In June 1982, the US FDA approved isotretinoin (Accutane®Footnote 1 or 13-cis-retinoic acid) for the treatment of severe, recalcitrant cystic acne unresponsive to conventional treatments. Isotretinoin was also approved in Canada in 1983 for use in this indication. Hoffmann-La Roche, the drug manufacturer, estimates that by April 2001 approximately 12 million patients worldwide had used isotretinoin, with 5 million of those patients in the US.[1] There is a clear trend in the US (and probably worldwide) for isotretinoin to be used for milder forms of acne with increasing numbers of patients being treated and exposed to the drug.[2] In the US, UK, Canada, Australia, New Zealand, the Netherlands, Mexico and Brazil the dosage recommended by the manufacturer is 0.5–1.0 mg/kg. In Canada, the US, Mexico, Brazil, Netherlands, Australia and New Zealand a dosage of 2 mg/kg is suggested for severe acne not responding to the usual dose.

From the outset, the major concern was the drug’ proven teratogenicity and programs to prevent isotretinoin-induced birth defects were instituted. The most recent addition to this program has been the System to Manage Accutane Related Teratogenicity (SMART), which was introduced in the US on October 31, 2001.[3]

While there were no reports of depression or mood changes in the New Drug Application (NDA) safety database, post-marketing reports of depression in the mid 1980s resulted in labeling changes listing depression as a possible adverse reaction.[4] In 1996, two cases of suicide associated with isotretinoin triggered a more indepth re-evaluation by the FDA.[4] In February 1998 changes were made to the product labeling. The Warnings section reads “Psychiatric Disorders: Accutane may cause depression, psychosis, and rarely, suicidal ideation, suicide attempts and suicide. Discontinuation of isotretinoin therapy may be insufficient; further evaluation may be necessary.” Similarly “psychiatric, suicidal ideation, suicide attempts, suicidal depression, psychosis, emotional instability” are listed in the Adverse Reactions section.[4]

In September 2000, the FDA’s Dermatology and Ophthalmology Drugs Advisory Committee discussed both the pregnancy prevention issue and a risk management strategy for the uncertain risk of psychiatric effects associated with the use of isotretinoin.[5] This committee, among other recommendations, suggested that the manufacturer should add the information about psychiatric adverse events to the informed consent document signed by patients and/or their parents or guardians prior to receiving isotretinoin.

In 2002 “aggressive and/or violent behaviors” was added to the Psychiatric Disorders of both the Warnings and Adverse Reactions sections in the US product monograph.[6]

These changes have focused the attention of the press on the safety of isotretinoin. Isotretinoin-induced depression and possible suicide has been reviewed in the dermatological literature[7,8] and has been featured in investigative television reports, talk shows and on the web.[9,10] The extensive media coverage, as well as the implicit warnings contained in the new consent forms and drug information sheets, has many patients and their parents fearful of the consequences of isotretinoin treatment.

This review will examine the complex relationship between isotretinoin and psychiatric symptoms. The MEDLINE database (January 1966–May 14, 2003) was searched including the following key words: depression, mood disorder, suicide, depressive disorders and isotretinoin, Accutane®, Roaccutane®, retinoids, etretinate, acitretin, vitamin A, and drugs. In addition, an internet search of the FDA database files and the Reports of the Committee on Government House Reform (US House of Representatives) for isotretinoin and depression was conducted.

1. Establishing a Causal Connection Between Isotretinoin and Depression and Suicide

In establishing a causal connection between isotretinoin use and depression and suicide it is useful to consider whether or not an association in fact exists, what is the nature of that association, and if there is a plausible mechanism of action (see Rothman and Greenland for a discussion of causality in epidemiology).[11]

Assuming equivalence in the reporting of depressive symptoms among patients receiving isotretinoin, the incidence of depression and suicide among patients receiving isotretinoin should be greater than that reported in the general population. Depression should be measured using established standard rating scales or diagnostic schedules with known psychometric properties so that the criteria used to establish depression are clear and reliable. It is important to clearly establish that the risk of depression is greater in isotretinoin users as opposed to non-users or that there are unique subsets of isotretinoin users at increased risk of depression. The association should be consistently evident in a variety of study designs whether case-control, cohort or randomized clinical trials. Case reports, while useful for hypothesis generation, are not of value in establishing a causal connection. A plausible biological mechanism of action should also be articulated that links isotretinoin to depression. Signals suggesting a relationship would include a temporal association (taking the pharmacokinetics of the drug into account), dechallenge (the symptoms clear when the drug is withdrawn), rechallenge, a dose-response relationship (whether that be monotonic, J-shaped or threshold), and a drug class effect. Finally, there should be no strong competing explanations for the observed associations.

