressant drug manufacturers met to discuss the withdrawal syndrome from SSRIs. The first panel observed that many of the reactions reported "are similar to those of tricyclic withdrawal, but a variety of novel symptoms are also associated with the stoppage of SSRI therapy" (Schatzberg et al., 1997b [331], p. 5). The new symptoms included "problems with balance, sensory abnormalities, and possibly aggressive and impulsive behavior" (Haddad, 1997 [193], p. 21).
That first panel found that the SSRI discontinuation syndrome encompasses the following frequently reported clusters of physical symptoms:
(1) disequilibrium (e.g,, dizziness, vertigo, ataxia), (2) gastrointestinal symptoms (e.g., nausea, vomiting), (3) flu-like symptoms (e.g., fatigue, lethargy; myalgia, chills), (4) sensory disturbances (eg., paresthesias [tingling, burning sensations], sensations of electric shock), and (5) sleep disturbances (e.g., insomnia, vivid dreams) (Schatzberg et al., 1997b [331], p. 8).
The panel also noted two "core psychological symptoms" of SSRI withdrawal - anxiety/agitation, and "dramatic" crying spells and irritability - as well as overactivity, depersonalization, decreased concentration/ slowed thinking, lowered mood, confusion, memory problems, and abnormal movements.
The 2004 panel confirmed all of the first panel's observations, as well as ours, and recognized both the astonishing variety of reported symptoms (over fifty) and the uniqueness of most cases. The new panel observed that the SSRI withdrawal syndrome "usually" appeared within one to seven days after discontinuation, even if the drug was taken for only one month. It estimated that the syndrome "usually" could last up to three weeks. The new panel reorganized the array of withdrawal symptoms into six clusters: (1) neurosensory (including vertigo, tingling and burning sensations, shock-like reactions, and various intense nerve and muscular pains), (2) neuromotor (tremor, spasms, loss of muscular control, and visual changes), (3) gastrointestinal (nausea, vomiting, diarrhea, loss of appetite and weight loss), (4) neuropsychiatric (anxiety, depressed mood, suicidal ideation, irritability impulsiveness), (5) vasomotor (heavy sweating, flushing), and other neurologic symptoms (insomnia, vivid dreaming, fatigue, chills). The panel observed that these groups of symptoms tend to occur together, with most patients experiencing symptoms from each group. It also confirmed previous observations that a full withdrawal syndrome could emerge from missed doses and drug holidays127.
Based on our own clinical experience, on the information reviewed by this panel of psychiatric experts, as well as on other recent studies and reviews128, we have found, as with tricyclic withdrawal, a vast assortment of symptoms associated with the SSRI discontinuation syndrome. Here are a few examples.
"A thirty-two-year-old man who discontinued Prozac after six months of use awoke with `painful extensor muscle spasms' and `protruding tongue movements' (Stoukides and Stoukides, 1991 [354]). A thirty-year-old woman tried to stop taking Luvox when she became pregnant but `was overwhelmed by strong feelings of aggression (she felt that she `could murder someone')'. This occurred on two distinct withdrawal occasions, and the woman was unable to stop the Luvox" (Szabadi, 1992 [359]).
"One of us has reported a case of `crashing' from SSRI withdrawal. A woman spontaneously decided to reduce her Zoloft from 100 mg to 50 mg per day. Within a couple of days, she lapsed into `exhaustion and fatigue, profound depression, and a compulsive, alien-feeling desire to kill herself'129. All of these symptoms disappeared soon after the woman resumed the 100 mg dose. In another publication, we describe a young woman who became suicidal when discontinuing Prozac" (Breggin, 1992a [50]).
"In two cases involving middle-aged men, neither with a history of major psychiatric problems, Paxil withdrawal led to severe symptoms. For twelve days after abrupt withdrawal, one man experienced marked hypomania. The second man developed various physical symptoms, then began to express `intense' homicidal thoughts - a condition that lasted for five weeks" (Bloch et al., 1995 [41]).
"One woman on Lexapro reduced her 10 mg per day dose to 5 mg a day and three weeks later stopped altogether. One week later, she began to experience `electric shock-like sensations or visual flashes lasting for about 1 second each. This was followed by a phase of spatial disorientation that lasted for about 30 seconds and was experienced as highly unpleasant and frightening'. This occurred about three times a day over two weeks" (Feth et al., 2006 [149]).
"One report describes three consecutive female patients who experienced `severe physical symptoms of withdrawal' when stopping Effexor. Effexor is not an SSRI, but like the SSRIs it stimulates serotonin. The women could not stop the drug, even after repeated attempts at tapering. These patients were finally able to discontinue the Effexor only by switching indefinitely to Prozac" (Giakas and Davis, 1997 [173]).
"After three months of taking Prozac, a young woman tried unsuccessfully to withdraw from the drug on three different occasions. Each time, she experienced extreme dizziness and instability. She was examined by specialists who ordered many tests, including magnetic resonance imaging (MRI) of the brain. The tests were negative. The symptoms were relieved each time by the re-administration of Prozac, but none of her doctors suspected a withdrawal reaction. She was eventually tapered successfully over twelve weeks" (Einbinder, 1995 [139]).
"A thirty-two-year old woman took 300 mg of Effexor daily for eight months. She tried to withdraw abruptly on three occasions but failed due to unbearable headaches, gastrointestinal distress, fatigue, and other symptoms. `She remains on a regimen of venlafaxine [Effexor], 100 mg t.i.d.' (Farah and Lauer, 1996 [145]). A man withdrawn from Effexor experienced `severe akathisia' (compulsion to move). This condition abated `within hours' of restarting the drug, which was later gradually tapered and withdrawn" (Wolfe, 1997 [392]).
In a lengthy paper on antidepressant withdrawal difficulties, and in a Web site chronicling his relentless querying of public officials in Britain about the safety of medicines, journalist Charles Medawar summarizes numerous reports of SSRI withdrawal reactions and the underreporting of these reactions by doctors, as well as the stonewalling by government agencies and drug manufacturers about their examination of SSRI withdrawal risks130. Included among these reports, which were dated from 1988 to 2003 and numbered more than one hundred and twenty-five, were a dozen that noted distinct withdrawal symptoms observed in newborn infants whose mothers took either SSRIs or other antidepressants during pregnancy. Much like the older antidepressants, the SSRIs can on occasion cause withdrawal mania (Goldstein et al., 1999 [181]).
