Notes on Depression and Anxiety

This is a non-comprehensive summary and clarification of some important information on depression (and anxiety), primarily sourced from two books The Noonday Demon and Lost Connections. 

  

The former was published in 2001, with the current version featuring an added Epilogue written about 15 years later. For that reason, at the very least, much of the book may be somewhat dated in terms of the more recent approach taken in understanding depression and the associated findings. For example, Solomon only mentions the biopsychosocial model once, which is now widely accepted as the dominant model for understanding depression –the biological, psychological, and social nature and origins of depression. He tends to take a much heavier biological approach, such as in focusing on chemical antidepressants for treatment. Incidentally, his father is (or was) a very high-level figure in the pharmaceutical industry which likely played some influencing role. Nevertheless, he does cite the endogenous (or internal) and reactive (or exogenous/ external) model for understanding depression a number of times. Solomon also highlights how depression is a condition which can be over-determined by many factors at the same time. In summary, this is a very comprehensive and long! book, and does to an extent address social factors (such as poverty, politics, and evolution –the latter perhaps even better than Hari) relating to depression. However, the book does not necessarily frame depression as appropriately or ideally as in Lost Connections –such as in fully appreciating the role certain social treatments may have over and above biological treatments. 

Lost Connections was published in 2018 and aims to balance the cultural and psychiatric scales towards better appreciating the social (and even psychological) causes of depression. He outlines seven core social reasons that are behind the significant levels of depression and anxiety in the world today, all framed in terms of a loss of connection. Hari comes back to the point a number of times that the biopsychosocial model of depression is and has been widely accepted by the psychiatric and psychological community for some time now, but that the biological role is unduly emphasised because of the political and economic ramifications around radically changing our culture, and the specific role of the pharmaceutical industry in favouring chemical treatments. Hari also ties the use of chemical antidepressants to our individualistic and materialistic society, and equates this treatment as being a product aligned to this form of culture. He effectively implores readers to start a revolution in remaking society to better serve our happiness and well-being –and at the very least to start taking the social origins of depression seriously, especially if we are to really make progress in treating depression.

Lost Connections and The Noonday Demon appear to diverge on some key points, all of which require further clarification and research. The scientific and other references in each book may be a good place to start (many of which are provided in this article), and it is also possible (or even likely) that these divergences are reflective of present uncertainties in the wider scientific community.

Perhaps the most important divergence is in their alternative views when it comes to the effective role of chemical antidepressants in treatment. Hari brings to light the work of Irving Kirsch and his book, The Emperor’s New Drugs, published in 2009, which details research undertaken to assess the role of placebo and the potential higher efficacy this may have over and above chemical antidepressants. Hari presents these findings as if it is generally conclusive that this is the case, with a minor caveat that for some people, chemical antidepressants may serve some appropriate or limited use. Moreover, Hari emphasises that chemical antidepressants may only be useful for a short period of time (citing Peter Kramer’s recommendation of usage for only 6 to 20 weeks), and that they come with possible side or adverse effects which likely negate any positive use.  

Solomon, on the other hand, throughout his entire book equates chemical antidepressants as one of two of the most accessible treatments for depression (the other being therapy), seems to advocate for their widespread (and continual use –even for years), and generally considers their risks, but often seems to make a case in which the positives outweigh the negatives. One compelling example of this is brought home when it comes to expectant mothers, and he tells a story (in the Epilogue – Page 487) of how a mother who went off her chemical antidepressants ended up committing suicide, killing herself and her baby, thus highlighting the important considerations to both sides of this debate. Solomon reflecting back years later in the added Epilogue at the end of the book seems to stick to his guns in his support of chemical antidepressants. His main response to the recent placebo findings (also in the Epilogue) was the following, 

“Substantive critiques of antidepressants focus on two main concerns. First, a number of researchers have argued that the drugs’ efficacy is entirely due to the placebo effect. Second, many have claimed that the drugs impel people to suicide. They further propose that psychiatry’s medicalization of normal states fuels the despair it professes to address; that the widespread use of antidepressants is almost entirely provoked by a rapacious pharmaceutical industry; and that our ability to map mental illnesses in the brain proves that there is no basis for developing pharmacological remedies. These arguments have been put forth most notably in Irving Kirsch’s The Emperor’s New Drugs, Robert Whitaker’s Anatomy of an Epidemic, Daniel Carlat’s Unhinged, several books by Peter Breggin, and influential essays by Marcia Angell, former editor of the New England Journal of Medicine. […] 