2. Measuring Depression

In any discussion on depression it becomes relevant to distinguish between depressive symptoms and depression the disease, as defined in the Diagnostic and Statistical Manual of Mental Disorders Fourth Edition (DSM-IV) of the American Psychiatric Association.[12] The word depression has several meanings, depending on the context. The simplest interpretation relates to mood. Synonyms would include feeling down, having the blues, feeling sad, unhappy, downhearted, and having low spirits. Most of us will have temporally experienced such moods. Depression in mood disorders takes on the added, more distressing meaning expressed by such synonyms as tearful, despair, hopelessness and, in general, an inability to find pleasure in those things that are normally pleasurable. When used to define a disease, a diagnosis or syndrome, criteria of severity, duration and clinically significant distress or impairment are required.

The DSM-IV lists a number of major categories of mood disorders with further sub-categories which include major depressive disorder (single episode, recurrent), dysthymic disorder, bipolar I disorder, cyclothymic disorder, bipolar II disorder, mood disorder due to a general medical condition, substance-induced mood disorder, and adjustment disorder with depressed mood. In order to be diagnosed with having a mood disorder, it is not sufficient that depressive symptoms are merely present, but rather, depressive symptoms have to persist for a specific length of time and impact upon a person’ social, occupational and interpersonal functioning. To be diagnosed with major depressive disorder, single episode, one must have had at least one major depressive episode.[12] Table I lists the criteria for major depressive episode noting the symptoms for the episode, length of time (2 weeks) over which they are required to persist, functional impairment impact, and the ruling out of substance abuse or physical disorders as competing explanations for the symptoms. Major depressive episodes may also occur in bipolar disorders. There are also explicit criteria for manic episodes and mixed episodes.

Table I
figure Tab1

Criteria for major depressive episode according to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV)[12] [reproduced from the DSM-IV. Copyright 1994 American Psychiatric Association, with permission]

Full size image

Determining whether or not an individual has a depressive disorder requires a clinical judgment by a trained clinician (i.e. a psychiatrist or clinical psychologist). In epidemiological research, the determination of depression is usually made with the assistance of a screening or rating scale or a structured diagnostic schedule. These types of instruments have been developed for use by the individuals themselves, trained interviewers, or clinicians (either psychiatrists or clinical psychologists). These instruments have been validated, are reliable and have known psychometric properties.

Examples of self-report screening scales are the Beck Depression Inventory,[14] the Center for Epidemiological Studies Depression Scale,[15] and the General Health Questionnaire.[16] The Hamilton Depression Rating Scale is an example of a rating scale designed for use by clinicians.[17] The Diagnostic Interview Schedule[18] and its successor, the Composite International Diagnostic Interview (CIDI),[19] are structured diagnostic schedules that can be used by trained interviewers. The Structured Clinical Interview for DSM-IV[20] and the Present State Examination[21] are structured diagnostic schedules for use by clinicians. The virtues of self-report screening scales are that they are brief, easily administered and economical. Structured diagnostic schedules are more valid but much more expensive to use. Measuring depression for research purposes, at the minimum, requires the use of a scale or diagnostic schedule with known psychometric properties.

Similar measurement issues apply to the determination of suicide.

3. Depression and Suicide in Adolescents and Young Adults

The population affected by acne includes mostly adolescents and young adults. This can be a turbulent stage in the life cycle and the symptoms of depression may be masked by aggression or disruptive behavior. Denial and concealment of depressive symptoms is common, and underreporting of depression is significant. In addition, alcohol, substance abuse, smoking, poor school performance, relationship problems including breakups, lack of social support, chronic conditions and previous depressive episodes are important co-morbid factors. Depressive episodes can be cyclical.