Withdrawal reactions affect between 20 and 80 percent of persons who stop SSRIs abruptly. Based on a conservative estimate that 50 percent of individuals suffer from withdrawal reactions, we would have to conclude that hundreds of thousands of people are affected every year in the United States alone. Most of these reactions are mild or moderate, yet can be distressing enough that patients would wish to avoid them. Few valid estimates of severe reactions exist: Rates of 20-30 percent were reported in a symposium (Thompson, 1998 [367]).
Withdrawal reactions from SSRIs typically occur after one to four days of discontinuation, although they may begin weeks later in the case of long half-life drugs such as Prozac. There seems to be a consensus in the published literature reporting on clinical studies and surveys of consumers that short half-life drugs, especially Paxil, seem most often implicated. On average, withdrawal reactions persist for seven to twenty-five days (from one day to thirteen weeks). One study of Cymbalta did find that patients taking the highest dose reported significantly more withdrawal symptoms than those on the lowest doses (Perahia et al., 2005 [302]). As yet, however, the number of studies remains too small to draw valid conclusions about the relationship between dose and length of drug use, on the one hand, and the risk of withdrawal reactions, on the other.
SSRI withdrawal reactions involve many different physical, emotional, and mental symptoms, the full range of which has probably not been charted (Hindmarch et al., 2000 [202]). They occur with both abrupt and more gradual withdrawal but seem to be attenuated by a truly gradual withdrawal that lasts three months or more. In our estimation, however, abrupt withdrawal is too frequent, regardless of whether medical supervision is provided.
Withdrawal reactions appear to be more frequent - or at least more acute - in the case of SSRIs with shorter half-lives, such as Paxil, Luvox, and Zoloft. Withdrawal reactions from longer-acting drugs like Prozac seem to appear much later. They can begin up to twenty-five days after the drug is stopped. Because they are so delayed, they are not visible in short-term studies; moreover, patients tend not to attribute these delayed withdrawal reactions to stopping the drug (Pollock, 1998 [310], p. 535).
GlaxoSmithKline, the manufacturer of Paxil, was recently required by the FDA to modify it's label to add emphasis to the danger of withdrawal reactions. This label change occurred at the time that Peter Breggin was acting as a medical expert in a California lawsuit aimed at requiring similar modifications. Under "Precautions", the new Paxil label lists the following reported withdrawal symptoms: "Dysphoric mood, irritability, agitation, dizziness, sensory disturbances (e.g,, paresthesias such as electric shock sensations and tinnitus), anxiety confusion, headache, lethargy, emotional lability, insomnia, and hypomania" (Physicians' Desk Reference, 2007 [308], p. 1533). The section refers to the occurrence of "serious discontinuation symptoms" and "intolerable symptoms".
We have found that a small number of patients cannot tolerate the anguish associated with withdrawing from Paxil and similar antidepressants, and instead reluctantly decide to remain on the drugs indefinitely. In such cases, we would advise consulting with more than one health provider about approaches to withdrawal. Further attempts to withdraw at a later date should not be ruled out since the intensity of withdrawal reactions may vary at different times.
As M. Lejoyeux and J. Adès (1997) [249] point out, "Patients who are classified as having a relapse while they are discontinuing therapy may, in fact, be suffering from unrecognized discontinuation symptoms" (Lejoyeux and Ades, 1997 [249], p. 11). As previously noted, the 1996 expert panel found that "dramatic" crying spells are a "core psychological symptom" of SSRI withdrawal. No wonder it's tempting to mistake this reaction for a relapse of depression. It is anybody's guess as to how many patients are resumed on antidepressants because they are suffering from withdrawal reactions that are mistakenly diagnosed as depressive relapses, but the number is undoubtedly large.
Withdrawal reactions to the atypical antidepressants are generally less well documented than for other types of antidepressants. There exists scarcely any information on the effects of withdrawing from Ludiomil (maprotiline). Withdrawal reactions from Wellbutrin or Zyban (buproprion), Serzone (nefazodone), and Remeron (mirtazapine) are likewise barely documented in the medical literature. One report describes a withdrawal syndrome from abrupt cessation of Serzone, consisting of "dizziness, nausea, vomiting, sweating, anxiety insomnia, and restlessness", which lasted three days (Benazzi, 1998a). Another report describes a similar reaction (which also included burning sensations under the skin) following the sudden withdrawal of the antidepressant Remeron (Benazzi, 1998b).
Withdrawal reactions to Effexor (venlafaxine) have been discussed in the previous section, as this drug has many effects in common with Prozac-like drugs that stimulate serotonin. One researcher found that serious withdrawal effects within mere hours of stopping Effexor or reducing the dose can affect physical and motor coordination to such a degree that patients should be explicitly advised not to drive a car (Campagne, 2005 [81]). Asendin (amoxapine), a combination of a neuroleptic and a tricyclic antidepressant, will share important withdrawal effects with antipsychotics and tricyclics (such as flu-like symptoms, agitation, and movement disorders), and readers should consult the appropriate sections in this chapter.
Several withdrawal effects have been documented with Desyrel (trazodone), including mania and hypomania, nausea and recurrent vomiting, and visual hallucinations131. Within 36 hours of trazodone withdrawal, one patient developed a "syndrome of overwhelming anxiety, depersonalization, insomnia, and nightmares" which took five days to subside (Menza, 1986 [276]).
As discussed in previous chapters, the absence or rarity of reports of withdrawal effects from specific psychiatric drugs in no way indicates that a drug is free of such effects. The reader withdrawing from atypical antidepressants should proceed slowly and expect almost any kind of effect associated with other antidepressants.
Stimulants are frequently used to control the behavior of children. The amphetamines Adderall, Desoxyn, Dexedrine, and Gradumet, as well as the amphetamine-like Ritalin, are the most commonly prescribed. Stimulant drugs are also given to adults to treat "attention deficit-hyperactivity disorder", narcolepsy depression, and obesity other stimulant-like drugs have been used for diet control, including phentermin (Fastin, Adipex) and mazindole (Sanorex). Another diet control drug, fenfluramine, was recently taken off the market in Canada and the United States because it harms human heart valves. Caffeine, of course, is a mild stimulant that is commonly used.