Most of these authors’ primary arguments have been refuted. Kirsch’s work demonstrating that placebos can be as effective as medication for depression has been challenged from multiple angles. Evidence suggests that the high placebo response he documented owes a great deal to the structure and duration of the study and the recruitment process. An analysis by Pim Cuijpers et al. of more extensive data than Kirsch’s indicates that placebos are highly effective but that antidepressants are consistently more so. Konstantinos Fountoulakis has found that Kirsch miscalculated the mean drug-placebo difference. Even Carlat has observed “an unequivocal if perplexing truth about psychiatric drugs –on the whole, they work.” Kirsch maintains that though antidepressants are somewhat effective for people with acute depression, they are nearly useless for people with milder depression. In the Journal of the American Medical Association, Robert Gibbons and colleagues at the University of Chicago pointed out the methodological flaws of studies such as Kirsch’s, reanalysed data from nearly five thousand patients, and concluded, “Patients in all age and drug groups had significantly greater improvement relative to control patients receiving placebo.” Studies indicate that while many people manifest a powerful initial placebo response –predicated in part on the close attention they receive in clinical trials –more than 40 percent have relapsed quickly, while fewer than 20 percent of those on medication have done so. Withdrawal studies make an even stronger case. When subjects who have improved on antidepressants are withdrawn under double-blind conditions, meaning that some continue to take antidepressants while others are switched to placebos, virtually every study finds more frequent relapse in those on placebos. Overall, people respond to placebos about one-third of the time and to a given medication about half the time –a sizeable difference.” [More can be read by searching for the book on Google Books, using the above passage]

Solomon thus raises an important factor to consider when weighing the efficacy of chemical antidepressants against the placebo response, and that is remission. It appears that Hari does not address this point, and so it may prove true that although placebo may be as powerful (or more) as any one chemical antidepressant initially, this may not be lastingly so. It is likely worth reviewing the scientific literature and Kirsch's views to see what else is said about the remission factor –along with other important matters such as possible methodological flaws with Kirsch’s study. At the very least, Kirsch has addressed Fountoulakis claims, and Hari does cite this response as a reference in his book [Kirsch et al., “Calculations are correct: reconsidering Fountoulakis & Möller’s re-analysis of the Kirsch data,” International Journal of Neuropsychopharmacology 15, no. 8 (August 2012): 1193–1198, doi: https://doi.org/10.1017/S1461145711001878]. 

A psychologist who reviewed Kirsch’s work provides at least a partial response to the remission factor [https://www.goodreads.com/review/show/665217205], wherein the following was stated, 

“According to Kirsch, the evidence for antidepressants' superiority to placebos (by a margin which is statistically but not clinically significant, according to Kirsch) comes from double-blind studies where, presumably, neither subjects nor researchers know who is getting the antidepressants and who is getting the placebos. However, Kirsch points out, subjects are likely to realize whether they are getting antidepressants or placebos based on the side effects. And when that happens, subjects who are experiencing side effects realize they are getting antidepressants and become hopeful about their improvement, leading to a self-fulfilling prophecy, while subjects who are getting placebos have the opposite experience. Further, Kirsch adds, once side effects' influence are controlled for by giving active placebos (i.e., placebos which mimic the side effects of antidepressants but are otherwise inert), the small difference between antidepressants and placebos is no longer statistically significant. 

But wait -- there's more. Apparently, drug companies are anxious for you to think their antidepressants work. So anxious, in fact, that if ten studies are performed comparing antidepressants to placebos and only three of those studies find a stronger effect for antidepressants, those three studies will get published. The remaining seven will not, skewing our views of antidepressants' efficacy. This is but one example of cherrypicking by the pharmaceutical companies.” [For more information on Kirsch’s work outside of Hari’s book, this may be a good, balanced summary – https://www.cambridge.org/core/journals/the-british-journal-of-psychiatry/article/clinical-trials-of-antidepressant-medications-are-producing-meaningless-results/E76CC5BDE0755D56DBD1FD22A564C747]

As concluded by Solomon on Kirsch’s views, there may be at least some efficacy of these drugs for those with acute depression (although unsure what this might be exactly on the Hamilton scale –is it merely 1.8 points?), but of course this would have to weighed against the possibility or likelihood of side or adverse effects occurring. Moreover, it is not quite known whether Solomon’s understanding of Kirsch’s views includes the factoring of publication bias and results found when active placebos are used. Yet, given neither of these two points were mentioned by Solomon, perhaps not? Ultimately, deeper research is needed. 

