Perspective

The Seventy Years Task Force

A history that might have been.

A note on genre

Uchronia is the literature of the point of divergence. Philip K. Dick asks what if the Axis had won. Robert Harris asks what if Heydrich had not been killed. Éric-Emmanuel Schmitt asks what if the young Hitler had been admitted to art school. Philip Roth asks what if Lindbergh had beaten Roosevelt. Each writer selects a small hinge, gives it a plausible push, and lets the world reorganize around it.

The pleasure of the form is not in the exotic setting. It is in the recognition. The reader knows what actually happened. What the writer offers is a mirror angled slightly differently, in which the outlines of our own present become visible for the first time.

What follows is an uchronia in the same spirit, set in a domain more prosaic than the fall of civilizations: the domain of physiological medicine, of vitamins and minerals and hormones, of what a body is entitled to receive when it is sick and what its physician is entitled to ignore. The hinges are small: a jurist’s opinion delivered in April 2005; a Wikipedia policy debate two years later; a WHO technical report annex issued at Geneva in 2006. The world they open up is not fantastical. Everything named below existed in some form on the day the divergence begins. The uchronic principle is respected throughout. No citation postdates the point of departure. No institution is invented that could not have arisen from the material already on the ground. No hero molecule is rehabilitated on evidence that did not already exist.

The reader is invited to hold two questions at once. First, is this plausible. Second, if it is plausible, what does it mean that we did not build it.

The departure point

By the late 1990s, three currents had come to a head at the same time.

The first was scientific. The evidence base for the therapeutic use of nutrients was thicker than any survey acknowledged. Bruce Ames on mitochondrial cofactors. Yukio Yamori on taurine and cardiovascular epidemiology across twenty-five populations. Guy Abraham on iodine. William Grant on vitamin D and geography. Joseph Kraft on insulin patterns during oral glucose tolerance. The literature was scattered across journals of second and third rank. Much of it was published by researchers who had spent their careers on the same molecule. It existed. It was reproducible. It was old enough that patents were no longer available on most of the molecules described.

The second was infrastructural. The internet had made the scattered literature searchable for the first time. A physician in Milan could read a case series published in Osaka in 1988. A patient in Winnipeg could find the Framingham cohort’s magnesium data. Wikipedia had launched in 2001 and was still, in the year of the divergence, a site of enthusiastic amateurs writing longform articles on obscure enzymes and vitamers, citing whatever they had read, with no formal doctrine on medical sources.

The third was political. The supplement industry was consolidating. Codex Alimentarius was negotiating a global framework for vitamin and mineral products. The European Union had adopted Directive 2002/46/EC, whose positive lists threatened to remove hundreds of forms of nutrients from national markets. Grassroots organizations, some of them well organized and some of them run by single individuals from home offices, were mobilizing against the framework. Media attention was rising.

In our timeline, each of these three currents was neutralized within a decade. In the uchronic timeline, each of the three encountered a small event that let it develop.

The first hinge: Advocate General Léger, April 2005

On 5 April 2005, the Advocate General of the Court of Justice of the European Communities delivered his opinion in the joined cases¹ challenging Directive 2002/46/EC on food supplements.² In our timeline the opinion is thoughtful, sides partly with the industry challengers, and is largely followed by the Court in its judgment of 12 July 2005: the directive is upheld, subject to procedural safeguards for the inclusion of new substances on the positive list. The practical effect is that hundreds of substances that were legally sold in the European Union before the directive are quietly delisted over the following decade. The burden of proof for inclusion falls on manufacturers of substances that no one owns the exclusive right to sell, which is to say, on no one at all.

In the uchronic timeline the Advocate General goes one step further. He accepts the industry argument that the directive constitutes a disproportionate restriction on the free movement of goods, given that the substances at issue are, on the whole, physiologically native to the human body and have long histories of safe consumption. He proposes that the positive list be construed as a rebuttable presumption rather than a closed catalogue, with member states retaining authority to permit substances demonstrated safe in national practice. The Court, on 12 July 2005, follows the opinion in substance.