Ritalin and the amphetamines are very similar to cocaine in terms of how they affect brain chemistry and function. Although the impact of stimulants on dopamine neurotransmission is usually viewed as the main cause for the euphoria or "pleasure" that encourages people to keep using these drugs, recent research has also focused on another neurotransmitter, serotonin. Cocaine and other stimulants block the removal of serotonin from the synapses between brain cells (Rocha et al., 1998 [319]). Since the SSRI antidepressants have similar effects, it is not surprising that withdrawal from stimulant drugs and from Prozac-like SSRI antidepressants share many features.
Described in many case reports for nearly three decades, "stimulant withdrawal was largely overlooked [by doctors] for years" (Dackis and Gold, 1990, p. 16). Clinicians have tended to dismiss the severity of stimulant withdrawal or to characterize it as "merely psychological". Because stimulant withdrawal typically lacks some of the more visible physical manifestations of withdrawal from drugs classically used to excess, such as alcohol and the opiates (Lago and Kosten, 1994 [241]) it is easier to overlook.
"Crashing" is the best-known effect of withdrawing from stimulants. During this state, you are likely to feel emotionally upset, and to lack energy and motivation. This depressive, fatigued state is the result of your brains attempt to overcome the previous state of artificial stimulation. Similarly if stimulants cause you to be less hungry a marked increase in appetite and weight gain may accompany withdrawal.
The deep depression and apathy usually last no more than three to ten days. However, as confirmed in a psychopharmacology manual, they may "reach serious clinical proportions" and "persist for weeks" in "unstable individuals" (Schatzberg, Cole, and DeBattista, 1997 [331], p. 352). Thoughts of despair and suicide may accompany this "crashing". A longer phase of general physical and mental slowing follows it. Paradoxically you may also suffer from insomnia, anxiety and irritation.
DSM-IV specifies certain criteria for a withdrawal syndrome associated with the group of stimulant drugs that includes Ritalin and cocaine. The symptoms it lists are fatigue; vivid, unpleasant dreams; insomnia or excessive sleep; increased appetite; and psychomotor retardation or agitation (DSM-IV APA, 1994 [6]), pp. 208-209, 225-226).
Rebound reactions are also common. In the case of amphetamines and Ritalin, rebound phenomena show up when children are taken off the drug abruptly or miss a dose. Typically they experience an increase in agitation, restlessness, excitability; and distraction (Whalen and Henker, 1997). Rebound reactions are, by definition, more intense than the same symptoms experienced before taking the drug. They may occur within hours of the last dose of a stimulant and can persist for days.
In a double-blind controlled study parents and teachers of "normal" boys given a single dose of amphetamine observed marked rebound phenomena five hours after the drug was administered (Rappoport et al., 1978 [315]). There is no doubt that parents, teachers, and doctors repeatedly mistake these rebound reactions for signs that the child's "attention deficit-hyperactivity disorder" is worsening and that the child "clearly needs" the drug.
There are case reports of delirium, psychosis, and severe confusional states upon amphetamine withdrawal132. Similarly, researchers have documented psychosis and deep depressive symptoms with suicidal thoughts upon withdrawal of Ritalin in children (Klein and Bessler, 1992 [233]). There is a case report of stuttering priapism (intermittent, long-lasting painful erections) in an adolescent withdrawn from Concerta (Schwartz and Rushton, 2004 [337]). As we discussed at the beginning of this chapter with regard to adverse reaction reports sent to the FDA, the relative paucity of reports of withdrawal-induced distress is not a good indicator of their true frequency.
Most of our knowledge about withdrawal reactions after stimulant use comes from studies of adult cocaine users. Amphetamine withdrawal has been studied less, and methylphenidate (Ritalin) withdrawal, even less still133. Millions of children take stimulants daily. It is truly remarkable that researchers are not systematically investigating this hazardous phenomenon.
Should withdrawal from stimulants be abrupt or gradual? The specialized literature contains recommendations to withdraw stimulants abruptly, on the grounds that no dangerous physical symptoms will occur. One psychopharmacology textbook is quite explicit: "When a patient who is dependent on stimulants is hospitalized, stimulant administration should be stopped abruptly. No tapered withdrawal is necessary" (Schatzberg, Cole, and DeBattista, 1997 [331], p. 352). The writers of that passage nonetheless confirm that withdrawal from large doses will "often" produce "a withdrawal syndrome consisting of depression, fatigue, hyperphagia [excessive eating], and hypersomnia" (Schatzberg, Cole, and DeBattista, 1997 [331]). Yet, rather than recommend a gradual withdrawal to attenuate incapacitating reactions, the authors speculate on the value of treating them with antidepressants!
We see no reason to endorse abrupt withdrawal from stimulants, unless they have been used for a short time or sporadically with few or no ill effects between episodes of use. Two doctors who described an arduous withdrawal of Ritalin from two older men emphasized that "a drug free state may not easily be achieved by following the widely accepted recommendation to stop daily stimulant doses abruptly" (Keely and Licht, 1985 [224], p. 123).
Lithium withdrawal raises the serious issue of confusing withdrawal difficulties with relapse. In fact, lithium withdrawal reactions exactly mimic the manic symptoms that lead to the start of lithium treatment. Doctors prescribe lithium primarily to treat mania; often, mania rapidly follows lithium withdrawal. How is the doctor or patient to know whether this condition is withdrawal mania or a return of the original psychiatric problem?
Some doctors refuse to see a specific syndrome of withdrawal from lithium. Others suggest that a true lithium withdrawal syndrome cannot merely resemble mania but that "tremor, dizziness, and sometimes epileptic seizures" should also be observed (Schou, 1993 [336], p. 515). We believe, however, that the latter opinions are further instances in which doctors have thoughtlessly attributed disturbed reactions after drug withdrawal to the patient's "underlying illness" rather than to the drug treatment itself.
