When the Compass Doesn’t Point North

Patients and clinicians work better together, not apart

This is a guest post by Alex Mendelsohn, PhD, who is a physicist based in the UK. His writing to date has focused on the therapeutic monitoring of lithium, and ways in which it could be improved, based on his experience as a patient. He has recently started a new Substack, The Psychiatric Multiverse.

---

Recently, several reviews, including one by Awais Aftab, have been published about Laura Delano’s memoir Unshrunk. The reviews summarize her story of iatrogenic harm, her rejection of the medicalized understanding of psychological anguish, and her new role as a guide for people wanting to reduce or come off psychiatric medications. Awais subsequently published an interview with Cooper Davis, Laura’s partner, about Inner Compass Initiative (ICI) that they both run. ICI is a website providing a knowledge base, online community and archive of personal stories to help users “flourish beyond the medicalized, professionalized offerings of the mental health industry.” When I read the reviews and the interview, my stomach twisted.

I find myself in a unique position of relevance. When going on and coming off medication, pharmacokinetics is heavily involved—and I have written about pharmacokinetics in the past.

I am someone who was harmed by psychiatric medication. After a severe reaction to an antidepressant, I went from someone who had never suffered an anxiety symptom to subsequently suffering almost all of them. There is a decent argument that had I not taken that antidepressant and had proper social and psychotherapeutic support, I would not have needed medication for my depression.

But I am also someone who was harmed by advocates trying to inform me about the harms of psychiatric medications. After my severe reaction, I, probably reasonably, was hesitant to trial medications. Some of the psychotherapists and counselors I saw would actively try to dissuade me from trialing them. They said antidepressants would “get in the way of the process.” I was vulnerable, and I believed them. Despite three years in counseling, with what I now realize was a slow decline in functioning, not one counselor ever suggested that perhaps seeing a psychiatrist might have been a good idea.

As I fell deeper and deeper into severe illness and dysfunction, I finally understood my specific condition required medication. But I also thought it was too late—I thought my disease had progressed beyond the rudimentary reach of psychotropics. Nevertheless, I started to see psychiatrists. I often felt scared, alone, and unsafe. I will admit I became very angry with the psychiatric system and for a while I held some of the views seen in the antipsychiatry community.¹

But in the end, I found a medication that started to help. I was wrong—it wasn’t too late. I was, however, extremely fortunate to have the support of my mum, who cared for me. She helped me with the simple ordinary tasks I could no longer do (like cook food) and provided emotional support. I was fortunate to have a home, without having to worry about money. The result is I am alive today.

Psychiatric medication both caused and treated my condition. My first-hand experience has provided evidence that there is no simple story here. There is no simple narrative. Psychotropic medications have the power to both destroy and rebuild lives—and everything in between. Psychiatry is nuanced.

When I became well enough, I wanted to contribute to the psychiatric literature. I had the desire as a patient to point out places where the psychiatric system could be improved.

But I was worried. Given my experience, I did not want to write something which could potentially be misunderstood by a reader. I quickly learned I needed help. I realized I had blind spots everywhere. I was taking my individual circumstance and generalizing it to everyone else. In other words: bias. The stronger the emotion, the stronger the bias. And I was very angry! So, therefore, very biased!

I reached out to psychiatrists. Awais first and then Owen Muir. I was astonished by their open-mindedness and their willingness to help. I’ve also since started to take part in research studies as part of lived experience advisory committees. I was surprised to learn about the growing patient involvement in psychiatric and medical research in general. This was not the story I told myself about the psychiatric system.

Awais and Owen’s input has been invaluable. Not only for their expertise as clinicians—having gone through the extensive medical education required to become psychiatrists—but also their experience of seeing many, many patients.

I agree with a lot of what Cooper says in the interview he had with Awais. I have gone through antidepressant withdrawal myself. I too have felt aggrieved by the lack of resources available. I am a strong advocate for patient involvement. But I can’t help but find myself disappointed.

