One Dragon at a Time: A Clinical Framework for Navigating Chronic Complex Illness
By Yoon Hang (John) Kim, MD, MPH · Board-Certified in Preventive Medicine | Integrative & Functional Medicine
After more than two decades of caring for patients with complex, multi-system chronic illness, I have come to believe that the most important clinical skill in this work is not diagnostic pattern recognition or pharmacologic knowledge. It is restraint. The sickest patients I see — those with overlapping post-infectious syndromes, mast cell activation, dysautonomia, tick-borne infections, mold-related illness, autoimmune disease, and the cumulative aftermath of decades of medical trial and error — are rarely made well by doing more. They are typically made well, when they get well at all, by doing less at any given moment and doing it more thoughtfully. The operating metaphor I return to again and again with patients is this: we fight one dragon at a time.
The Core Principle: One Intervention, Slow Titration, Graceful Retreat
In clinical practice, the temptation to layer interventions is enormous. A patient arrives exhausted, in pain, mast-cell reactive, and sleep-deprived, and it is natural to want to address every dimension at once. This is almost always a mistake. When three or four therapies are introduced simultaneously and the patient deteriorates — as patients with fragile physiology frequently do — there is no way to know which intervention caused the problem, which one may have helped, and what to adjust. The information is lost. The patient loses confidence. The clinician loses leverage.
The discipline I ask of patients, and of myself, is to introduce one change at a time, begin at the lowest plausible dose, and titrate upward slowly enough that the body has time to speak. If a side effect emerges, the rule is simple and non-negotiable: stop, reset, and restart lower. This approach feels painfully slow in the first few weeks. It is, in fact, the fastest route to durable improvement, because every step taken is a step the patient can actually stand on.
Low-Dose Naltrexone as a Cornerstone
Among the tools available for immune and neuroimmune modulation, low-dose naltrexone has earned a place I would describe as foundational rather than optional. Its utility across conditions as varied as fibromyalgia, Crohn's disease, multiple sclerosis, complex regional pain syndrome, Hashimoto's thyroiditis, long COVID, and mast cell activation syndrome reflects the fact that LDN does not act on a single pathway so much as on the regulatory architecture of the immune and endorphin systems. For patients whose fundamental problem is dysregulation rather than deficiency, a medication that restores regulatory tone is often exactly what is needed.
What continues to surprise clinicians new to LDN is the astonishing range of dose sensitivity. Patients exist across the full spectrum from those who tolerate and benefit from the traditional 1.5 to 4.5 mg range, to those who require microgram dosing, to a small but real subset who respond only to nanogram preparations and who experience unmistakable side effects from doses most physicians would consider homeopathic. There is no universal correct dose. The correct dose is the one that produces benefit without cost in that particular person.
LDN also carries an underappreciated downside that bears direct acknowledgment. In patients whose endorphin systems are already depleted — from chronic pain, prolonged opioid exposure, chronic infection, or chronic stress — the pulse of opioid receptor blockade that LDN produces can transiently worsen symptoms before the compensatory upregulation of endogenous endorphin production occurs. When this happens, the answer is not to abandon the medication but to recognize what is happening, lower the dose, and proceed more gently. This is one of many reasons LDN should not be self-directed; the titration conversation is where the clinical judgment lives.
MCAS and the Common Pathway of Immune Dysregulation
Mast cell activation syndrome has moved, in my view, from being a discrete diagnosis to being better understood as a final common pathway — the route by which a great many insults ultimately express themselves as daily illness. Viral infections, including SARS-CoV-2, Epstein-Barr, and others, can provoke it. Tick-borne disease, particularly Lyme and its co-infections, frequently does. Autoimmune conditions travel with it. Environmental exposures, most notably water-damaged buildings and the mycotoxins they produce, can both trigger and perpetuate it. By the time a patient meets the clinical picture of MCAS, there is almost always an upstream driver that deserves attention.
The clinical implication is important. Treating mast cell symptoms alone — antihistamines, stabilizers, dietary restriction — will reduce symptom burden, and in some patients that is enough. But in patients whose mast cells are being provoked daily by an unaddressed infection, an untreated autoimmune process, or ongoing mold exposure, symptom management without root-cause work tends to produce a ceiling that is difficult to break through. The more durable gains come from identifying and addressing what is keeping the immune system agitated in the first place.
Thoughtful Combination, Not Polypharmacy
There is a meaningful difference between combining therapies thoughtfully and accumulating them. A common and useful pairing is LDN with a mast cell stabilizer such as ketotifen, where the two agents work on complementary aspects of immune regulation. What makes this kind of combination effective is that each component was introduced, titrated, and stabilized before the next was added. What produces the opposite outcome — a patient on fourteen medications, none of which are clearly helping or clearly harming — is the habit of adding before confirming. The discipline, again, is to layer only after stabilizing, and to know why each agent is in the regimen.