The rapid recurrence of mania can happen even among patients who have been taking the drug for years, are seemingly "well-stabilized", and are withdrawn from it for only four or five days134. In one study researchers abruptly switched twenty-one previously manic patients to placebos. They had taken lithium continuously for an average of four years. The authors write: "Within 14 days on placebo, 11 patients relapsed into severe psychotic states with paranoid, manic and depressive syndromes. ... Most of the other patients not relapsing into psychotic states reported anxiety nervousness, increased irritability and alertness, [and] sleep disturbances" (Klein et al., 1981 [232], p. 255).
One review systematically examined fourteen published studies of lithium withdrawal involving 257 "manic-depressive" patients with an average of thirty months of stable lithium treatment. Of the new manic episodes that occurred, more than 50 percent were experienced within three months of withdrawal. Also observed was a staggering 28-fold increase in the risk of new manic episodes for patients just withdrawn from lithium (Suppes et al., 1991). Another review of published and unpublished studies on lithium withdrawal yielded similar conclusions (Faedda et al., 1993 [144]). This evidence led one doctor to state candidly: "[F]rank manic symptoms are the defining feature of significant withdrawal effects and appear to be of a comparable severity to those seen in manic illness generally often requiring hospital admission" (Goodwin, 1994 [182], p. 149).
Withdrawal from lithium increases the likelihood of rebound depression as well as mania. A study of 14 patients and 28 controls showed that lithium withdrawal led to "a high immediate relapse rate". However, once these patients recovered from the rebound withdrawal reaction, their overall outcome over the next seven years was "not worsened by discontinuation". This is good news for patients who are considering withdrawal from lithium. According to this study once they have recovered from the withdrawal reaction, their prognosis is as good as for patients who remain on lithium (Cavanaugh et al., 2004 [84]).
Increases in energy and alertness, heightened emotional response, increased concentration, and decreased thirst are sometimes reported even in patients who do not experience mania during withdrawal (Balon et al., 1984 [27]).
An editorial in the British Journal of Psychiatry openly states that when patients take lithium for less than two years, they risk undergoing frequent recurrences of manic episodes shortly after withdrawal. The author, a psychiatrist, believes that these are genuine withdrawal effects and recommends informing patients of the danger before they decide to undertake lithium treatment (Goodwin, 1994 [182]). Nonetheless, the same psychiatrist maintains that taking lithium for more than ten years does not carry this withdrawal risk. This conclusion contradicts well-accepted clinical experience indicating that the longer a drug is taken, the greater is the risk of withdrawal reactions.
Psychiatrists are belatedly beginning to realize that the rapid recurrence of mania after lithium withdrawal is a real withdrawal effect. However, many practicing doctors undoubtedly continue to attribute withdrawal-induced mania to their patients' "chronic disease". These doctors then persist in exposing their patients to lithium despite it's many long-term adverse effects.
Over the past decade, the use of lithium to treat mania or bipolar disorder has been mostly replaced by a crop of drugs originally developed to treat epilepsy. Most of these anti-seizure or anticonvulsant drugs, which include older drugs such as Tegretol and Depakene, and newer drugs such as Topamax and Neurontin, have not been approved by the FDA to treat mania or bipolar disorder. However, Lamictal and Depakote are now approved for treating manic disorders.
Several previous reports described anticonvulsant withdrawal reactions, focusing on the risk of seizures in patients having received these drugs to treat epilepsy or other seizure disorders. Now that hundreds of thousands of people who do not suffer from seizures are prescribed anti-seizure drugs, withdrawal reactions are likely to be more varied. Nonetheless, as has occurred previously with every wave of "newer" drugs, there is still a dearth of published reports of withdrawal reactions.
Tegretol (carbamazepine) is an anticonvulsant drug widely used as a treatment for many problems, including mania. Upon withdrawal, Tegretol can provoke serious emotional flare-ups - including paranoia, hostility, and agitation-in previously disturbed people (Heh et al., 1988 [201]). These withdrawal reactions can also occur in mentally stable individuals, as in one documented case involving a patient treated for a physical disorder (Darbar et al., 1996 [127]). Although probably less frequently than withdrawal from lithium, withdrawal from Tegretol can also cause rebound mania (Jess et al., 2004 [216]). Other anticonvulsants or anti-epileptic drugs such as Depakene (valproic acid), Depakote (divalproex sodium), and Dilantin (phenytoin), are also widely prescribed in psychiatry. The risk of seizures needs to be considered when any anticonvulsant drug is withdrawn, whether or not a seizure disorder was previously present. Other symptoms of anticonvulsant withdrawal may commonly include anxiety, muscle twitching, tremors, weakness, nausea, and vomiting. A case of withdrawal reaction in an eighty-one-year-old man maintained for five years on Neurontin (gabapentin) and then tapered over one week described "severe mental status changes, severe somatic chest pains, and hypertension" appearing ten days after the last dose. Upon reintroduction of the drug, the symptoms subsided within 24 hours. The authors propose that "a gabapentin taper should follow a course similar to that of a benzodiazepine taper-slowly and over a period of weeks to months" (Tran et al., 2005 [369]).
Antipsychotic or neuroleptic drugs include Clozaril, Haldol, Mellaril, Navane, Prolixin, Risperdal, Zyprexa, and others listed in Chapter 4. They produce many physical and psychological withdrawal reactions, including weight gain, abnormal movements, and psychosis. These reactions vary from merely unpleasant to life-threatening, from transient to irreversible. Antipsychotics have been widely used for fifty-five years, and distinct withdrawal syndromes were documented in the first decade of their use, let, as with most psychiatric drugs, these syndromes have not been sufficiently studied (Tranter and Healy, 1998 [370]).
Interest in antipsychotic withdrawal syndromes has been renewed, V however. As we discuss in this section, researchers now realize that abrupt withdrawal - an exceedingly common practice - escalates rates of "schizophrenic relapse". Furthermore, after withdrawal from atypical antipsychotics such as Clozaril, a violent reaction characterized by delirium with psychosis may occur. Interest in withdrawal has also been sparked by the nursing home reform initiated by the Omnibus Budget Reconciliation Act of 1987 (OBRA). The act mandated nursing homes and other institutions receiving federal funds to implement dose reductions and withdrawals of neuroleptic and sedative drugs.
