Presentation of benefits and harms of antidepressants on websites: A cross-sectional study

2.1.Selection of websites

We tried to mimic what most people would do when searching for information on the Internet and therefore chose easily accessible websites, which appeared at the top of a Google search. We searched the term depression and when this did not provide information about pharmacological therapy, we searched the terms antidepressants or depression therapy. Due to language restrictions, we chose ten countries where English or a Nordic language is spoken: Australia, Canada, Denmark, Ireland, New Zealand, Norway, South Africa, Sweden, UK, and USA. We avoided websites that were clearly aimed only at health practitioners or that required logins. Each webpage was archived and dated; the time frame was April to August 2018. We reviewed at least two websites per country (see list of the websites in Supplement S1). Doubts about inclusion were resolved by discussion between the authors.

2.2.Data extraction

One researcher focused on the websites in English, the other on those in Nordic languages. Doubts about data extraction and their interpretation were resolved by discussion. We classified the websites into three groups, primarily based on the information provided in the “About us” section, according to whether they were governmental institutions, advocacy groups, or consumer organisations. We determined that the purpose of advocacy groups was to promote the interests of patients and families, whereas consumer organisations generally provided information about healthcare services that are offered to patients and citizens (Table 1).

When information about funding was unclear, we emailed or phoned the organisations. Judgments about whether the websites met the standards for informed consent were based on the guidelines by the Medical Board of Australia [9], Australia’s National Health and Medical Research Council, (NHMRC) [10] and the Canadian Medical Protective Association’s Good Practices Guide [11] (see Table 2).

Based on a list of 14 predefined criteria, we assessed whether the information presented was trustworthy, given the available evidence (see Table 3) [12–40]. These criteria were either of obvious importance considering the research literature on antidepressants, or represented controversial issues, e.g. whether the drugs are safe during pregnancy, which is also important to know for the users.

2.3.Patient and public involvement

We did not involve patients or the public in our research project. One of us (PCG) communicates very often with patients, relatives and the public about psychiatric drugs and is Protector for the Hearing Voices Network in Denmark.

3.1.Effectiveness of antidepressants

All 39 websites mentioned the uses and benefits of antidepressants. For example, an Australian website claimed, “most people (60–70%) will recover in 6 to 8 weeks of taking antidepressants” [w30]. Another Australian website stated that, “around 70% of people with major depression start to feel better with the first type of antidepressant they are prescribed” [w29].

Others took a more measured approach stating that, “a significant percentage of people may not respond to a prescribed antidepressant” [w20] or that, “For some people with mild depression, antidepressants seem to have little effect. However, for people with more severe depression, antidepressants often make a big difference” [w21].

3.2.Relapse of depression and duration of treatment

Twenty-two websites (56%) claimed that antidepressants could prevent relapse, advising that medications should be taken typically for six to nine months, and some websites even suggested for a lifetime. For example, a Swedish website stated that, “after you are healthy from your depression, you need to continue your treatment for a period of time. In some cases, antidepressants must be taken life-long” [w38]. A Norwegian website stated that, “if you have previously had recurrent depression, recommendations may be continuous drug treatment” [w36]. However, to our knowledge, no scientific studies have been conducted to support the need to take antidepressants for a lifetime [17–19].

3.3.Chemical imbalance

Twenty-nine (74%) websites attributed depression to a “chemical imbalance” or claimed they could fix or correct that imbalance. Examples of misleading claims are shown in Table 5.

3.4.Functioning/Quality of life

Only two websites described the effect of antidepressants on functioning or quality of life. A Danish website wrote: “you may notice for example, that you sleep better at night and begin to take better care of yourself. You regularly eat, fix your hair, dress properly ... You tackle everyday tasks easier and you feel you have more energy and better appetite” [w34]. A UK website claimed antidepressants “have a good effect on your sex life as your mood lifts and you become interested in life and relationships again” [w19]. However, the same page stated that, “the side-effects of escitalopram might put a strain on your friendships and relationships, especially in the first few days of taking it” and that “you may have a lower sex drive” [w19].

3.5.Emotional numbing

Five (13%) websites mentioned that antidepressants may cause emotional numbing. For example, one website stated that, “Some people report a general dulling of emotion while taking SSRIs [w5].

3.6.Sexual dysfunction

Twenty-three (59%) websites mentioned that antidepressants were associated with sexual dysfunction. For example, an Irish website stated that SSRIs are known to cause “low sex drive, lack of orgasm and, in men, abnormal erection or ejaculation” [w1].

3.7.Suicidality

Twenty-five (64%) websites mentioned that antidepressants may cause increased suicidal ideation, but 23 (92%) of them contained misleading or contradictory information related to this.