The National Population Health Survey (NPHS) is a continuing longitudinal study omnibus health conducted by Statistics Canada.[22] The NPHS incorporates a short-form adaptation of the CIDI major depression section assessing the nine diagnostic symptoms for major depression (as outlined in the DSM-IV). In the survey, a diagnosis of major depression required a total of five depressive symptoms including depressed mood or loss of interest. The information was gathered from the Canadian general population, and anyone institutionalized was not included in the results. The interviews were conducted in 1994/1995 and participants were re-interviewed in 1996/1997. An analysis of new cases of major depression arising in this large population sample showed that the annual incidence of depression for 12–24-year-olds living in households in Canada for 1996/1997 to be 7.1% for females and 2.9% for males.[23] The incidence for major depression in this study was higher amongst participants in younger age groups and within all age groups was generally higher in women than in men. Women between the ages of 15 and 19 were the most likely to report symptoms of depression. The duration of depression was also likely to be longer for women than men.

Mortality-cause statistics for Canada show that suicide is the second leading cause of death in both sexes in the 15–34 age group, representing 33% of all suicides.[24] The incidence rate for suicide was estimated at 12.3 per 100 000 in 1997 and 79% of completed suicides were by men.[24] These age/gender patterns are consistent with the epidemiological literature.

A similar situation prevails in the US. Major depressive disorder affects about 5% of the general population;[25] 3.3% of men and 6.6% of women.[26] The average age of onset is in the mid-20s.[26] The suicide rate in the 15–24-year-old age group nearly tripled between 1952 and 1996, and in 1997 exceeded 13.8 per 100 000. It is the third leading cause of death in the US among 15–24-year-olds.[27]

4. Psychiatric Aspects of Acne

If the impact of drug-induced depression is to be assessed, it is important to examine the incidence and prevalence of depression in patients with acne. Acne has a profound psychosocial legacy.[28,29] Acne patients experience problems with self-esteem,[30] self-confidence, body image, embarrassment, social withdrawal, depression, anger, anxiety, preoccupation with acne, and impaired social functioning including poorer academic functioning, unemployment,[29] decreased dating and decreased participation in sport.[28,31,32] The cosmetic impact may be disproportionate to the observed severity of the acne. It has been suggested that adolescents and young adults are more vulnerable to the cosmetic problems imposed by skin diseases and that acne might act as a precipitating factor for depression in this population.[33] These authors found that even mild-to-moderate acne can be associated with significant depression and suicidal ideation. Cotterill and Cunliffe[34] described seven acne patients who committed suicide. Two of these patients had been noted to be depressed prior to their suicides. Another patient committed suicide during the fourth month of isotretinoin treatment. Although one other patient in this group is said to have received isotretinoin, no further details are given.

Despite suggestive literature, there have been no solid studies that clearly document the effect of acne on depression.

4.1 Dysmorphophobia and Acne

Dysmorphophobia refers to an intense dislike of a part of an individual’s own body, which is both irrational and inappropriate. These patients often have minimal objective evidence of acne and are unhappy with the results of treatment. They are regularly anxious and depressed. The mental state of a group of such patients is said to have improved on isotretinoin treatment.[35]

5. Drug-Induced Depression

Drug-induced depression is classified in the DSM-IV as a substance-induced mood disorder. The diagnostic features include a prominent and persistent disturbance in mood, judged to be the physiological effect of the substance (drug).[12] The disturbance may involve depressed, expansive or irritable mood, and may develop while a patient is taking the substance or following the withdrawal of the substance. Other criteria include the exclusion of other mood disorders unrelated to the drug, and the symptoms must cause clinically significant distress or impairment in social, occupational or other important areas of functioning.[36]

Whitlock and Evans[37] reported that more than 200 drugs have been said to induce depression, with 80 drugs having at least four reports in the literature. Drugs implicated in causing depression include anabolic steroids, β-adrenoceptor antagonists (β-blockers), cimetidine, corticosteroids, fenfluramine, interferon-α, nifedipine, oral contraceptives, ranitidine and verapamil.[38] Concern has been expressed that some cholesterol-lowering drugs (pravastatin and simvastatin) may increase suicide rates.[39] However, the accepted causal association between drugs and depression is often based only on evidence from case reports, and rigorous studies are often absent.[38] Even when epidemiological studies suggest a relationship, other studies may reach the opposite conclusion. Other candidates such as underlying medical illness, psychosocial stress, major depressive disease, and invalid patient attribution further complicate this enigmatic relationship. The reality is that it is very difficult to be certain that a drug is not responsible for a patient’ symptoms of depression.[40,41]