A common type of antipsychotic withdrawal reaction resembles withdrawal from tricyclic antidepressants. It involves rebound of the brains cholinergic neurotransmitter system. Because neuroleptics such as Mellaril, Navane, Thorazine, and Stelazine suppress this system, it rebounds, or "goes into overdrive", when the drugs are removed (Eppel and Mishra, 1984 [143]). Other neuroleptics, such as Haldol and Compazine, have strong antinausea effects; when they are withdrawn, nausea is a likely reaction.
Cholinergic rebound produces unpleasant symptoms similar to a bad attack of flu, such as nausea, vomiting, diarrhea, headaches, chills, sweating, and a runny nose. Emotional upset frequently occurs as well. These symptoms usually last about one to four weeks, depending on how long the drug has been used. However, we are aware of some patients who have suffered lasting effects. Especially in cases where the vomiting becomes severe or breathing becomes difficult, medical attention may be required. In children, "the reaction is often more severe, and it may on occasion occur after missing a single dose"135. Clinicians have noted that "[f]igures of up to 50 percent of subjects affected by [cholinergic rebound] withdrawal seem standard" (Tranter and Healy, 1998 [370], p. 308).
A second group of withdrawal reactions involves abnormalities of movement. Because antipsychotics suppress the dopaminergic neurotransmitter system, which controls voluntary movements, removal of the drug can result in involuntary spasms, twitches, tics, tremors, and other muscular movements. Withdrawal reactions may include a "Tourette-like syndrome" characterized by motor tics and vocalizations136. Many of these movements involve the face and neck, but any muscle function may be affected, including talking, swallowing, and breathing. These withdrawal reactions can be painful, disfiguring, and disabling. They are more fully described in Chapter 4 in connection with tardive dyskinesia.
In some cases, the movements gradually cease after a few weeks; this phenomenon is known as withdrawal emergent dyskinesia. (Other names for it are withdrawal parkinsonism, withdrawal dystonia, and withdrawal akathisia). If the movements persist for four weeks or longer, they are diagnosed as tardive dyskinesia or one of it's variants. In about one third of cases, these movements lessen substantially or, in very few cases, disappear gradually over a period of months. Most cases persist indefinitely. Accompanying these movement disorders are unpleasant mental states such as depression and indifference in cases of parkinsonism, and anxiety despair, and anger in cases of akathisia137.
Not everyone views these abnormal movements as true withdrawal reactions, for the simple reason that they are ordinarily first observed during drug use (Chapter 4). In many cases, however, abnormal movements become apparent or worsen only when the patient is reducing or stopping the drug138. Astonishingly, one review of antipsychotic withdrawal reactions does not mention abnormal movements (Dilsaver, 1994 [132]).
A third group of withdrawal reactions involves a wide range of psychological and behavioral symptoms, including insomnia, anxiety, agitation, irritability, and organic psychosis. Psychotic withdrawal symptoms are variously called tardive psychosis, super-sensitivity psychosis, or withdrawal psychosis139. Frequently accompanied by abnormal movements, they include hallucinations, delusions, confusion, and disorientation.
Tardive psychosis is considered controversial by some investigators, and researchers have reported widely different rates. However, we believe that clinical and scientific evidence confirms it's existence (Reviewed in Breggin, 1997a [55]). After years of suppression of the dopamine system by these drugs, the brain compensates for their effects. When the drugs are discontinued, the hyper-aroused dopamine system takes over. Psychotic reactions upon abrupt withdrawal have been observed in individuals with no history of psychotic symptoms, such as patients taking antipsychotics for tic disorders (Silva et al., 1993 [345]).
Among developmentally disabled individuals who have been treated with neuroleptics, withdrawal frequently leads to profound emotional disturbances. This worsening of behavior, called "withdrawal-induced behavioral deterioration" (Sovner, 1995 [348], p. 221), usually lasts for several months (Gualtieri, 1993 [191]) and can become permanent. However substantial number of people may improve if kept off the drugs. One author notes that "it is not generally appreciated that severe behavioral symptoms which emerge during neuroleptic dosage reduction can be transient". As a result, "[i]f symptoms do appear during drug taper, the usual clinical practice is to reintroduce the drug, often at doses higher than the original" (May et al., 1995 [268], p. 156).
Cases of neuroleptic malignant syndrome (see Chapter 4) have also been documented upon neuroleptic withdrawal. NMS involves impaired consciousness, abnormal movements, fever, and other symptoms, sometimes ending in the patients death140.
Withdrawal reactions from Clozaril (clozapine) involve not only cholinergic and dopaminergic super-sensitivity but also rebound. Several case reports describe a quickly occurring withdrawal syndrome "with rapid onset of agitation, abnormal movements, and psychotic symptoms". According to one author, "These appear to be much more rapid than syndromes seen after stopping conventional [neuroleptics]" (Pies, 1998 [309], p. 176). Others have found that the "severe agitation and psychotic symptoms ... resolved rapidly and completely upon resumption of low doses of clozapine" (Stanilla et al., 1997 [352]).
Two cases of sudden emergence of brand-new obsessive-compulsive symptoms, one of which also involved "Tourette's syndrome-like tics", have been described after clozapine withdrawal. In both cases, clinicians observed a "complete disappearance" of symptoms upon resumption of clozapine (Poyurovski et al., 1998 [311]). Another person treated for two years developed "severe insomnia, restlessness, and chills" after clozapine was withdrawn. These symptoms disappeared when clozapine was reinitiated, appeared three months later when doctors again withdrew it, and disappeared once more upon readministration of the drug (Staedt et al., 1996 [351]). In another study, twenty-eight patients were evaluated seven days after abrupt clozapine discontinuation; of these, 61 percent developed withdrawal reactions. Most of the reactions were rated "mild" with agitation, headache, or nausea; but four patients experienced more uncomfortable withdrawal involving nausea, vomiting, or diarrhea. Another experienced "a rapid-onset psychotic episode requiring hospitalization" (Shiovitz et al., 1996 [343]). The study does not specify how many patients with "mild" agitation during the first seven days after withdrawal went on to develop more severe episodes of agitated behavior. According to a recent report describing four serious cases of Clozaril withdrawal, two patients were unable to walls, another had a lurching gate, and two gagged while eating or drinking (Ahmed et al., 1998 [3]).