Some of the erroneous information was dangerous. For example, one website claimed that the association between antidepressants and suicidal risk “has not been proved” and that the FDA is still investigating the possibility of such a link [w5]. Another stated that, “there is no clear evidence of an increased risk of self-harm and suicidal thoughts in adults of 18 years or over” [w17], and one noted that taking antidepressants means that “you no longer think of suicide” [w37]. Another claimed that antidepressants “Prevent coercive thoughts and forced acts” [w35], which is also incorrect [19,26–29].

Only two websites mentioned the FDA’s warning that the suicide risk may be observed in people of all ages [29]. Health Canada stated that, “new warnings indicate that patients of all ages taking these drugs may experience behavioural and/or emotional changes that may put them at increased risk of self-harm or harm to others” [w13] and the US National Institutes of Mental Health recommended that, “the warnings from the US Food and Drug Administration (FDA) also says patients of all ages taking antidepressants should be watched closely, especially during the first few weeks of treatment” [w22] although this was not evident on the FDA website we included in our review.

One website acknowledged that, “there is a possibility that taking an antidepressant could make you feel suicidal – even if you didn’t experience suicidal feelings before” [w18] whereas other websites, specifically aimed at “youths”, did not mention the harms at all [w7, w30].

Thirteen of the 25 websites that mentioned the suicidal risk in patients misstated that the risk only occurs “when they first start to take antidepressants” or similar [w3, w9-10, w12-14, w18, w22, w27-28, w31-32, w36]. This is contradicted by evidence that the suicide risk increases whenever the dose is changed [16,19,29].

3.8.Withdrawal, tapering and addiction

Twenty-eight websites (72%) informed patients about the withdrawal effects caused by sudden cessation of antidepressants but 13 of these (46%) had misleading or incorrect information. For example, a Norwegian website stated that, “When you finish treatment, you sometimes get so-called withdrawal symptoms” but “it does not mean that you have become addicted” [w36]. A New Zealand website wrote that antidepressants “are very safe medications and are not addictive but need to be stopped gradually so that the body can adjust to not having them” [w8].

Twenty-three websites (59%) mentioned “addiction” with reference to antidepressants but 22 (96% of those) stated that SSRIs are not addictive. The remaining 16 websites did not make any reference to addiction at all. Only one of the 39 websites (3%), from Norway, acknowledged the addictive effects of antidepressants, warning that people using antidepressants “may get abstinence symptoms” [w36].

A Danish website with drug advertisements wrote that, “antidepressants cannot be abused … and you experience no such hassle by taking them as you can experience with heroin” [w35]. An Australian website wrote that, “antidepressants are not addictive and you will not become dependent on them” [w29]. The website for the UK Royal College of Psychiatrists took it one step further and suggested that if people are having trouble stopping their medications, “Most doctors would say that it is more likely that the original condition has returned” [w17].

Only 20 websites (51%) warned patients to slowly taper the dose of their medication but 16 (80%) of these did not detail time frames.

3.9.Foetal harms in pregnancy

Thirteen (33%) websites mentioned the effects of antidepressants on foetal malformations or foetal withdrawal symptoms [w3, w4, w8, w15-19, w20-23, w25]. The remaining 26 websites either made no mention of the harms of antidepressants during pregnancy or suggested women consult their doctor.

3.10.Psychotherapy

Thirty-seven websites (95%) mentioned psychotherapy as an alternative or adjunct to medications. The websites contained varied levels of detail about talking therapies ranging from cognitive behavioural therapy (CBT) to interpersonal therapy, to group therapy.

3.11.Off-label use of drugs

Fourteen websites (36%) mentioned that SSRIs are prescribed off-label, meaning the medication is not approved for that condition. For example, in Australia, SSRIs are not approved for depression in people less than 18 years of age [40], yet many doctors prescribe the drugs for children.

4.1.Benefits

All websites suggested that antidepressants are effective or used to treat the symptoms of depression. However, none of them explained that the effect in randomised trials is so small that it is below the minimally clinically relevant effect [12–14,19] or might be non-existent because the trials have not been effectively blinded [13,19] or that the benefits could be outweighed by the harmful effects [12,19].

When the benefits were described with numbers, they were widely exaggerated, e.g. SSRIs “help to relieve symptoms of depression and anxiety in approximately 50 to 70% of people” [w9]. It is misleading to write that two-thirds of the patients taking antidepressants will recover [w30] or feel better [w29] because this reflects the spontaneous remission that would have occurred in any case, also in untreated patients [19,20].