6. Depression and Suicide Linked to Isotretinoin

6.1 Literature Reports

A MEDLINE search (January 1966 to May 2003) found 24 cases linking isotretinoin to depression and suicide (see table II). Hazen et al.,[42] in 1983, reported that 6 out of 110 patients with acne or keratinizing disorders experienced depressive symptoms (crying spells, malaise or forgetfulness were the only symptoms mentioned) while receiving isotretinoin. One patient had a history of depression. Five patients continued taking isotretinoin despite the symptoms, which resolved rapidly once the drug was stopped. The dosages administered were between 1 and 2 mg/kg/day.

Table II
figure Tab2

Literature reports linking depression/suicide to isotretinoin use

Full size image

Scheinman et al.[43] reported that 7 out of 700 patients with cystic acne, cutaneous disorders of keratinization, psoriasis or basal cell carcinoma fulfilled diagnostic criteria for the diagnosis of major depressive disorder. Symptoms of depression resolved within 1 week of stopping isotretinoin. Two of the seven patients had received previous isotretinoin treatment without episodes of depression. The mean duration of treatment prior to the onset of depression was 14 weeks (range = 6–47 weeks). One patient was rechallenged and the symptoms recurred during the third month of treatment.

Mood changes (depression) occurred in two teenagers reported by Duke and Guenther.[44] One, a 15-year-old girl, threatened suicide and set fire to her clothes in protest to the stopping of the isotretinoin. This patient recovered after 6 months following the withdrawal of isotretinoin.

Byrne and Hnatko[45] described three patients with depression while taking isotretinoin, two of whom had been actively suicidal. For one patient, depressive symptoms were still present 5 months after having stopped isotretinoin. All three patients had high scores on the Hamilton Depression Rating Scale. One woman made a remarkable recovery once the isotretinoin was stopped.

Bravard et al.[46] reported three patients (two males and one female) who developed depression while receiving isotretinoin. The two men attempted suicide; one after 4 months of treatment and the other, successfully, 3 months after completing 4 months of treatment.

Gatti and Serri[47] described a 17-year-old male whose acne had shown considerable improvement on isotretinoin but who developed a major psychiatric illness 1 month after completing treatment. He committed suicide despite receiving psychiatric care.

Burket and Storrs[48] reported a 2-year-old child with nodulocystic acne who, while receiving isotretinoin, changed from a cheerful, happy child to being sullen, withdrawn, cranky and crying easily. When isotretinoin was discontinued she returned to her former cheerfulness.

Cott and Wisner[53] reported that a woman with bipolar disorder was monitored weekly using several psychiatric measures, including the Hamilton Depression Rating Scale, throughout a course of isotretinoin treatment given for severe acne. After a month, she became suspicious of and preoccupied with a casual comment made by a relative and was unable to sleep. These symptoms were interpreted as early signs of decompensation and she was started on perphenazine and divalproex. She completed a 6-month course of isotretinoin without further behavioral symptoms.

Cotterill and Cunliffe[34] described a patient, who had responded poorly to isotretinoin, committing suicide during the fourth month of treatment.

Three patients receiving isotretinoin for severe acne attempted suicide by taking an overdose of the medication.[49–51] In none of these cases were references made to preceding depression or to a psychiatric assessment.

A report from Australia described a 17-year-old man with moderately severe acne who developed acute depression and suicidal ideation after 2 weeks of isotretinoin at a dosage of 0.63 mg/kg/day. The depression improved when the dose was lowered and sertraline was added. The depression deteriorated when the dosage of isotretinoin was increased to 1 mg/kg/day and he attempted suicide. Within 2 weeks of discontinuing isotretinoin he had improved and was no longer suicidal. After 6 weeks he had no depressive symptoms and had returned to work.[52]

In a study of 94 US patients who were taking isotretinoin, 10% reported insomnia and mild depression.[54] During a period between October 1982 and June 1985, the Adverse Drug Reaction Reporting System (ADRRS) of the American Academy of Dermatology reported 104 suspected reactions occurring in 93 patients taking isotretinoin.[55] This report identified three patients with depression while taking isotretinoin. One patient who was taking concomitant nortriptyline experienced a worsening of his depression while a second patient taking lithium for manic-depressive psychosis became suicidal after starting isotretinoin. Details of the third patient were not provided.