One article entitled "Clozapine as a Drug of Dependence" confirms the existence of a "clear somatic withdrawal sign" in laboratory animals after treatment with clozapine (Goudie et al., 1999 [185]). The drug was seen as the prototype for atypical neuroleptics when first introduced in the United States in 1990. Over time, so-called atypical neuroleptics or antipsychotics such as Risperdal, Seroquel, Zyprexa, Geodon, Ability and Invega - enthusiastically promoted as safer drugs than the older neuroleptics - seem to increasingly resemble the conventional drugs they were meant to replace.
Like the SSRI antidepressants, Seroquel has actions on the serotonin system, and it's not surprising that some withdrawal reactions from it might resemble those of SSRI drugs. One case report described a thirty-six-year-old woman on 100 mg a day of Seroquel who, upset over her large weight gain on the drug, decided to stop the drug and was advised by her doctor to reduce the dose by half. "After one day, she reported nausea, dizziness, headache and anxiety severe enough to preclude her normal activities." Her symptoms continued despite returning to a higher dose. A more gradual reduction of 12.5 mg every five days, aided by an anti-nausea drug, also failed. She was successful on a third attempt when her doctors added an older antipsychotic drug to her regimen, although her nausea persisted for two more days after her last dose (Kim and Staab, 2005 [226]).
In cases where patients are withdrawn from extended antipsychotic use, much of what gets called "schizophrenic" or "psychotic" relapse may actually be unrecognized withdrawal reactions. Withdrawal symptoms such as agitation, restlessness, and insomnia are also likely to be mistakenly attributed to the patients mental condition. By now, this situation may be depressingly familiar to the reader. It reflects the same confusion that we noted in regard to withdrawal from antidepressants, minor tranquilizers, lithium, and stimulants.
Indeed, because of the resemblance between many withdrawal symptoms and patients' prior emotional problems, clinicians not only blame the reaction on the "underlying disorder" but also recommend continued treatment with the offending agent. These clinical errors persist despite telltale signs such as rapid onset of the symptoms after the drug is discontinued and rapid relief after it is reinstated. They persist even despite observations that "[r]eal psychotic relapse rarely occurs during the first weeks of withdrawal" (Perenyi et al., 1985 [303], p. 430).
The controversy over the value of antipsychotic treatment will likely increase as clinicians continue to observe differing relapse rates between abruptly and gradually withdrawn patients; as it turns out, many patients who are gradually withdrawn do not need any further medication. According to one study (Viguera et al., 1997 [380]), "There is a nearly twofold difference in relapse risk between abrupt and gradual withdrawal of [antipsychotic] agents" (Pies, 1998 [309], p. 176), whereas an earlier comprehensive analysis concluded that the relapse rate among patients withdrawn from antipsychotics is three times greater than that among patients kept on the drugs (Gilbert et al., 1995 [175]). However, when other researchers reanalyzed these data to distinguish between patients quickly withdrawn and those more gradually tapered, the latter had one-third the relapse rate of the former (Baldessarini and Viguera, 1995 [25])! In terms of relapse rates, then, prolonged drug treatment appears to be no better than a gradual tapering (See Cohen, 1997a [92]).
Of course, prolonged treatment is much more dangerous. Indeed, researchers are beginning to suggest that brain changes resulting from long-term drug treatment may create "pharmacologic stress factors" or "iatrogenic-pharmacologic stress effects" that, combined with abrupt withdrawal, increase the vulnerability to relapse (Baldessarini and Viguera, 1995 [25]). Taking drugs and then abruptly stopping them might weaken patients, capacity to respond to stress in the future.
The circumstances in which antipsychotics withdrawal should be strongly considered or attempted are so numerous that full coverage is not possible. We can, however, provide a few guidelines that reflect generally recognized medical knowledge and practice:
Given the high probability of irreversible effects of antipsychotics, any doctor following patients who are undergoing long-term treatment must seriously consider lowering the dose to the absolute minimum. In many such patients, the minimal dose will actually be zero. In general, patients should be given neuroleptics for the shortest possible time in the smallest possible doses. We believe that a more rational practice of psychiatry would eliminate the use of such dangerous medications.
Research on OBRA-87, the congressional legislation regulating the use of psychotropics and physical restraints in nursing homes, shows that it had the intended impact, especially in terms of reducing the use of neuroleptics (Snowden and Roy-Birne, 1998 [347]). The use of benzodiazepines, also targeted by the legislation, has likewise decreased. We still need better documentation of how these changes are improving the residents, quality of life. Meanwhile, however, it is clear from several studies that up to half of institutionalized elderly patients can be successfully withdrawn from neuroleptics. Tapered withdrawals are usually successful, residents infrequently display increased levels of agitation, and most are able to remain off the drugs for extended periods (Bridges-Parlet et al., 1997 [78]). Overall, these drugs are so dangerous to the elderly that they should not be prescribed to them.
Antipsychotic drugs should be completely stopped at the first sign of any abnormal movements. Too often doctors try to guess whether or not a particular abnormal movement, like eye blinking or grimacing, is related to tardive dyskinesia. Any abnormal movement should be considered a probable manifestation of tardive dyskinesia when the patient is taking antipsychotic drugs, including the newer ones; and the drug should be stopped as quickly as possible.
Some physicians believe that the newer antipsychotic drugs are less likely to cause tardive dyskinesia. At present, the data is insufficient to make this claim and the FDA requires the same tardive dyskinesia warning for all antipsychotic drugs. Nearly all of the antipsychotic drugs, including Risperdal and Zyprexa, suppress function of the dopamine nerves and therefore can cause tardive dyskinesia.
An occasional unscrupulous report claims that the newer antipsychotics can "improve" tardive dyskinesia because they tend to suppress the symptoms. In reality; any drug that suppresses tardive dyskinesia is likely to cause it as well. This phenomenon is called masking. Giving the newer antipsychotics to control tardive dyskinesia is likely to worsen the disorder in the long run.