Inner Compass Initiative’s current team does not seem to include a clinician, psychiatrist or clinical researcher. The ICI website, at the time of writing, only seems to have a single webpage where reviewers are mentioned.² The webpage tells us, “Several medical doctors, psychiatrists, pharmacists, academic researchers and others have also generously provided vital feedback on key sections of the website.” However, according to the Internet Archive Wayback Machine, the webpage appears to be a recent addition.³ Further details about the reviewers or their feedback are not available.

I feel this is important to point out given their website provides an educational section which utilizes sources from clinical psychiatric literature. In my opinion, based on current evidence, it does not look like it has been reviewed by a medical professional. Given my current (still quite limited) knowledge, I can only comment on the lithium part of Inner Compass Initiative’s website “learn” section. I have recognized some of the blind spots I used to have.

I would categorize some of the claims as misleading and unnuanced—weakly backed up by a sparse set of cherrypicked studies. This is my generous opinion. Take the following examples from ICI’s website about lithium toxicity.

The “How effective is lithium as a mood stabilizer?” section starts:

Lithium was determined to be so toxic to humans and lacking in clinical usefulness that it was banned from all food and medical uses by the FDA in the 1940s.

The FDA banned lithium chloride⁴ in March 1949.⁵ John Cade published the first paper on the antimanic properties (and therefore clinical utility) of lithium six months later in September 1949.⁶ You might already see the problem here.

The clue to this predicament is in the “chloride” of “lithium chloride.” It became available in 1948 as a substitute for sodium chloride, or in other words, table salt. Patients that were advised to switch to a low sodium diet, such as those with hypertension or congestive heart failure, started sprinkling lithium chloride on their food. The amount was completely uncontrolled, so it is not surprising that lithium toxicity cases started to show up.⁷ The FDA did not initially ban lithium’s use as a medication, they banned its use as a condiment.⁸

Inner Compass Initiative’s page makes a further lithium toxicity claim in a later section:

Lithium is extremely toxic to the human body and can be very dangerous and potentially lethal at blood serum levels that are very close to the ‘therapeutic’ levels that are commonly prescribed by psychiatrists.

Allow me to make an equally misleading, unnuanced statement: Did you know that one formulation of lithium, lithium orotate, is so safe that it is sold over the counter (OTC) as a supplement worldwide? In fact, there has only been one recorded toxicity case in over 30 years.

Both statements fail to acknowledge a general principle of medicines—the dose makes both the benefits and poison. Compared with placebo, people with standardized 12h lithium serum concentrations (henceforth I will call “Amdisen levels”⁹) under 0.6 mmol/L were not associated with being at any greater risk of adverse events. There is evidence to suggest that lithium may have some beneficial effects at these lower serum concentrations.¹⁰

The therapeutic range is generally considered to be between 0.6 – 0.8 mmol/L.¹¹ This corresponds to the Amdisen levels usually required to prevent manic and depressive episodes for bipolar patients who respond to lithium.¹² The difficulty with lithium is that for most people, there is not much of a buffer between the upper therapeutic Amdisen level and concentrations where lithium toxicity symptoms start occurring (generally reported in literature and guidelines as around 1.5 mmol/L and above, with moderate to severe toxicity starting at 2.0 mmol/L).¹³ Indeed, there have been a small number of cases where people have suffered toxicity symptoms within the recognized therapeutic range.¹⁴

It is nevertheless important to recognize the nuance of lithium toxicity. Firstly, despite the paucity of available data, the annual prevalence of lithium toxicity is thought to be around 1-2%.¹⁵ For patients who suffer toxicity symptoms, the mortality rate has been estimated to be between 0 and 2 percent.¹⁶

There are three types of lithium poisoning: acute, acute on chronic, and chronic.¹⁷ Acute (and acute on chronic to a degree) poisoning generally results in a better prognosis as lithium has not yet been able to accumulate in the brain and other tissues. To my understanding, chronic poisoning is the most dangerous, which occurs when lithium intake exceeds elimination over the timespan of (usually) weeks.¹⁸

A significant reason that patients on lithium require blood tests in the long term (every 3 – 6 months) is to check kidney function. Lithium can cause kidney damage, which can eventually progress to chronic kidney disease (reported rates between 2 – 40%). However, this process is very slow. While there have been occasional reports of it occurring over several years on lithium, for most affected patients this will take decades.¹⁹ Nevertheless, it is in therapeutic monitoring of lithium where I personally believe there are valid issues. I’ve written previously on my experience of confusing guidelines, a lack of notation in the literature and misrepresentations of monitoring data by clinicians.