The immediate practical effect is modest. A few dozen substances remain legally accessible in Denmark, Sweden, and the United Kingdom that in our timeline are pushed out. But the constitutional effect is large. The Court has connected the free movement of nutrients to a proportionality analysis grounded in bodily need. Bodily need has become, for the first time in European jurisprudence, a factor that regulators must weigh.

This is the first hinge. It does not solve any problem. It creates a legal grammar in which the problem can be discussed.

The second hinge: the Wikipedia debate, autumn 2007

By 2007, the English Wikipedia has grown to two million articles. Its medical content is chaotic and contested. A group of editors, largely physicians and pharmacists, proposes a formal policy on medical sourcing. In our timeline, the policy that emerges is called WP:MEDRS, medical reliable sources. It restricts claims about health to secondary sources in indexed journals, with a strong preference for systematic reviews, meta-analyses, and statements from major medical bodies. The effect on nutrient articles is rapid. Case series are removed. Clinical monographs are removed. The literature that had begun to be summarized in the enthusiastic amateur period is pruned back to what the major textbooks say, which is to say, to the recommended dietary allowances and the deficiency syndromes described in the 1930s.

The mood of the policy is captured in the response the co-founder of Wikipedia, Jimmy Wales, gives on 23 March 2014 to a petition from a coalition of holistic and integrative practitioners asking that the site’s medical policies be revised.³ “No, you have to be kidding me. Every single person who signed this petition needs to go back to check their premises and think harder about what it means to be honest, factual, truthful. Wikipedia’s policies around this kind of thing are exactly spot-on and correct. If you can get your work published in respectable scientific journals, that is to say, if you can produce evidence through replicable scientific experiments, then Wikipedia will cover it appropriately. What we won’t do is pretend that the work of lunatic charlatans is the equivalent of true scientific discourse. It isn’t.” The premise Wales does not examine is that “respectable scientific journals” and “replicable scientific experiments” are the appropriate filter for a class of physiological molecules whose evidence base is chronic, ecological, epidemiological, and mechanistic rather than acute, blinded, and randomized. The petition is answered. The premise is not addressed.

In the uchronic timeline the debate turns differently. The pivotal contribution comes from a librarian at the Cochrane Collaboration who intervenes on the talk page to point out that the proposed policy conflates two evidentiary questions. The first is whether a claim is likely to be true. The second is whether the mechanism used to answer the first question is appropriate to the domain. Randomized controlled trials, she observes, are the correct instrument for pharmaceutical interventions that admit of blinding, placebo control, and short-horizon endpoints. They are the wrong instrument, or at least an insufficient one, for nutrients that are physiologically native, chronically dosed, and evaluated over decades. To restrict medical content to sources that use the pharmaceutical instrument is to filter out the domain rather than the noise.

The policy that is adopted, WP:MEDSCOPE, distinguishes between claims about pharmaceutical interventions and claims about physiological molecules. It permits case series, ecological studies, and specialized monographs for the second category, provided they are attributed and cross-referenced. Its authors do not imagine that they are altering the epistemic landscape of medicine. They believe they are keeping Wikipedia honest about what it can and cannot answer.

The effect over the following decade is that nutrient articles on Wikipedia become substantive. A reader arriving at the article on magnesium finds a section on intracellular versus serum measurement, with citations to Rasmussen, Elin, and Costello. A reader arriving at potassium finds a section on tissue depletion in the presence of normal serum values. A reader arriving at iodine finds Ghent and Eskin on breast tissue uptake alongside the thyroid content. The material has always been there. It is now visible to anyone with a browser.

This is the second hinge.

The third hinge: WHO Technical Report Series 916, 2006 revision

The World Health Organization published Technical Report Series 916, Diet, Nutrition and the Prevention of Chronic Diseases,⁴ in March 2003. In our timeline the report is a compromise document. Its statements on saturated fat and refined carbohydrate are weakened during drafting under industry pressure. Its treatment of specific nutrient deficits, apart from iron and iodine in developing populations, is thin. It is cited widely and changed little.

In the uchronic timeline a revised edition is issued in 2006 under the same reference. The revision is driven internally by the nutrition division and externally by an unusual coalition of orthomolecular clinicians and cardiovascular epidemiologists, some of whom the Advocate General’s opinion has emboldened. The revised edition retains the original chapter structure but adds a fifth technical annex, dedicated to intracellular deficits of undetected clinical significance. The annex is short, twenty-eight pages. It lists magnesium, potassium, iodine, and B12 as candidate deficits routinely missed by serum panels, and it recommends that member states consider adding functional tests to standard practice.