In non-emergency situations where tardive dyskinesia is not suspected, antipsychotic drugs should be withdrawn slowly, keeping in mind that it is not unusual to take at least one month to withdraw for every year of exposure, so that a patient who has taken neuroleptics for two years is likely to require at least two months to withdraw (see Chapter 8). Also keep in mind the risk of withdrawal psychosis. A temporary worsening of the patient's condition is likely toward the end of withdrawal or shortly after the drug is stopped. This should not in itself prevent eventual, if gradual, withdrawal from antipsychotic drugs.
Primate research confirms that withdrawal should proceed very slowly because drug-induced changes may last for several months after the last dose. In a laboratory study, thirty-eight healthy monkeys were observed daily for 108 weeks. For the first 25 weeks, "baseline" information was gathered. From week 25 to week 72, the monkeys received long-lasting injections of fluphenazine (Prolixin). The main findings during this period were "highly significant decreases in self- and environment-directed behaviors and affiliation", meaning that the animals paid much less attention to themselves, their companions, and their surroundings. As of week 73, drug injections were ceased, and "[a]ggression showed some increase during early drug discontinuation, accentuated by stress". (This outcome resembles the withdrawal agitation observed in humans). Importantly, the investigators state that "[r]ecovery of normal (baseline) behavioral scores began by week 7 after the last treatment". In other words, it was not until two months after the last dose that the animals began to display normal behavior. Ultimately, "tardive dyskinesias persisted in 30 percent of the animals for a prolonged time" (Lifshitz et al., 1991 [257]).
Based on clinical experience bolstered by research, we suggest that if you are withdrawing from neuroleptics, you should avoid making major changes in your life during and shortly after withdrawal. Allow for a "coasting" period of at least several weeks to shake off drug effects, with emphasis on improved nutrition, exercise, and general well-being. In particular, strengthen your social relationships and avoid new, unpredictable sources of stress and tension.
Rebuilding your life without resort to drugs after years of antipsychotics can be especially difficult. You may have acquiesced to taking drugs that you actually experienced as toxic and mind-numbing, and this submissiveness, compounded by drug effects, may have weakened your ability to make independent decisions. You will benefit from careful planning and preparation, building a support network, and practicing nondrug options to deal with stress and anxiety.
Antiparkinsonian drugs such as Cogentin, Kemadrin, Artane, and Symmetrel are frequently prescribed to suppress the movement disorders, such as parkinsonism, commonly caused by neuroleptics. About half of the patients who receive older antipsychotics also receive antiparkinsonian drugs.
Antiparkinsonian drugs are also described as "anticholinergic" because of their chemical action in suppressing cholinergic activity in the brain141. Rebound cholinergic hypersensitivity causes a large proportion of the withdrawal effects of antiparkinsonian drugs, which resemble those of tricyclics and neuroleptics and can be equally complex and varied. Several studies have demonstrated the typical flu-like symptoms of nausea, vomiting, chills, weakness, and headache, as well as the insomnia and restlessness that occur upon withdrawal from these drugs (Luchins et al., 1980 [259]).
In a rare double-blind, placebo-controlled withdrawal study the authors found "a recognizable withdrawal syndrome" - namely, an increase in anxiety manifested by "irritability; tension, palpitations and headache", various physical complaints (relating especially to the gastrointestinal tract), dizziness upon standing, and irregular heartbeats. In addition, psychotic symptoms flared up within days of withdrawal and lasted up to three weeks (McInnis and Petursson, 1985 [270], p. 297). Eight of the eleven withdrawn patients (73 percent) displayed these "obvious withdrawal-related symptoms".
Other studies, too, have observed psychotic flare-ups, which are usually accompanied by a resurgence of abnormal movements. Indeed, investigators directly relate these psychotic symptoms - including delusions, hallucinations, suicide attempts, and isolation - to physical symptoms of rigidity; restlessness, akathisia, and parkinsonism142.
In one such study, the authors state plainly that "the anticholinergic withdrawal syndrome resembles the negative schizophrenic syndrome in many important respects. Its occurrence in schizophrenic patients may be mistaken for psychotic decompensation and result in inappropriate treatment" (Tandon et al., 1989 [360], p. 712). A case report also describes a severe catatonia (immobility) upon abrupt withdrawal of Symmetrel (Brown et al., 1986 [79]).
Furthermore, one case of a severe reaction resembling neuroleptic malignant syndrome was reported upon withdrawal of antiparkinsonians, even though the patient had never received neuroleptics (Toru et al., 1981 [368]).
Since antiparkinsonian drugs worsen tardive dyskinesia and impair memory, withdrawal can provide extra benefits. In particular, withdrawal can improve the symptoms of tardive dyskinesia (Yassa, 1985 [396]) and can cause a "dramatic increase in memory scores on the Wechsler Memory Scale" (Baker et al., 1983 [24], p. 585).
As with all psychiatric drugs, withdrawal from antiparkinsonians should be gradual and tapered, thereby decreasing the recurrence of abnormal movements originally suppressed by the drugs. Gradual withdrawal is also less likely to lead to a reinstitution of the drugs (Ben Hadj et al., 1995 [36]).
As we saw in Chapter 7, the so-called gradual or tapered antidepressant withdrawals mentioned in the published literature are often too rapid. In one report, for example, "gradual" refers to a seven-to-ten-day withdrawal period (Barr et al., 1994 [30]); in another, it refers to withdrawal over four days (Rauch et al., 1996 [316]); in a third, it refers to a 50 percent reduction for five weeks, then no drug each alternate day for one week, then abrupt cessation (Amsden and Georgian, 1996 [9]). In two additional reports, "gradual" means a 50 percent reduction in dose for one week, then abrupt cessation143. In a review of forty-six case reports of SSRI withdrawal reactions, twenty-three involved abrupt withdrawal, twelve involved tapering over two weeks or less, and only six involved tapering over more than two weeks (Therrien and Markovitz, 1997 [366]). None of the last eighteen tapers used something like the gradual, patient-centered methods described in Chapter 8. Finally in a review of nearly sixty studies involving patients with diagnoses of schizophrenia or other psychoses who were withdrawn from neuroleptic drugs, the authors report that in most studies specifying the speed of withdrawal, the process was completed in less than a week, usually in one clay (Gilbert et al., 1995 [175]).
