Science-Based Medicine: Sometimes “n = 1” Is Enough

Posted on by

Yves here. KLG gives an example of how a single case is challenging widely-held views about how Alzheimers afflicts the brain, through amyloid plaques acting as cognition kudzu (note that there are competing schools of thought, perhaps a topic for a later post).

Here, a patient had a genetic disorder that led to early onset of Alzheimers. Yet his case progressed roughly twenty years behind the normal schedule, despite imaging and his autopsy showing his brain to be a plaque-y mess. This led to an examination of “Why was he different?” which in turn looks to be providing important new insights into how Alzheimers might be checked.

By KLG, who has held research and academic positions in three US medical schools since 1995 and is currently Professor of Biochemistry and Associate Dean. He has performed and directed research on protein structure, function, and evolution; cell adhesion and motility; the mechanism of viral fusion proteins; and assembly of the vertebrate heart. He has served on national review panels of both public and private funding agencies, and his research and that of his students has been funded by the American Heart Association, American Cancer Society, and National Institutes of Health.

As has been noted here before, Alzheimer’s disease (AD) is one of the most frightening diagnoses imaginable, unforgettably expressed in the words of Dr. Alois Alzheimer’s index patient Auguste D, “Ich habe mich verloren – I’ve lost myself.” (registration at The Lancet required, but worth it if you are so inclined).  Every time those of us of a certain age forget the name of a person or pause while searching for a word in conversation or in our writing, a mental twinge immediately follows.  Research over the past 50 years has identified many of the concomitants of AD, but both the cause and agent [1] of the disease have remained elusive, despite a predominant focus on the amyloid hypothesis of AD (here and here).

The current consensus is that the development of amyloid plaques of pathological fragments of amyloid precursor protein (APP) and Tau tangles [2] in the brains of those with AD may be both cause and agent of AD.  A paper published in Nature on 15 May 2023 casts some doubt on this hypothesis: Resilience to autosomal dominant Alzheimer’s disease in a Reelin-COLBOS heterozygous man (open access).  There is a lot to unpack in this title, but understanding this research is no more difficult than understanding a “derivative,” and I am not referring to the derivative function from the first semester of calculus.  From the Abstract:

We characterized the world’s second case with ascertained extreme resilience to autosomal dominant Alzheimer’s disease (ADAD).  Side-by-side comparisons of this male case and the previously reported female case with ADAD homozygote for the APOE3 Christchurch (APOECh) variant allowed us to discern common features.  The male remained cognitively intact until 67 years of age despite carrying a PSEN1-E280A mutation.  Like the APOECh carrier, he had extremely elevated amyloid plaque burden and limited entorhinal (area of the temporal lobe of the brain involved in memory and navigation) Tau tangle burden…He was heterozygous for a rare variant in RELN (H3447R, termed COLBOS after the Colombia–Boston biomarker research study), a ligand that like apolipoprotein E binds to the VLDLr and APOEr2 receptors. RELN-COLBOS is a gain-of-function variant showing stronger ability to activate its canonical protein target Dab1 and reduce human Tau phosphorylation in a knockin mouse.  A genetic variant in a case protected from ADAD suggests a role for RELN signaling in resilience to dementia.

A few definitions required for navigating this research report:  Autosomal dominant means that a genetic disease is caused by a mutation in only one of the two copies of the gene present in each cell.  Most common genetic diseases are autosomal recessive (e.g., Tay-Sachs disease), in which one copy of the gene, especially if the gene encodes an enzyme, is enough to produce a normal phenotype.  Individuals with one copy of the mutation are carriers; the child of two carriers has a 25% chance of getting the disease.  PSEN1 (Presenilin-1) is the gene that encodes a subunit of an enzyme that cleaves proteins, including amyloid precursor protein (APP) and others.  PSEN1-E280A carries one and only one mutation in which amino acid-280 (of 463 amino acids) is changed from glutamic acid (E) to alanine (A).  From the same OMIM link [3]: “With this and other missense mutations in the PS1 gene, increased levels of amyloid beta-peptides ending at residue 42 are found in plasma and skin fibroblast media of gene carriers.  A-beta-42 aggregates readily and appears to provide a nidus for the subsequent aggregations of A-beta-40, resulting in the formation of innumerable neuritic plaques.”  These neuritic plaques are the hallmark of AD.  A large kindred of 1,200 individuals in Colombia that carries this presenilin mutation suffers from ADAD, which is characterized by mild cognitive impairment at age 44 and dementia by age 49; dementia in non-familial AD usually occurs much later in life.  Dr. Alzheimer’s first AD patient, Auguste D, died at the age of 51.