While I have focused on misleading statements about lithium toxicity for this piece, I don’t want to create the assumption it is the only issue. There are quite a few others. For instance, returning to ICI’s website section on lithium efficacy, the use and interpretation of sources is questionable. Only three references are used: two systematic reviews from the early 2000s (when more recent reviews exist) and a (arguably) polemical blog post. When referencing the 2004 review by Geddes et al. ICI says, “Lithium was not found to significantly reduce depressive feelings,” but the review was about lithium’s efficacy for relapse prevention in bipolar mania and depression (i.e. prophylaxis). Geddes et al. states that “the evidence from our review is consistent with a moderate beneficial effect, rather than an absence of an effect against depressive relapses.”²⁰

I can understand why the team at Inner Compass Initiative may have overlooked these details. Literature review is one of the most difficult skills to learn. It is so easy to fall into confirmation bias. Even the best literature reviewers can’t stop their own confirmation bias—they can only self-correct.

Which is why I am struggling to understand why ICI does not seem to have reached out to any medical professional or researcher for help. In the case that they have, it is puzzling to me why they have not been fully transparent about who specifically has helped them.

I am deeply worried about ICI’s education section in its current state. On the banner of their website, it says that ICI is “helping you make more informed choices about all things mental health: diagnoses, drugs, and drug withdrawal.” The strapline not only explicitly states a much wider scope than medication discontinuation, it implies the information on the site is applicable to visitors who have not yet made a decision about whether they want to discontinue their medication.

I can foresee a psychiatric patient, otherwise doing well on lithium, visiting ICI’s website. After reading the lithium education section they might start to believe they are ingesting poison. Subsequently, the psychiatric patient may very well start thinking about coming off lithium, weighing the risks as much higher than they actually are.

All of the psychiatrists I saw during my time as a psychiatric patient did not seem to be fully aware of just how much power and influence they held over me. It is a point Cooper Davis makes himself in his interview with Awais.

On the other hand, I don’t have the impression that ICI’s team is aware of the influence they have over the patients that visit Inner Compass Initiative’s website (or people who decide—and are able—to pay to become part of Inner Compass’s Exchange community²¹). Especially given the publicity afforded to Cooper Davis and Laura Delano, alongside their relatable experiences as psychiatric patients. I fear that ICI will simply end up on the other side of the iatrogenic harm coin.

In Cooper’s interview with Awais, he said he was “always interested in hearing the ways in which psychiatrists account for the unknowable, especially when the cost of hubris can be so high.”

So, my question to Cooper Davis and ICI’s team is: given you find yourself in a position of significant influence, offering a service to some of the most vulnerable people in society, what are the ways in which you are accounting for the unknowable, and for your own biases, especially when the cost of hubris can be so high?

---

Psychiatry at the Margins is a reader-supported publication. Subscribe here.

1. As an example, see the first article I wrote. Looking back, there are some parts of the article I don’t feel very proud about.

2. Just before publication, an instance on a second webpage was spotted (it is not a recent addition). On the about us page it says:

“Though we have sometimes received comments and feedback from medical, pharmacological, and clinical practitioners of different kinds as we’ve created this website, and some of our team members, contributors, and volunteers may work as practitioners, all of the final editorial decisions related to the content on this website have been made by laypeople and are provided for general educational purposes only.”

3. The Wayback Machine crawls the web and periodically takes “snapshots” of webpages and archives them according to the date the snapshot was taken. ICI’s homepage has snapshots dating back to 2017, whereas the webpage where reviewers are mentioned has only one archived snapshot from January 14^th 2025.