One possible rebuttal to Kirsch’s study on chemical antidepressants and placebo is that often optimal treatment is not received by patients, including that no one psychiatric drug will be effective for an individual, and it may take testing of a few before one is found to be efficacious. Moreover, there is the issue of tolerance, perhaps acute and short-term (known as tachyphylaxis), which is the condition of losing a previously effective response while still on adequate treatment. Solomon points out, “I did not learn until much later on that while more than 80 percent of depressed patients are responsive to medication, only 50 percent are responsive to their first medication—or, indeed, to any particular medication.” In alignment with this understanding, Hari cites the results from the Star-D Trial which involved patients potentially testing multiple medications, given that the first for example may not be effective. It found that although some 67% of patients did feel better using chemical antidepressants, within a year half of the patients were fully depressed again, and only one in three of the people who stayed on the pills had a lasting proper recovery from their depression. He says that, “This evidence has been followed up several times since –and the proportion of people on antidepressants who continue to be depressed is found to be between 65-80%.” Whilst this figure seems low (particularly because [according to Solomon] at least 75% of patients come off chemical antidepressants, often due to side effects by 6 months), what would the figure be without chemical antidepressants –a natural recovery rate? Solomon points to John Greden of the University of Michigan, who says that “This illness has an eighty percent relapse rate within a year without medication, and an eighty percent wellness rate with medication”. Greden’s view (at least the latter one) seems to conflict with the Star-D Trial results, and unless a study is being referred to by Greden, as opposed to say anecdotal experience, it may be better to go off the Star-D Trial findings. It is possible (maybe even likely), however, that he is equating a wellness rate with responsiveness to medication (improvement in depressive symptoms), yet a reference for this statement of his was not provided by Solomon so it cannot be confirmed.

Another relevant question in this debate on the efficacy of chemical antidepressants is, if a brain becomes impaired on a long-term or permanent basis (such as having a damaged cortisol system due to chronic stress), do chemical antidepressants serve more of a role in treatment? Hari quotes the neuroscientist Marc Lewis as basically saying, “The pain caused by life going wrong can trigger a response that is “so powerful that [the brain] tends to stay there [in a pained response] for a while, until something pushes it out of that corner, into a more flexible place. And if the world keeps causing you deep pain, of course you’ll stay trapped there for a long time, with the snowball growing.” Hari’s emphasis on the possibility of long-term brain dysfunction thus seems quite minimal in this framing. The possibility of ultimately “inescapable” brain changes is not brought up –outside of the “snowballing effect”. Conversely, Solomon brings much more attention to the potential biological ramifications (long-term and permanent) arising from say chronic stress or depression. He cites studies (see Page 517-518 in Notes for his references on brain changes) that apparently show how, “the first episode of major depression is usually closely tied to life events; the second, somewhat less; and by the fourth and fifth episodes life events seem to play no part at all.” Whilst I personally believe in the powerful capacity of the body to heal itself, and the resilience and adaptability in general of the human organism (coming from someone who has lived with chronic stress), it is still worth thinking about possible longer-term or even permanent changes to the structure and biochemistry of the brain, and how this might influence possible treatments. In this analysis, it would seem that addressing social and psychological factors to depression would still be highly relevant, but in certain circumstances (such as for those with acute, or recurring depression) biological treatments such as chemical antidepressants may be just as relevant or more-so.

In other biological circumstances, psychiatric or other drugs may be irrelevant and/ or counter-productive. Robert M. Sapolsky, professor of biology and neurology at Stanford University, wrote a book called, Why Zebra’s Don’t Get Ulcers (2004, third edition), which explains a possible scenario in which this might be the case. Endocrinological in nature, it relates to how “abnormal levels of a number of different hormones often go hand in hand with depression” (Page 289). Those whose secretion of thyroid hormone is too low can develop major depressions, and when depressed, may be atypically resistant to chemical antidepressants working. He says that many people with what seems to be depression of a purely psychiatric nature, turn out to have thyroid disease. Sapolsky also touches on a major feature of female depressions, which is that at certain reproductive points (postpartum, menstruation, and menopause) women are at a particularly higher risk for depression. This relates to fluctuating levels of estrogen and progesterone –hormones which “can regulate neurochemical events in the brain, including the metabolism of neurotransmitters such as norepinephrine and serotonin.” In support of this view, Sapolsky says that researchers cite evidence that women can get depressed when artificially changing hormone levels (such as when taking birth control pills). Solomon also discusses this aspect of female depression in his Populations chapter, wherein he states the mood effects are not consistent or predictable – “Sudden lowering of estrogen levels will cause depressive symptoms, and high levels of estrogen promote a sense of well-being.”