The subject of this report had the RELN-COLBOS-H3447R variant (histidine-3447 replaced by arginine) and carried the PSEN1-E280A mutation.  He remained cognitively intact until age 67, or more than 20 years after others who carry this presenilin mutation present with AD.  By age 72 his language skills had deteriorated, and by age 73 he required assistance with basic activities consistent with moderate dementia.  He died at 74 of aspirated pneumonia.  His relatives agreed to donate his brain for postmortem analysis of his neuropathology.

But prior to his death he underwent neuroimaging analysis at Massachusetts General Hospital (MGH) as part of the Colombia-Boston biomarker research study.  His cortical amyloid-beta (A-beta) plaque burden was higher than younger impaired carriers of this kindred (PSEN1-E280A).  The heat maps (red = higher plaque burden) and dot plots of Figure 1a,b [4] illustrate this very well.  Regarding the tau tangle burden, it was much lower in the in the RELN-COLBOSpatient than the younger patient with typical MCI (mild cognitive impairment) as shown in Figure 1c [middle panel (top, our patient) with blue-green rather than red-orange middle panel (bottom typical younger patient with MCI)].  The conclusion from these data is that “in this patient and in the APOECh homozygote case, protection against ADAD dementia occurred even in the face of high amyloid burden.

Which led to the question: What is different about this patient?  Genomic analysis of samples from this patient led to the identification of a mutation in the RELN gene, which encodes the protein reelin.  This variant was found only in this individual and his sister, who also had late-onset cognitive decline associated with ADAD, although not as late as her brother.  The focus on reelin was natural because this protein is involved in Tau phosphorylation (the addition of phosphoryl groups to Tau; phospho-Tau has been implicated in AD and may be a clinical marker of disease progression).

If the reelin hypothesis is correct, the next step is to consider the normal role of reelin in the brain.  Defects in reelin have been shown to be the cause of the gait abnormalities in the reeler mouse (first identified in the late 1950s, this mouse is much less agile than the typical mouse and has other neurological defects).  The reeler mouse has been a model for understanding the development of the central nervous system (CNS) for the past 50 years.  This mutant mouse lacks reelin, an extracellular protein shown to be essential for normal wiring of the brain.  Reelin has roles in axon guidance, development of synapses in the CNS, network activity in the CNS, synaptic plasticity, and learning and behavior.  A substantial literature associates reelin with schizophrenia, autism, and AD.  The protein Dab1 (Disabled-1) is a well-characterized neuronal signaling component downstream of reelin, here and here (paywalled).  Thus, reelin “fits” as a protein that could modulate the effects of the RELN-COLBOS variant.

But how does it fit?  This paper in Nature shows that at the molecular and cellular levels the RELN-COLBOS mutant results in the increased levels of the downstream component Dab1 in normal (primary) neurons from mouse brain (Figure 2) and reduces Tau hyperphosphorylation and preserves motor function in the mouse model of tauopathy (Figure 4).  This research also showed that the RELN-COLBOS mutant binds better to receptor molecules on the cell surface than the wild-type protein (Figure 2 and Figure 3).  These data are frankly “too much information,” but based the work done in my previous research laboratory, the techniques are well suited to this research.  The structural diagrams of the protein show how the mutant reelin is thought to interact with heparin.  Heparin is a type of extracellular GAG (glycosaminoglycan) that serve as co-receptors of reelin on the cell surface.