In a prospective study examining adverse effects in acne patients treated with isotretinoin at a dosage of 1 mg/kg, symptoms of depression were noted in 4% of patients.[56] Depression persisted in most patients for the duration of the treatment. None of the patients required treatment and all completed the course of isotretinoin.

6.2 Roche Drug Safety Database

By April 2001, Hoffmann-La Roche estimated that approximately 12 million patients worldwide have used isotretinoin, while in the US about 5 million patients have received the drug.[1] In presentations to the FDA, Hoffmann-La Roche[57] reviewed all cases of psychiatric adverse events received worldwide by Roche Global Drug Safety between 1982 and April 30, 1999. The cases were classified with the preferred term in WHO-ART SOC 500 Psychiatric Disorders, and then placed into the Roche Drug Safety Database. This classification used the verbatim wording of the report. Although the word depression might have been used in the original report to convey a symptom, it would be classified in the same way as major depressive disorder. There were 1247 cases of mood disorders; 367 were cases with positive dechallenge. Twenty-three of the 367 cases indicated both positive dechallenge and positive rechallenge. There were a total of 168 cases of suicidal behavior worldwide, which included 104 suicide attempts and 64 (53 males and 11 females) completed suicides; 38 completed suicides were in the US.[58] These reported suicides occurred both during treatment and up to 10 years following treatment. Of the 64 individuals committing suicide, 30 were on treatment, 24 were off treatment and in 10 cases information was lacking.[58,59] No relationship between isotretinoin dose and suicide was found.[58] In the presentation to the FDA, Hoffmann-La Roche concluded that the number of suicides observed in the US is much less than would have been predicted from suicide statistics for an age- and sex-matched general population group.[57]

Clearly the reports to the manufacturer are voluntary and, almost certainly, seriously underestimate the frequency of the adverse events, but the denominator is also extremely large. A more critical examination of this data is required and should form the basis of an easily accessible publication.

6.3 Governmental Adverse Drug Monitoring Agencies’ Reports

6.3.1 US FDA Reported Cases

The FDA, like most national adverse drug reaction monitoring bodies, requires that drug manufacturers expedite the reporting of serious adverse events and periodically report on other events. The adverse event reporting system (AERS) for drug products is a computerized database based on Medwatch forms (Adverse Drug Event Reports). These forms are submitted on a voluntary basis by physicians or pharmacists either directly to the FDA or more usually to the drug manufacturer, and it is thought that there is substantial underreporting.[60] However, according to the FDA, the information collected does allow the detection of rare adverse events.[60]

Wysowski et al.[61,62] reported that in the 18-year period that isotretinoin has been available in the US (i.e. from June 1982 to May 2000), the FDA had received reports of depression, suicidal ideation, suicide attempts and completed suicide in 431 exposed patients. Thirty-seven US patients committed suicide; 24 while taking isotretinoin and 13 after stopping the drug. Males constituted 87% of the group. The median age was 17 years. For individuals committing suicide whilst taking isotretinoin, the median time from initiation of isotretinoin to suicide was approximately 3 months. For individuals who committed suicide after stopping the drug, the median time to suicide, after isotretinoin cessation, was 2.5 months. There was a past history of psychiatric disorders in 22% of cases. In 57% of the suicide cases, co-morbid factors were identified including substance abuse, stressful life events, personal relationship problems and a family history of psychiatric disorder. The median peak dosage was 1 mg/kg/day. About half of these reports were received after labeling changes warning of suicide and depressions.

In addition to suicide reports, the FDA received reports of 110 US isotretinoin users who were hospitalized for depression, suicidal ideation and suicide attempts; 85 of whom were taking isotretinoin and 25 of whom had stopped taking the drug.[63] The median age was 17 years and 56% were female. A history of a preexisting psychiatric disorder was reported in 44% of patients, while other possible contributing factors were reported in 52% of patients.