4.2.Chemical imbalance

The marketing of depression as a “chemical imbalance” was widely promulgated by Eli Lilly, the makers of fluoxetine (Prozac) [21], the first widely used SSRI, and it has been very harmful. Despite being discredited many years ago [22,23], this myth appeared on three-quarters of the websites we examined. An Australian study showed that 88% of people think it’s the likely cause of depression, and those who believe they have a “chemical imbalance” are more likely to take medication [41]. These false beliefs are likely to discourage people to eventually discontinue their SSRIs, particularly if they believe it is like needing insulin for diabetes [w29], the implication being that such conditions require treatment for life.

4.3.Withdrawal, addiction and relapse

What is called relapse of the depression is very often an abstinence depression, which consists of withdrawal symptoms that would not have occurred if the patient had not been on an antidepressant agent [15,16,19,31,32,36]. Psychiatrists often mistake these withdrawal symptoms for relapse [31,36].

The withdrawal reactions to SSRIs are very similar to those for benzodiazepines, which are widely acknowledged for their addictive properties [30]. It is therefore contradictory to claim that SSRIs are not addictive and the patients’ experience is just the same: Once their body has adjusted to the medications, they often feel they have become addicted because they cannot get off the drugs again [16,19,31,32].

The effectiveness of antidepressants to prevent relapse is questionable. The STAR*D trial involving 4,041 patients found that only 3% of the patients who entered the trial remitted, and then stayed well and in the trial during the year-long follow-up [18]. There is no good evidence in support of patients being advised to take antidepressants for 6-9 months or for their entire life [17–19], as suggested by many websites.

4.4.Suicidality

The increased risk of suicide caused by antidepressants in young people is well known. It is less well known that it has been shown in systematic reviews that this risk exists at all ages, including middle-aged women with urinary incontinence and healthy adult volunteers [19,27–29]. In 2007, the FDA upgraded its warning to include everyone under 25 years of age [42]. Yet, one website trivialised the black-box warning by stating that suicidality was rare and only observed in those “under the age of 18 years” [w35], and the UK Royal College of Psychiatrists claimed that, “There is no clear evidence of an increased risk of self-harm and suicidal thoughts in adults of 18 years or over” [w17].

Several websites had contradictory statements, including switching between scientific evidence and opinions. For example, the Mayo Clinic’s website cited the 2007 FDA warning about the suicide risk [42] but undercut the message by stating that, “not all mental health researchers believe these warnings are necessary. Newer research indicates that the benefits of antidepressants may be greater than the risk of suicide” [w21]. Thus, the Mayo Clinic suggested that antidepressants protect against suicide, which is misinformation.

Several websites trivialised the harms by implying that suicidal and homicidal ideation or behaviour aren’t as serious as suicide. The Mayo Clinic stated that, “None of the children in any of the studies actually took his or her own life” [w21], and the UK Royal College of Psychiatrists stated that, “There is some evidence of increased suicidal thoughts (although not actual suicidal acts)” [w17]. An Australian website, specifically aimed at a young audience, stated that, “The majority of young people who experience suicidal ideation will not go on to take their lives” [w28]. Using such logic, one might say that drunk driving is safe because most drunk drivers do not die in car crashes. There is a close link between suicidal events and suicide, and deliberate self-harm is one of the strongest predictors of suicide [43].

One argument for softening the warnings is to avoid deterring people from taking their medications. Such paternalism in lieu of honest information is common. For example, an Australian website [w28] blamed a drop in the use of antidepressants in the US on media publicity of the 2004 Black Box warning. However, the decline in the use of antidepressants had already begun in 2003 [44]. The argument that antidepressants protect children against suicide is based on flawed observational studies, and the proponents of this erroneous message consistently ignore observational data that demonstrate the opposite [19].

Several websites claimed that fluoxetine is safer than other SSRIs for young people, e.g. “Fluoxetine is considered to be safer than other antidepressants for people under the age of 18” [w19] and “Fluoxetine, an SSRI antidepressant, can be used in the under-18s” [w17]. However, this claim has been refuted [12]. Antidepressants for children are all unsafe.

4.5.Psychotherapy

Psychotherapy has less harms than antidepressants and is often superior, particularly in the long run [45,46], which is why we included it as one of the criteria for informed consent. Psychotherapy was mentioned on most websites, but the information varied considerably, from calling it very effective to only being effective in conjunction with antidepressants. A meta-analysis of ten trials showed that psychotherapy halved the occurrence of new suicide attempts in patients admitted after a suicide attempt [39]. Since research suggests that antidepressants increase the risk of suicide in people of all ages [19,27–29] and have marginal benefits [12–14,19], we believe that psychotherapy should be the prevailing treatment for depression of all severities.

4.6.Limitations

Our study is limited to websites that ranked highly on Google searches from each country. On the other hand, this makes our results generalisable.