When selecting among agents, the mechanism matters more than the brand. A generic compounded preparation that delivers the active molecule at a dose the patient can tolerate will usually serve better than a more expensive branded product chosen because it is familiar. The question to hold in mind is always: what is this molecule doing, and is that what this patient needs right now?
Dosing, Pharmacology, and the Reality of Individual Variability
Clinicians trained in conventional pharmacology are accustomed to dose ranges that vary perhaps two- or threefold across patients. In the population I see, it is not unusual for individual dose requirements to vary by three or four orders of magnitude — from milligrams to micrograms to nanograms to, occasionally, picograms. The genetic and epigenetic contributions to this variability are real and include polymorphisms in cytochrome P450 enzymes, in methylation pathways, in opioid receptor density, and in mast cell reactivity. A dose that is therapeutic in one patient is toxic in another and inert in a third.
The consequence is a principle I repeat often: more is not better. Pushing a dose upward in search of effect, when effect has not appeared, is one of the most common ways patients are made worse in the name of being helped. The correct move when a dose is not working is almost never to increase it reflexively; it is to pause, reassess the diagnosis, reconsider whether the mechanism applies to this patient, and only then decide whether a dose change is warranted.
Practical Principles at the Bedside
Three working rules guide most of my decisions in clinic. If something works, do more of it — whether that means continuing the medication, reinforcing the lifestyle change, or expanding the therapy that is clearly moving the patient forward. If something does not work, stop and reassess rather than persisting out of hope or habit. And aim for consistent small improvements rather than dramatic transformations. In chronic complex illness, home runs are rare and often illusory; the patients who recover do so by stringing together singles and doubles over months and years. A patient who is five percent better every month is, at the end of a year, in a fundamentally different place.
Adjunctive Therapies: Useful, Not Universal
Several adjunctive approaches deserve mention because they come up frequently in patient conversations and because each has a legitimate but bounded role. Neurofeedback has meaningful evidence in specific applications — sleep regulation and autonomic dysregulation among them — but the cost, the need for consistent sessions, and variability between providers make it a reasonable option for some patients and impractical for others. Exogenous ketones can be useful in selected neurological and metabolic contexts; they are not a general-purpose tool, and the enthusiasm around them has outpaced the data. Methylene blue, at appropriate doses, has interesting antimicrobial and mitochondrial properties and has a growing literature behind its careful use, but it is strongly dose-dependent, interacts with serotonergic medications in ways that matter, and is not a supplement to be casually added. Each of these deserves a specific rationale in a specific patient, not a blanket endorsement.
The Reality of Chronic Illness Recovery
One of the more honest conversations I have with new patients is about what recovery actually looks like. In chronic complex illness, a realistic goal is often seventy to eighty percent of baseline function, achieved through patient iterative work, rather than a full return to who the patient was before they got sick. That is not a failure of medicine; it is an accurate description of what the literature and clinical experience support. Patients who arrive expecting a cure in six weeks are almost universally discouraged within six months. Patients who arrive willing to work, willing to titrate, and willing to measure progress in small increments tend to do remarkably well over years.
This is also why these patients are not well served by going it alone. The internet has democratized access to information about LDN, MCAS, mold illness, and post-infectious syndromes, and that is mostly a good thing. What it cannot replace is the judgment of a clinician who knows the patient's full history, who can integrate signals across systems, and who can make the next call when the last one did not produce the expected result. Self-directed experimentation in this population tends to reproduce the same error the physicians made: too much at once, too fast, without a plan for retreat.
The Larger Picture
If I had to reduce the framework of this work to its essentials, I would say that chronic complex illness care rests on three pillars: precision dosing, which respects the enormous variability between patients; patience, which accepts that the nervous and immune systems heal on biological rather than consumer timelines; and iterative experimentation, which treats each intervention as a question the body answers over weeks and months rather than a command it obeys.
The goal in the first phase of care is stability, not optimization. A patient who is stable — whose mast cells are quieter, whose sleep is more reliable, whose flares are less frequent and less severe — has a platform from which further gains can be built. A patient who has been optimized on paper but who is lurching from crisis to crisis has no such platform, regardless of how impressive the regimen looks on a medication list. One dragon at a time, slowly, is not a concession to the difficulty of the work. It is the method by which the work actually succeeds.
About the Author
Yoon Hang (John) Kim, MD, MPH is board-certified in Preventive Medicine and in Integrative/Holistic Medicine. He founded Direct Integrative Care, a membership-based telemedicine practice capped at 99 active patients, and also practices at Hill Country Integrative Medicine in Fredericksburg, Texas. He is a UCLA-trained Medical Acupuncturist, an Osher Fellow of the Andrew Weil Center for Integrative Medicine at the University of Arizona, an IFM Scholar, and a recognized author and educator on low-dose naltrexone. He is licensed in Iowa, Illinois, Missouri, Texas, Georgia, and Florida.