A few other worthwhile points to consider on chemical antidepressants –

1) It’s important to better understand the placebo response, and the powerful role that belief plays. Solomon acknowledges, “Frankly, I think that the best treatment for depression is belief, which is in itself far more essential than what you believe in. If you really truly believe that you can relieve your depression by standing on your head and spitting nickels for an hour every afternoon, it is likely that this incommodious activity will do you tremendous good.” So, maybe you don’t need an authority medical figure telling you everything will work out (or not) with x intervention or y condition –if you “really truly believe” in the powerful capacity for your body to heal itself, and that everything will work out better than best. And even if possible hard limits on individuals’ belief were to exist (such as natural laws or social constraints), might it still not be the case that an individual would often or generally best be served by believing in the most hopeful future that serves them? Such a positive mindset would also seem to characterise the “learned optimism” of CBT. Incidentally, it appears that Kirsch (as per the psychologist’s review) seems to consider psychotherapy/ CBT to be the best form of treatment. Yet interestingly, Hari seems to downplay the usefulness of CBT in particular, stating (on page 223) “the evidence suggests that this kind of therapy has a small effect, and it doesn’t last very long—but its effect is real nonetheless. (To be fair, the main champion of CBT, Professor Richard Layard, says it should be combined with social change if you want the best results.)”. The reference Hari provided for this statement was a meta-analysis study which is generally considered one of the better types of evidence given it pools research from multiple studies on the same subject to see what is commonly found [D. Lynch, “Cognitive behavioural therapy for major psychiatric disorder: does it really work? A meta-analytical review of well-controlled trials,” Psychological Medicine 40, no. 1 (Jan. 2010): 9–24, doi: https://doi.org/10.1017/S003329170900590X.]

2) Part of the efficacy of chemical antidepressants that Solomon talks about is that it may derive from having a synergistic effect in combination with talk therapies, such as CBT. He says (in the Epilogue – Page 475),

“Depression is a disease of loneliness, and substantial evidence holds that informed human contact is among its best solutions. The perception that someone is paying attention to what you are experiencing is greatly reassuring. The need for medication makes people feel broken; therapy makes people feel whole. Further, solid evidence supports the conclusion that on average the combination of therapy and medication works better than either one alone. Therapy can help people avoid severe relapse, which is economically preferable (even for insurers) to bouncing in and out of hospitals. Cognitive behavioural therapy has a solid track record in addressing mild to moderate depression, but it and other similarly effective strategies are used too seldom. The dangers are less acute than those attached to incompetent brain surgery, but things can go awry where trust is violated, bad advice is offered and illness is allowed to escalate.” 

Hari does not touch on this possible beneficial synergy effect however, so more confirmation on what the scientific evidence might say about this would be useful, such as any specific quantification of the positive effects. One study though that Solomon points to (referenced on Page 519 in his Notes), and that may contradict Hari’s position, found that “less than half [of depressives] experienced significant improvement with just medication; that less than half experienced significant improvement with cognitive behavioral analysis; and that more than 80 percent experienced significant improvement after being treated with both.” Finally, even if Hari’s cited study does accurately reflect the scientific truth on CBT’s effectiveness (with or without therapy), the finding is still likely an average, which means that for some people, who for example may have a more psychological (as opposed to say biological) based depression, it may be more relevant and efficacious.

3) When it comes to the biological mechanism of how antidepressants work, Hari constantly brings attention to how the serotonin theory has been discredited as being responsible for depression (although does cite a study in the references [Page 273] as showing an indirect relationship), but he does not discuss the more recent “neurotrophin hypothesis” of depression, which Solomon does touch on (in the Epilogue – Pages 455, after stating that the “chemical imbalance” theory has been out of circulation for more than a decade – 478-479), which suggests “that depression and stress lead to impaired neuroplasticity, and that antidepressant therapies -psychotherapies, medication, electroconvulsive therapy, deep brain stimulation (DBS), and even sleep deprivation –all raise levels of brain-derived neurotrophic factor (BDNF), which aids the formation of new neurons and the development of synapses.” This new hypothesis may better explain the biological mechanism and therefore offer more accurate and evidence-based chemical antidepressant treatments in the future, although as Solomon (again in roughly 2016) states, “while it helps explain existing treatments, it has so far not been the basis for new ones.”