The recommendation has no immediate force. Member states are under no obligation to act on WHO annexes. But the annex has been adopted at Geneva, with the signature of the Director-General, and it exists as a reference document that any national health authority can cite when it wants to justify a change of practice.

This is the third hinge. It is the smallest of the three in its immediate effect and the largest of the three in what it makes sayable.

The Vioxx acceleration

Merck withdrew rofecoxib from the market on 30 September 2004.⁵ The FDA’s David Graham estimated, in a Senate Finance Committee hearing that November, that the drug had caused between 88,000 and 139,000 excess cardiovascular events in the United States alone. The subsequent litigation and settlements ran for years. The public understanding of the Vioxx episode remained, in our timeline, largely private. It was a scandal about a specific molecule and a specific company. It did not become a scandal about the pharmacovigilance system that had let the molecule stay on the market for five years while internal data suggested harm.

In the uchronic timeline the Graham hearing lands differently, in part because a Wall Street Journal series in early 2005 develops the systemic angle at length. The FDA’s post-marketing surveillance division is restructured. A new agency is created, first as an office within HHS and then as an independent body, with the specific mandate of collecting real-world evidence on long-marketed molecules and on molecules that were never marketed at all because they are not patentable. It is called the Institute for Post-Approval Physiology. Its founding director is a Vioxx whistleblower.

The Institute’s first three years of work are unremarkable. It develops a database. It commissions a few observational studies. Then, in 2008, it publishes its first major report, on the mineral panel used in routine hospital admission. The report documents that serum magnesium and serum potassium miss the majority of clinically significant tissue depletion, that the fact has been known in the specialist literature for four decades, and that the panels have not been updated because the molecules involved are not owned by any manufacturer with an incentive to sponsor a change.

The class action that reshaped incentives

The Institute’s report is picked up in 2009 by a plaintiffs’ firm in Houston. The theory of the case is not that any manufacturer sold a defective product. The theory is that the class of hospital admissions between 1995 and 2009, whose serum panels showed normal magnesium and potassium in the presence of clinical presentations consistent with intracellular deficit, received substandard care as measured against a body of specialist literature published between 1960 and 1995. The defendants are the hospital systems and their laboratory contractors, alleged to have failed to update their panels in the presence of published knowledge that the panels were unreliable.

The theory is imaginative. It survives a motion to dismiss in 2010 largely because the WHO annex of 2006 is cited as evidence that the specialist knowledge had reached the level of a technical consensus.

The class action does not succeed at trial. It settles in 2012, without admission of liability, for a sum that is small on the scale of pharmaceutical litigation but that pays for a national program to introduce erythrocyte magnesium and 24-hour urinary potassium into standard hospital admission workups in the settling systems. Other systems follow, defensively.

By 2015 the intracellular mineral question is no longer a matter of specialist knowledge. It is a matter of routine practice.

Molecules from the cold

Lithium had been described in three American epidemiological studies by 1990. Dawson had shown in 1970 that Texas counties with higher lithium in the drinking water had lower rates of mental hospital admission.⁶ Schrauzer and Shrestha had shown in 1990 that twenty-seven Texas counties displayed inverse correlations between drinking water lithium and rates of homicide, suicide, and rape.⁷ The methodology was ecological, the effect sizes were substantial, and the mechanism was consistent with the known role of lithium in serotonergic and glutamatergic modulation at doses far below the pharmaceutical range used for bipolar disorder.

Nothing was done with the finding in our timeline. Low-dose lithium was not patentable. It was associated in the medical imagination with the high-dose bipolar drug, whose therapeutic window is narrow and whose side-effect profile is severe. The two dose ranges are separated by two orders of magnitude and are pharmacologically distinct, but the association was strong enough that the low-dose literature stayed in the specialist journals.