Wysowski et al.[62] reported that a further 284 US isotretinoin users had depression that did not require hospitalization. This group included 24 patients where depression recurred when rechallenged. Forty-five percent of these reports were received in 1998 after depression and suicide were added as a warning to the labeling, and the fact that half of these reports were from isotretinoin users or their relatives is of significance. The warnings associating isotretinoin and depression may have stimulated additional reports.

The FDA performed a rank order for depression of all drugs in the spontaneous reporting and adverse event-reporting database. Isotretinoin ranks number 4 for depression, 5 for serious depression, and 10 for suicide attempts.[62]

Assuming a 4-month exposure for isotretinoin use, the 5 million exposed isotretinoin users in the US would have the equivalent of 1.6 million years of exposure. Based on the expected suicide rate of 11.4 per 100 000 population per year for persons between the ages of 15 and 24 years, the predicted number of suicides would be 190, which is substantially more than the 37 cases reported.[62] Using an incidence rate of 1.6 per 100 per year for the onset of major depression,[64] then 26 666 cases of depression could be expected to have arisen among isotretinoin users in the US.

FDA Reported Cases of Dechallenge/Rechallenge

Marilyn R. Pitts,[60] a safety evaluator in the FDA’s department of Post Marketing Drug Risk Assessment, gave details of 41 depressed patients with positive dechallenge/rechallenge of isotretinoin. In 10 patients there was a reported past psychiatric history. The series included 28 patients with depression, five patients with psychosis or psychotic depression, five patients with mood disorder and three patients who had suicidal ideation. There were almost twice the number of females as males. Six patients required hospitalization. The median time to onset was 30 days, but the range was as short as 1 day and as long as a year. The median time to onset in the second course (rechallenge) was only 10 days with a range of 1–60 days. Significantly, symptoms persisted in 10 of the 33 patients whose follow-up was reported.

While positive dechallenges and rechallenges are important in signaling a possible association between the adverse event and the drug, the data is complicated by the inclusion of patients with previous psychiatric illnesses and cases where the late onset makes an association seem unreasonable. Despite this, there are clearly cases suggesting an association of severe depression with isotretinoin.

6.3.2 Cases in Canada, Australia and Ireland

Isotretinoin was approved in Canada in 1983. From 1983 to January 1999 the Canadian Adverse Drug Reaction Monitoring Program received 16 reports (8 since March 1998 after depression and suicide were added as a warning to the labeling) of depression and other psychiatric problems.[65] These included depression (5), aggressive reaction (4), emotional lability (4), irritability (3), suicidal tendency (3), amnesia (2), aggravated depression (1), abnormal thoughts (1), manic reaction (1) and suicide attempt (1). In those reports where the dose was documented, all but one patient was on the maximum recommended dosage of 1–2 mg/kg; the remaining patient had received dosages up to 3 mg/kg. Wooltorton[66] reported that by November 22, 2002, Health Canada had received 56 adverse event reports associating depression and suicidal ideation with use of isotretinoin.

The Australian Adverse Drug Reactions Advisory Committee has received a total of 12 reports of depression in association with isotretinoin from 1985 to June 1998.[67] All the patients involved were young (age range was 15–40 years, median 19 years), mainly male (M : F = 9 : 3) and were taking the drug for the treatment of acne. Two reports described the re-emergence of depression and in the other 10 reports depression was noted for the first time. Two cases were described as severe and four patients had psychotic features. Three patients developed suicidal thoughts and two patients attempted suicide. One of these had a fatal outcome. Of the other 11 patients, three had recovered after withdrawal of isotretinoin, one was improving with the use of an antidepressant, and the other seven patients had not recovered at the time the report was submitted.

The Irish Medicines Board has received a total of six reports of psychiatric events including one case of suicide occurring during treatment with isotretinoin.[68]

The value of case reports suggesting a possible linkage between isotretinoin and depression is strengthened when the symptoms of depression develop during treatment, subside rapidly after the drug is discontinued, and occur in patients without a history of previous psychiatric problems or other confounding issues. Governmental adverse drug-monitoring agencies regularly fail to present their data using meaningful criteria for depression. In addition, patients with pre-existing depression should be identified and analyzed separately. Chronological limits should be set. The majority of substantive mood disorder cases should resolve within 30 days. Most of those reported in the literature did so within 15 days,[69] yet in many of the reported cases, patients continued to have symptoms months and years after isotretinoin exposure.