4. And lithium salts in general

5. H. Aaron, ‘Dangerous drugs’, Consumer Reports, vol. 14, pp. 171–173, 1949.

6. ‘Lithium salts in the treatment of psychotic excitement’ Cade (1949).

7. See ‘Lithium poisoning from the use of salt substitutes’ Corcoran et al. (1949) & ‘Lithium chloride as a substitute for sodium chloride in the diet: observations on its toxicity’ Hanlon et al. (1949)

8. For further information see Chapter 4 of The History of Lithium Therapy by Frederick Johnson for a through account. For freely available shorter versions see Edward Shorter’s 2009 article, Mark Ruffalo’s 2017 article, or Alan Strobusch & James Jefferson’s 1980 article (click “Alternate access options”).

9. In a short paper I wrote with Awais and Owen Muir, we proposed “Amdisen levels” as an attempt to better communicate standardised lithium serum concentrations.

10. For a concise guide on the benefits and risks for different Amdisen level ranges, see ‘Lithium: how low can you go?’ Strawbridge and Young (2024).

For low dose lithium benefits and risks see ‘Identifying the neuropsychiatric health effects of low-dose lithium interventions: A systematic review’ Strawbridge et al. (2023), ‘Beyond its Psychiatric Use: The Benefits of Low-dose Lithium Supplementation’ Hamstra et al. (2023) & ‘Lithium in the Kidney: Friend and Foe?’ Alsady et al. (2015).

For a balanced review of the potential suicide prevention properties of lithium, see ‘Prevention of suicidal behavior with lithium treatment in patients with recurrent mood disorders’ Tondo and Baldessarini (2024).

11. The International Society for Bipolar Disorders (ISBD) and the International Group for the Study of Lithium Treated Patients (IGSLi) assembled a group of experts (a “Task Force”) to tackle the question of what the optimal Amdisen level range should be.

In their review, published in 2019, they recommend the 0.6 – 0.8 mmol/L range. But they also caveat this with the recommendation for slightly different levels for different responses and older aged people (e.g. 0.4 – 0.6 mmol/L for sufficient response but poor tolerance).

However, the reviewers also noted there was “a lack of well-designed, double-blind randomized dose finding … studies assessing optimal serum level […]. This paucity of data makes it difficult if not impossible to formulate confident evidence-based recommendations….”

12. Initiating and monitoring lithium levels for patients with bipolar disorder is not straightforward. Episodes of mania and depression can occur, which may require adjustments of lithium dose. Malhi, Gershon and Outhred developed the “Lithiumeter2.0” (published in 2016) as a lithium level clinical aide for psychiatrists managing patients with bipolar disorder.

13. See ‘The use of lithium for the treatment of bipolar disorder: Recommendations from clinical practice guidelines’ Malhi et al. (2017) and ‘Lithium Toxicity’ Hedya et al. (2023)

14. E.g. ‘Lithium-related neurotoxicity despite serum concentrations in the therapeutic range: risk factors and diagnosis’ Megarbane et al. (2014).

15. See ‘Lithium monitoring and toxicity management’ H. Godden (2024).

16. See the review ‘Lithium side effects and toxicity: prevalence and management strategies’ Gitlin (2016).

17. The definitions for each are here

18. See ‘Lithium Poisoning’ Baird-Gunning et al. (2016) for a thorough review of lithium toxicity. See ‘Lithium Toxicity’ Hedya et al. (2023), for a free brief review.

19. See the review ‘Key questions on the long term renal effects of lithium: a review of pertinent data’ Gitlin & Bauer (2023) and the article ‘Association between lithium treatment and renal, thyroid and parathyroid function: A cohort study of 6659 patients with bipolar disorder’ Wiuff et al. (2023).

20. The confusion might have occurred because Geddes et al. stated that they did not find a statistically significant effect for bipolar depression relapse prevention. On this point, however, it is important to note that in the more recent updated review, with more data, one of the statistical models did reach statistical significance (2004: p = 0.06, 2014: p < 0.001).

21. Inner Compass Exchange offer scholarships for people who are currently not in a financial position to afford membership.