In the uchronic timeline the Institute for Post-Approval Physiology commissions a replication study in 2010, using the National Longitudinal Suicide Registry that a different Institute report had just recommended be built. The replication succeeds. A joint statement from the American Psychiatric Association and the American Public Health Association in 2013 recommends that municipal water authorities test for ambient lithium and consider fortification in counties below a defined threshold, on the same principle as fluoridation. The recommendation is controversial and is implemented in some jurisdictions and not others. Suicide rates in the counties that implement it decline over the following decade at rates consistent with the ecological literature.

Lithium is one molecule. In the same period, taurine is rehabilitated on the basis of Yamori’s WHO-CARDIAC data from the 1980s and 1990s, and iodine is rehabilitated on the basis of the Abraham and Brownstein clinical series and the older Japanese literature on tissue iodine sufficiency. None of these rehabilitations depends on new evidence. Each depends on the older evidence becoming legible again, which is to say, on the second hinge.

The Seventeen Years Task Force

In 2000, Balas and Boren had published in the Yearbook of Medical Informatics a synthesis of translational research showing that the average time between the publication of a clinical finding and its incorporation into routine practice was seventeen years.⁸ In 2011, Morris, Wooding, and Grant published in the Journal of the Royal Society of Medicine a review that confirmed the figure across seventeen further studies.⁹ The number was cited widely in translational medicine circles. It did not become the name of an institution.

In the uchronic timeline it does. In 2014, following a series of reports from the Institute for Post-Approval Physiology on specific translational failures, the United States Congress creates, over the objection of the pharmaceutical trade associations, an executive body called the Seventeen Years Task Force. Its mandate is to identify clinical findings whose translation into routine practice has been delayed by more than the mean, to publish annual lists, and to recommend remedial action. Its budget is small. Its convening authority is large.

The Task Force publishes its first list in 2015. It contains 340 findings, ranging from the delayed adoption of tight glycemic control in critically ill patients to the delayed abandonment of routine episiotomy. The intracellular mineral question is item 4 on the list.

The Task Force is not universally popular. It is understood within a year of its creation as an instrument that any actor within the medical system can use to embarrass any other actor. It is preserved because the embarrassment tends to move practice.

The Seventy Years Task Force

Within two years of the Seventeen Years Task Force publishing its lists, an obvious pattern has emerged. The translational lag for pharmaceutical findings clusters around the reported mean. The translational lag for findings involving non-patentable physiological molecules is much longer, and does not cluster around any single figure.

In 1956, E. B. Flink published in the Journal of the American Medical Association the first coherent clinical description of magnesium deficiency in humans.¹⁰ The syndrome was mapped. The unreliability of the serum measurement for its detection was already flagged. In the timeline the reader inhabits, the finding has not been operationalized in the seventy years since. The routine hospital admission panel still uses serum magnesium as its operative test. The routine primary care panel does not measure magnesium at all.

Similar histories can be told about protein, whose recommended dietary allowance was set at approximately 0.8 grams per kilogram per day in the 1940s on the basis of nitrogen balance studies in young men and has not been revised despite the coherent literature from Rand, Wolfe, and Volpi establishing its insufficiency for older adults and for populations recovering from illness. And about iodine, whose extra-thyroidal roles have been documented since the 1960s. In each case the delay is measured in generations.

In 2017 the Seventeen Years Task Force spins off a companion body dedicated to what it calls, in its foundational document, “endogenous molecules of chronic clinical relevance for which no commercial actor has a translational incentive.” The founding memorandum takes the emblematic case as its calibration point. Sixty years have elapsed since Flink’s magnesium description. A further decade of institutional work, its authors argue, will be needed to close the residual gap. The colloquial name that attaches to the new body is the Seventy Years Task Force. Its list, first published in 2018, is shorter than that of the parent body and its findings are older. The oldest items reach back to the 1930s.

The two Task Forces work in parallel. The pairing carries a message that neither body would have carried alone. The medical system has two distinct kinds of blind spot. One arises from the ordinary friction of institutional learning. The other arises from the systematic underinvestment in molecules that no one owns.

The patient as body with needs

The framework that develops in parallel with the Task Forces is legal and administrative, and its central novelty is a reframing of the patient. In our timeline the patient has become, over the past four decades, a person with preferences. Informed consent, patient autonomy, shared decision-making, patient satisfaction surveys: each of these instruments treats the patient as a locus of preference-formation and choice. The instruments are valuable. They are also incomplete.