6.4 Epidemiological Studies

A large case-control cohort study, sponsored by Hoffmann-La Roche, found no increased risk of depression in individuals treated with isotretinoin when compared with those treated with antibiotics.[70] This study used two validated databases, one Canadian, the other British, to compare the prevalence rates of neurotic and psychotic disorders, suicide and attempted suicide between isotretinoin and antibiotic users. In the Saskatchewan Health data files there were 7195 isotretinoin and 13 700 oral antibiotic users, while the United Kingdom General Practice Database contained information on 340 isotretinoin and 676 antibiotic users. The relative risk estimate for depression or psychosis was 1.1 (95% CI 0.8–1.6). The relative risk estimate for suicide and attempted suicide was 0.9 (95% CI 0.3–2.4). The small number of suicides in the populations studied makes an estimate of relative risk difficult. The study asserts that the use of isotretinoin is not associated with an increased risk of depression, suicide or other psychiatric disorders. This study represents the first attempt to formally evaluate an association between isotretinoin use and depression and suicide. However, there are a number of weaknesses in the study design. The diagnostic codes chosen in the Canadian study were the International Classification of Diseases (ICD)-9 codes 296–301. While the codes for major depressive disorder were included, other important codes for depression were omitted, including 309 — adjustment reactions and 311 — depressive disorder, not classified elsewhere. These patients would not have been ascertained. Much of the study was not about depression. Other weaknesses include the failure to use drug data to maximize the identification of the study disease and lack of an attempt at validation.[71] Wysowski[72] has also commented on the limitations of this study, the summary of which must be considered to be inconclusive.

A second formal study by Hoffmann-La Roche has appeared in abstract form.[73] This was a retrospective study using a major, nationwide US Health plan database and included 5130 first time isotretinoin users and 25 600 age-, gender- and health plan unit-matched controls. A 6-month screening period was included to identify pre-existing depression and the follow-up period extended to 30 days after the last dose of isotretinoin. The results presented in this abstract do not support an increased incidence of depression in the isotretinoin group. A more detailed analysis of this study awaits the publication of its details.

7. Depression Linked to Vitamin A and Other Retinoids: Is There a Drug Class Effect?

A drug class effect might be expected if isotretinoin produced depression by a pharmacological effect on cerebral retinoid receptors. Vitamin A is an essential nutrient that is fat-soluble and mostly stored in the liver. Widespread consumption of vitamin and mineral supplements has raised concerns about hypervitaminosis A. However, symptoms of toxicity generally require a daily dosage in excess of 100 000 IU. Hypervitaminosis A results in many of the well known adverse effects of isotretinoin, including raised blood lipids, dry skin, cracked lips, fatigue, hair loss and teratogenicity. Hypervitaminosis A may also be associated with CNS effects such as sleeping difficulties, loss of appetite, tiredness, irritability, ataxia and increased intracranial pressure. Bauernfeind did not list depression as a symptom in 200 cases of hypervitaminosis (as reported by Bendich and Langseth).[74] Restak[75] described an 18-year-old female patient with acne who developed severe psychosis and depression following 6 months of self medication on high dose (100 000–150 000 IU) retinol (vitamin A). She had no antecedent psychiatric problems. This episode preceded the development of pseudotumor cerebri. Both resolved rapidly when retinol was stopped. A 54-year-old with no prior psychiatric history developed depression following ingestion of 12 times the normal dietary supplement of retinol 25 000 IU/day for 2 years.[76] This patient recovered within 2 months of stopping the retinol.