These data were subsequently confirmed with a knock-in mouse that expresses the RELN-COLBOS variant.  The gain-of-function mutation associated with this mutant recapitulated the molecular and cellular data.  Dab1 levels were increased and there was a difference between males and females that corresponded to the differential protection of RELN-COLBOS against ADAD in the male versus the female case.  Crossbreeding of the RELN-COLBOS mouse with a transgenic mouse that expresses a human Tau mutation leading to the accumulation of tau tangles and neuronal loss was consistent with the hypothesis that RELN-COLBOS protects against Tau-mediated brain pathology.  Male knock-in mice expressing the RELN-COLBOS mutation had a significant reduction in Tau phosphorylation compared control mice lacking the mutation.  Further studies are needed, of course, but the data support the overarching hypothesis that RELN-COLBOS is a gain-of-function mutation responsible for resilience to Tau-mediated pathology in regions of the brain associated with AD.  Note that a gain-of-function mutant would be active (approximately “co-dominant”) in the presence of the normal gene, which would retain wild-type activity/function.

So, to summarize this research:

  • The RELN-COLBOS variant rendered the male carrier resistant to cognitive impairment associated with autosomal dominant Alzheimer’s disease (ADAD) associated with the presenilin mutant PSEN1-E280A. Nevertheless, the comparative neuropathology was determined in a very small sample size, and the results cannot be considered definitive.  They have generated hypotheses that are likely to be fruitful, however.
  • A female sibling showed similar resistance to cognitive decline, but with a different clinical history that included a severe head injury. Other research has identified RELN as a candidate gene associated with AD pathology in cognitively healthy individuals.  Moreover, DAB1 variants have been linked to AD risk in APOE4 homozygotes, which further links the RELN/DAB1 pathway to AD.
  • Biochemical and cellular studies demonstrated a molecular basis for the involvement of reelin in AD in the presence of the RELN/COLBOS
  • Defects in signal transduction from the outside of the cell to the inside are often the molecular cause of disease. This includes cancer, for which a first step is dysregulation of the cell cycle in which cells ignore extracellular cues to stop dividing, and other developmental diseases.  It is reasonable to view defects leading to AD to be related to dysregulated neuronal signaling, perhaps in the pathway implicated here.
  • Amyloid pathology and tauopathy are complex in AD, but the genetic variants studied here suggest that the ERC (entorhinal cortex of the temporal lobe) is spared of tauopathy. The ERC is essential for memory and is probably the first part of the brain affected in subclinical AD.  Thus, the protection of the ERC from tauopathy associated with the RELN/COLBOS variant may be a key finding that stimulates further research in the origins of AD and prevention of AD progression.
  • This research also demonstrates that individuals with the mutation responsible for ADAD and the RELN/COLBOS variant are resistant to AD progression even in the presence of high levels of amyloid plaques that are the postmortem hallmark of AD.

What is meant here by science-directed medicine?  Clinical observation followed by experimental studies that elucidate a biological mechanism consistent with the pathology observed.  In this case a mutant in one man was found to be associated with resilience against disease progression in ADAD, and subsequent studies in vitro (literally “in glass”) of isolated components, in cells, and in an animal model, i.e., in vivo, strongly supported the initial hypothesis.

How does this differ from evidence-based medicine (EBM)?  That is a difficult question, but while EBM always has its genesis in a biological hypothesis, sometimes the accepted practice becomes divorced from the original theory and evidence for various and sundry reasons.  This is not to say that science-directed medicine stands alone, however.  The original research must be followed by unambiguous confirmation at the level of the lab bench before translation to the bedside, “bench to bedside” in the current vernacular.  Sometimes this transition is fraught, as in the case of Ignaz Semmelweis, whose observations on maternal mortality from childbed fever were ignored, until they could not be. This Semmelweis Reflex (Basically: That is not the way to think about it!) is often difficult to overcome.  Sometimes this transition from biomedical science to clinical practice is subverted.  Sometimes an intervention is assumed to be benign without a convincing demonstration that this is the case.  We may have been dealing with an example of this as a result of a crisis-driven rush since early 2020.  In the case of the amyloid hypothesis of Alzheimer’s disease, there is ample reason to believe the Semmelweis Reflex is still very strong.

Which leads us once again to the realization that correlation is not (necessarily) causation, in this case that amyloid plaques cause AD.  The amyloid hypothesis has dominated AD research for a long time, but progress in understanding the cause(s) of the disease and improvements in clinical care have been halting to nonexistent.  Could another mechanism be responsible?  Of course, and in the absence of apparent progress under the amyloid hypothesis, this seems likely.  Although it is much too early to state that on the basis of this paper on ADAD, the amyloid hypothesis of AD has been refuted, this research does show that amyloid plaques are not always a concomitant of AD.  Defective signaling from reelin to Dab1 to Tau is the potential mechanism identified here.