The uchronic reframing does not displace the patient-as-preference. It adds, alongside it, the patient-as-body. A body has needs that are not preferences. It needs a certain concentration of magnesium in its cells whether the person in whom it is embedded prefers to have that concentration or not. It needs a certain concentration of thyroid hormone whether the person prefers that concentration or not. The needs are not private in the way that preferences are private. They are physiologically determinable, at least in principle, and their non-satisfaction produces measurable dysfunction.

The Council of Europe adopts, in 2016, a supplementary protocol to the Oviedo Convention on human rights and biomedicine, titled “Protocol on the Physiological Needs of the Patient.” The protocol enumerates a small number of rights: the right to have one’s physiological state measured with instruments appropriate to the molecule in question, the right to have the results explained, the right to receive replacement therapy for documented deficits, and the right to appeal a physician’s refusal to measure. The rights are conceived, in the drafting history, as extensions of the general right to health in Article 12 of the International Covenant on Economic, Social and Cultural Rights, and specifically of the language in General Comment 14 on functional accessibility.

Ratifications proceed slowly through the second half of the 2010s. By 2022 twenty-eight member states of the Council of Europe have ratified. The United States is not a party. A parallel American instrument, developed by a bipartisan working group at the National Academy of Medicine, is adopted in 2019 under the name Patient Physiological Rights Act. Its scope is narrower and its enforcement is thinner, but it establishes, for the first time in federal law, that a patient has an actionable interest in the measurement of physiological parameters that a competent physician would measure.

The reframing has a philosophical name in the uchronic literature. It is called somatic subjecthood, the recognition that a person is not only a bundle of preferences to be respected but also a body whose needs must be met if the preferences are to be meaningfully exercised. A hungry person cannot deliberate. A magnesium-depleted person cannot regulate his mood. A hypothyroid person cannot maintain the cognitive functioning that the doctrine of informed consent presupposes. Somatic subjecthood is offered, in the drafting history of the Oviedo protocol, as the physiological completion of the older doctrine, not as its replacement.

The uchronic present

By 2026 the medical system in the uchronic timeline has a shape that a visitor from our timeline would find both familiar and strange.

At the primary care level, an initial workup includes a first-line physiological panel of eighteen parameters, of which twelve are measured intracellularly or functionally: erythrocyte magnesium, twenty-four-hour urinary potassium and sodium, whole blood selenium, plasma zinc with copper, homocysteine as a functional B-vitamin marker, methylmalonic acid, ferritin with transferrin saturation, 25-hydroxyvitamin D, thyroid stimulating hormone with free T3 and free T4, morning cortisol, and fasting insulin. The panel is covered by the public payer in every country of the European Union and by Medicare and most private insurers in the United States.

At the specialist level, the Institute for Post-Approval Physiology maintains a national longitudinal registry that links prescribing data, laboratory results, and outcome data at the individual level, with strong privacy protections and with automated flagging of translational lag patterns.

At the educational level, medical curricula have been rebalanced. A student graduating from a European medical faculty in 2026 has completed a mandatory course of 180 hours in what is called physiological medicine, distinct from pharmacological medicine, covering the diagnosis and management of deficits and dysregulations in endogenous molecules. The course is taught by a joint faculty of nutritionists, endocrinologists, and clinical biochemists. The old orthomolecular literature, which in our timeline was pushed into the margins, is present in the curriculum, corrected where it was wrong and preserved where it was right.

At the administrative level, member states of the European Union have created ombudsman offices for physiological rights. The Quebec office, established in 2019, receives about four thousand complaints a year, of which the majority concern refusal to measure. The office does not adjudicate individual cases but produces annual patterns reports that feed the national committee on medical practice standards.

At the industrial level, the incentives are visibly different. A pharmaceutical company that develops a synthetic molecule for a condition treatable with an endogenous molecule must, since 2020, provide comparative data against the endogenous alternative in its marketing authorization application. The requirement has slowed the approval of several drugs whose superiority over inexpensive alternatives is not demonstrable. It has accelerated the approval of drugs whose superiority is clear.