Rarely do investigators suggest that the taper period needs to be extended for several weeks; in fact, twelve weeks and fourteen weeks are the longest periods cited in the above-mentioned reviews. In one case of severe dizziness occurring upon successive, failed trials of Zoloft withdrawal in a twenty-nine-year-old man, the doctors contacted the drug manufacturer for help. They were advised to proceed with "extremely slow dosage titration" (Amsden and Georgian, 1996 [9], p. 686). Specifically a reduced dosage was prescribed every other day for three weeks, then every third day for the next six weeks, then every fourth day for two more weeks, at which time the Zoloft was finally discontinued. Overall, the withdrawal lasted eleven weeks.
No such recommendation to proceed with "extremely slow" withdrawal appears in the official product monographs for Zoloft or any other antidepressant. As noted, as a result of FDA requirements, makers of SSRI and other antidepressants have begun to include warnings about discontinuation symptoms and recommendations of "gradual reduction in the dose rather than abrupt cessation ... whenever possible" (as cited in the Cymbalta label, in the 2007 Physicians' Desk Reference [308]), but no guidelines are given, in contrast with guidelines for initiating drug taking. Thus, both doctors and patients should be actively encouraged to proceed more slowly and carefully when psychiatric drugs are withdrawn. The gradual methods described in Chapter 8 constitute a sensible guideline in this regard.
In an authoritative 500-page book focusing solely on adverse effects of psychiatric drugs, fewer than five pages are devoted to withdrawal effects associated with all categories of drugs (See Kane and Lieberman, 1992). This finding reflects the inadequate level of psychiatric interest and knowledge in such reactions.
Psychiatric drugs may induce a wide range of adverse effects when they are taken, and they may induce a wide range of adverse effects when they are withdrawn. Recognized withdrawal syndromes are a regular, common feature of the use of all psychiatric drugs. But as we have seen, doctors who prescribe these drugs too often fail to warn the patients who take them.
Because they produce unpleasant withdrawal reactions, psychiatric drugs must be considered drugs of dependence. In other words, at least some users will restart the drugs due to withdrawal-induced discomfort. These individuals will continue their drug use simply in order to avoid withdrawal reactions. The countless first-person reports posted on the Internet by users of SSRI antidepressants, for example, provide unusually vivid evidence that this phenomenon occurs all too frequently. Unfortunately, when withdrawal reactions lead to prolonged drug use, users risk experiencing more severe withdrawal reactions later.
As noted, the relatively small quantity of published case reports describing withdrawal reactions cannot be taken as a valid indicator of the actual frequency of their occurrence. As a rule, withdrawal reactions from most psychiatric drugs have been ignored or simply not recognized as such. However, studies specifically designed to look for such reactions have found them in 60-80 percent of patients. A full 20 percent may undergo "severe" reactions.
Doctors often focus on the physical consequences of withdrawal, such as nausea, tremors, or seizures, while failing to identify the emotional withdrawal symptoms that so often contribute to the resumption of drugs. Emotional withdrawal reactions such as anxiety, depression, insomnia, confusion, and irritability can actually have a greater impact on patients than purely physical symptoms.
The three main categories of emotional and behavioral withdrawal reactions are anxiety, depression, and psychosis. Anxiety reactions appear to be common upon withdrawal of central nervous system (CNS) depressants such as benzodiazepines and other tranquilizers, most of the antidepressants, antipsychotics, lithium and anticonvulsants used as mood stabilizers, and antiparkinsonians. Depressive reactions appear to be common upon withdrawal of stimulants and SSRIs. And psychotic reactions appear to be common upon withdrawal from neuroleptics, lithium, and antiparkinsonians. Still, evidence from case reports and studies suggests that any psychiatric drug may produce any of these withdrawal reactions.
In addition, over a dozen studies so far have implicated expectant mothers' use of antidepressants during pregnancy with the appearance of a peculiar "neonatal abstinence syndrome". In one well-controlled study a full 30 percent of 60 infants whose mothers took antidepressants for prolonged periods, including during the third trimester, developed the syndrome, which lasted up to four days; 13 percent of the infants had severe reactions. The most common symptoms were tremor, gastrointestinal problems, an abnormal increase in muscle tone (hypertonicity), sleep disturbances and high-pitched cries. None of the 60 infants without exposure to SSRIs developed the syndrome (Levinson-Castiel et al., 2006 [252]). There is some debate whether this represents an actual withdrawal reaction or a sign of direct drug toxicity in the serotonin system.
Furthermore, clinicians and researchers do not always correctly describe the symptoms that occur upon withdrawal. In many cases, prodrug bias and habitual resistance to acknowledging withdrawal effects lead them to use relatively neutral, ambiguous terms (such as agitation, restlessness, anxiety, and psychomotor retardation) in place of terms like psychosis.
Undoubtedly; many withdrawal reactions from all categories of drugs are mistakenly treated as "relapses" by prescribing physicians, thus moving the focus away from possible drug involvement and toward the patient's "underlying disorder". Many researchers also approach the problem timidly often failing to describe or emphasize withdrawal reactions. As a consequence, patients suffer, and the public is kept uninformed.
So that patients can provide truly informed consent for the drugs they are taking, doctors must fully describe withdrawal or discontinuation in terms of adverse drug reactions with a "frequent" probability of occurrence, it is the doctors duty to make sure that patients actually grasp and then remember the nature and likelihood of all important adverse reactions, including withdrawal reactions. Similarly, in their official drug monographs, drug manufacturers need to provide accurate summaries of reports of withdrawal reactions, as well as detailed guidelines for tapering their products, just as they provide guidelines for how to initiate treatment.
The best way to minimize the risk of severe withdrawal reactions is not to take psychiatric drugs in the first place. The second best approach is to plan a slow, gradual withdrawal involving close monitoring and a systematic, ongoing program of information, counseling, and reassurance. Unfortunately, however, abrupt withdrawal remains too common in clinical practice. Abrupt withdrawal is imprudent and may result in additional distress and disability. Except in emergencies, patients who are stopping their use of psychiatric drugs and the professionals who are assisting them should proceed gradually and maintain this gradual pace until complete cessation is accomplished, even if the early stages of withdrawal present no difficulties.