When high dose oral tretinoin (vitamin A acid) [100–200 mg/day] was used in early clinical studies in 30 patients, 3 patients developed changes in their psychological state.[77] Henderson et al.[78] described three cases of depression associated with etretinate treatment. These patients all had psoriasis, with no past history of depression. In all three patients, depression resolved when etretinate was withdrawn; in one patient depression recurred on rechallenge. A 49-year-old man with symmetrical erythrokeratoderma was treated with etretinate 1 mg/kg/day. After 1 month on treatment, he developed profound depression, fatigue, aggression, somnolence and reduced work capacity. These symptoms resolved after stopping etretinate and did not recur when rechallenged at a dosage of 0.5 mg/kg/day.[79]

7.1 Mechanism of Action

Isotretinoin is a retinoid acid receptor agonist. Retinoid receptors are found in the brain. The retinoic acid receptors (RAR)-α, and retinoid X receptors (RXR)-α and RXR-β are uniformly distributed, whereas RAR-β, RAR-χ and RXR- χ are restricted in their expression patterns. If there is a causative association, perhaps dose-dependent, between retinoids (including vitamin A) and psychiatric symptoms, the relationship between retinoid receptor function and psychiatric effects has not yet been established.

8. Improvement of Psychiatric Symptoms with Isotretinoin

Isotretinoin treatment has also been shown to significantly improve anxiety and depressive symptoms in acne patients.[35,80,81] Rubinow et al.[80] evaluated the psychiatric morbidity and mood characteristics of a group of 72 patients with cystic acne. They used a number of self-reported instruments to assess psychiatric symptoms both before and after treatment. In addition, 55 patients took part in a psychiatric interview before treatment and 34 were re-interviewed after treatment. Of the 55 interviewed before treatment, four patients met the criteria of the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders Third Edition (DSM-III) for major affective disorder, one had generalized anxiety disorder, and one had major depression. The latter patient was not included in the analysis. After treatment, these patients had significant reductions in anxiety, depression and interpersonal sensitivity. This improvement was most marked in those showing the greater clinical improvement in their acne.

Kellett and Gawkrodger[82] noted that 6 (18%) of 34 patients were significantly depressed prior to isotretinoin treatment, as assessed using the Hospital Anxiety and Depression scale. Following treatment, none of the patients had depression scores indicating that they required a further psychiatric assessment. Macdonald-Hull et al.[35] reported that 16 patients with acne who had inappropriate depression of mood, some with dysmorphophobia, were treated with isotretinoin. Of the 16 patients, 14 felt they had benefited from the treatment. Only one of these patients accepted psychiatric help. Psychiatric measures of mood disturbance were not used to monitor the treatment.

Layton et al.[83] used a validated questionnaire to assess the psychosocial disability produced by acne both before and after isotretinoin treatment. The Assessments of the Psychological and Social Effects of Acne (APSEA) questionnaire contained elements derived from several instruments used to measure psychosocial disability including the Hospital Anxiety and Depression scale and the Zung Self-Rating Depression Scale. These researchers demonstrated significant differences between patients with acne and age- and sex-matched controls. When patients who had received isotretinoin were assessed 6 months after the completion of treatment, there was a dramatic improvement in scores of psychological disability. These authors advocated early use of isotretinoin to reduce the psychosocial disability in patients with acne.

While these studies demonstrate that mood improves as acne improves, the studies do not conclusively prove that there is a cause and effect relationship between improvement in acne and improvement in mood; although they provide strong evidence of such a connection. Ideally studies would require an untreated or placebo group to definitively determine such a relationship.

9. Conclusion

There is evidence from case reports that isotretinoin (and other retinoids such as vitamin A and etretinate) may be associated with the development of depression. However, there is also evidence that isotretinoin may improve psychiatric symptoms in patients with acne. The results of the single-published, formal, large scale, case-control cohort study on the adverse effects of isotretinoin use are inconclusive. Epidemiological evidence for an association between isotretinoin and depression is lacking. Epidemiological studies using well designed, appropriate study methodologies, including reliable and valid measures, and adequate sample sizes, may bring us closer to the answer. Perhaps there are subsets of patients who are at increased risk. Such studies should not exclude patients who might be depressed, as some of these patients may also benefit from treatment. Further sound research is needed.

In the meantime, for the practitioner, the obvious benefit of isotretinoin in treating acne should determine its continued use; however, patients and their relatives must be informed and be encouraged to report depressive symptoms promptly. At each visit the physician should inquire specifically about symptoms of depression. Such questions should include an inquiry into mood changes, persistent feelings of sadness or depression, loss of interest in normal activities, loss of appetite, loss of sleep and excessive tiredness. Then if necessary, antidepressant treatment, referral to a psychiatrist or discontinuation of isotretinoin should be considered.