4) The focus on chemical antidepressants here, in part as a result of the conflicting positions held by Solomon and Hari, may have unduly emphasised the possible importance of this one biological treatment in favour of other biological treatments, and indeed social and psychological treatments. So, it is important to fully understand the array of treatments available and place chemical antidepressants in their proper context of use. For example, it may have been just as relevant to dedicate as much attention to discussing how to motivate oneself and take positive action when it comes to nutrition, such as not buying unhealthy foods to begin with so one isn’t later tempted, thinking about ones’ living or work environment and other’s possible negative influences on ones’ choices, habituating taking ones’ lunch to work so one doesn’t buy fast food, seeing a dietician to formulate a specific plan relevant to oneself, growing an herb and vegetable garden, working out your self-discipline muscle, journaling ones’ thoughts about the subject, such as meeting certain goals, and so on and so forth. Yet, perhaps because there isn’t much apparent controversy over the beneficial effects of healthy eating on mental health it is somewhat taken for granted that we already know what to do. Still, it is worth properly exploring these other areas.

To summarise this review, there are some strong differences of perspective evident from reading The Noonday Demon and Lost Connections. The disparity between the biological focus and the social focus may be the most apparent, which results in treatments being advocated that are reflective of these two approaches. In the former, we have chemical antidepressants and talking therapies taking centre stage as the most accessible treatments for people with depression, with a much more eclectic set of non-social treatments also discussed. In the latter, we mostly have a set of social causes outlined (such as loss of connection to other people, nature, and meaningful work and values), with treatments reflective of these ails. The divergence of view characterised by Solomon and Hari thus manifests in contrary positions held over the efficacy of chemical antidepressants over other forms of treatment. The limitations and risks that come with these drugs, along with the role that placebo and positive belief has, is also part of the debate –the uncertainties possibly also a result of lack of answers in the wider scientific community.

Despite all the disagreement and contradictions though, there is also strong agreement (explicit and implicit) between both authors as-well. Both do recognise at least endogenous and reactive causes of depression, and indeed, although Solomon barely discusses it, appears to accept the biopsychosocial model of depression as-well. Both appear to accept that chemical antidepressants may serve at least some useful role, and likewise acknowledge the issue of side or adverse effects that factor into considerations. Both accept that adequate nutrition, exercise, social connections, supportive family and friends, adequate financial means and work, lack of stress, and belief are critically important as a foundation to ones’ well-being and mental health. They also simultaneously acknowledge that religion or prayer, and meditation can serve a positive role. Both Solomon and Hari would agree that much more needs to change in our culture to better serve our mental health, and that in many ways, we do have a significant task at hand both in terms of how many are affected by depression and anxiety, and in terms of what will need to be done in-order to accomplish positive changes on a societal level. Finally, both appear to show what may be described as cautious optimism about making progress on these fronts, and likewise also acknowledge the important strides that we have made thus far in treatment and understanding depression.

In conclusion, I would recommend both books, although due to the amount of necessary time to invest to do this, if only one could be read it would be Lost Connections. If both are read, I would recommend reading The Noonday Demon first, followed by Hari’s book. This is because Solomon’s book was written earlier and so there is a good, natural progression built up in ones’ understanding, at least in part as a result of the unfolding knowledge of the wider scientific community in the elapsed time. 

---

Thank you for your time and attention! I hope you have found this an engaging and thoughtful read. If you think you have derived value from this content, your support would be highly appreciated :) 

ETH:  0xFcde7A9af2da9B200D22fC21C806579a2a5d4642

LTC:  M9HB5G8krgb82VyzcCQs8natBsrkUyyNXH

ZEC:  t1XSqrexTHoq4r52bUoDbkqCcAxbWrZifSq

I also look forward to any comments and/or questions you may have. 

---

[For access to scientific papers linked in this document, the use of an extension called ‘Unpaywall’ (http://unpaywall.org/products/extension), and the website ‘Sci-Hub (which may be in a legal grey area for use) (https://whereisscihub.now.sh/) may be helpful]