At the informational level, Wikipedia’s medical content is substantive. The article on magnesium runs to sixty-eight thousand words. The article on iodine cites four hundred and thirty sources. The article on insulin has a dedicated section on measurement and its history, discussing Kraft, Reaven, Bikman, and the delayed integration of the fasting insulin measurement into routine panels.

None of this is utopia. Medical error persists. Iatrogenic harm persists. Diseases of aging remain incompletely understood. There is a lively literature complaining that the ombudsman offices are underfunded, that the Seventy Years Task Force lists are ignored more often than they are acted on, that the primary care panel misses parameters that a more ambitious version would include. What has changed is the frame within which these complaints are made. The complaints are made in the register of rights, against institutions that acknowledge the rights, on the basis of measurement that recognizes the underlying physiology.

Our timeline, seen from theirs

A visitor from the uchronic timeline, arriving in our world in 2026, would notice, among many other things, four features.

First, the panel would look impoverished. A patient presenting with fatigue, cognitive fog, muscle cramps, and mood dysregulation would receive a serum magnesium, a serum potassium, a thyroid stimulating hormone, and a complete blood count. Three of the four measurements would be near-guaranteed to return within reference range. The patient would be told, in the great majority of cases, that no cause for the symptoms had been identified.

Second, the informational environment would be shallow at exactly the points where a patient would need it to be deep. The Wikipedia article on magnesium deficiency does mention, in a single sentence within a longer physiology section, that plasma magnesium levels may underestimate deficiency. It does not treat the point as a central diagnostic problem, does not cite the seminal reviews on serum-tissue divergence, and does not draw the clinical conclusion that the routine serum test is inadequate for the question it is asked to answer. The article on hypokalemia notes that ninety-eight percent of body potassium is intracellular, then proceeds as if serum potassium were nonetheless a reliable index of total body stores, without drawing the diagnostic consequence that a normal serum value is consistent with severe tissue depletion. The article on iodine confines itself to the thyroid, with no mention of the extra-thyroidal roles documented in the older Japanese literature or in the Ghent and Eskin work on breast tissue uptake. The article on insulin discusses measurement in a paragraph on units and reference values, without naming Kraft, Reaven, or Bikman, and without addressing the fact that a fasting insulin measurement is absent from every standard metabolic panel. The specialist literature that in the uchronic timeline informs primary care would be, in our timeline, findable only by patients who already know the names of the researchers to search for.

Third, the industrial map would look distorted. A visitor asked to draw the incentives operating on the medical system would find that the great majority of arrows point toward patentable molecules, and that the small number of arrows pointing toward non-patentable molecules terminate in loose grassroots organizations rather than in institutions.

Fourth, and most tellingly, the visitor would notice that the patient in our timeline has no vocabulary for the demand that his body’s physiological state be measured. He can invoke his right to informed consent, his right to autonomous decision-making, his right to a second opinion. He cannot invoke a right that his doctor consider the possibility that his cells are depleted of a molecule the doctor has not thought to measure. The right does not exist because the institutional apparatus to receive the claim does not exist, because the hinges that would have built the apparatus were not pushed.

The limits of the counterfactual

There is no claim here that the counterfactual would have been better in every respect. Counterfactuals of this scale carry costs that historians of the imagined future can only speculate about. A more assertive Court of Justice in 2005 might have unsettled other regulatory areas. A weaker Wikipedia medical policy might have permitted the propagation of claims worse than those it filtered out. A more directive WHO annex might have produced iatrogenic harm in populations where over-supplementation is a real risk.

The claim is more modest. The hinges existed. They were small. The counterfactual is plausible. The physiological rights framework toward which it points is not utopian. It is a framework toward which existing law, existing evidence, and existing institutions could still be bent. The Seventeen Years Task Force does not exist in Washington. The Institute for Post-Approval Physiology has no premises in Rockville. The Council of Europe has not adopted its supplementary protocol. But the ICESCR is in force, General Comment 14 is in force, the Oviedo Convention is in force, and every clinical finding cited above was published on or before its stated date, in our timeline, in journals that anyone can consult.

The point of departure is April 2005. It is twenty-one years ago. The distance between the two timelines is smaller than it looks.