The major question here is whether “n = 1” is really enough to represent a valid finding rather than remain an anecdote?  In my view, the answer is yes in this case (in which n = 2, including the primary subject’s unstudied sister).  One patient has provided solid evidence of a biological mechanism likely to be important to the etiology and progression of AD.  He was identified in a cohort study, which is essential in determining the underlying mechanisms of an inherited disease.  The effort led by Nancy Wexler to identify the gene associated with Huntington’s disease (HD) depended on a 10-generation cohort in Venezuela that was afflicted with familial HD.  The functions of the protein huntingtin are still obscure, but as a triplet repeat disease HD is an exemplar of a brain disorder with the same autosomal dominant genetic mechanism as ADAD.  Interestingly, as more has been learned of HD, environmental and genetic factors contribute to the progression of disease.  This is similar at the population level to AD, and it has been suggested that something in our environment may lead to AD.

We should also note that if scientists had not been supported in research on the mutation that produces the reeler mouse, none of the work reported here could have been completed with these results.  You never know what a result will mean at the time of discovery, with the rare exception exemplified by this one-page paper from 1953 in Nature.  Abraham Flexner, the founding director of the Institute for Advanced Study, wrote of The Usefulness of Useless Knowledge in 1939.  He was right then and remains so 84 years later.  If we want to make scientific progress in biomedicine and beyond, we as a society and polity would do well to remember this.

And finally, in a relevant update that is nevertheless somewhat off-topic, the strange case of Dr. Sylvain Lesné is still ongoing, one presumes after nearly a year.  This has been covered previously, and it has put one of the major results underpinning the amyloid hypothesis of AD (>2,300 citations to date) under a dark cloud [5] that is something more than the Semmelweis Reflex.  The research described here may provide a ray of sunshine from another direction.


[1] Agent and cause are frequently conflated when discussing a disease.  Most people, including scientists, when asked to name the cause of tuberculosis will reply “Mycobacterium tuberculosis.”  In my view it is more correct to say that M. tuberculosis is the agent of the disease.  Tuberculosis has been known for thousands of years dating back to ancient Egypt.  But the primary cause of tuberculosis, which was the leading cause of death in the 19th century when TB was known as the “White Plague,” was overcrowding in inner city tenements along with malnutrition and generally poor living conditions of an immiserated working class and those who fell below that mark.  Two paintings by the Venezuelan artist Cristobal Rojas here and here illustrate the plague that was TB (both jpg).  Which is not to say that the fortunate do not also get TB.  The young physician Walker Percy contracted TB in Bellevue Hospital.  The novelist Walker Percy subsequently became the author of several works, both fiction and nonfiction, that describe the second half of the 20th century very well.

[2] Tau (tubulin associated unit) is a microtubule-binding protein.  Microtubules can be understood as “railroad tracks” in the cell that provide the path to move proteins other intracellular cargo in both directions from a “microtubule organizing center” near the nucleus to the cell periphery.  This trafficking is essential in all cells but especially those in the brain, which can have relatively long distances to move things.  Dysregulation of microtubules in neurons will have a deleterious effect.

[3] OMIM (Online Mendelian Inheritance in Man) is a project begun by Victor McKusick of Johns Hopkins University School of Medicine.  It is the accessible successor of the 4-volume The Metabolic and Molecular Bases of Disease, 8thedition (2000), which was first published in 1960 as the one-volume The Metabolic Basis of Inherited Disease. the go-to source for genetic diseases and I recommend it as Wikipedia for smart medical students, a few of whom listen.  The clinical features, for example ADAD, can be useful to all.

[4] As with the previous NC post on the pancreatic cancer vaccine, I will link directly to the open-access figures of this paper.  They are more legible at the links than when inserted into the text, and sometimes the ability to enlarge them is useful, for example in visualizing the blue dot representing the APOEch homozygote carrier (i.e., both chromosomes have this mutation) in Figure 1b.  They are much more usable on a screen larger than that of a smart phone, however, if you want to dig deeper.

[5] “14 July 2022, Editor’s Note: The editors of Nature have been alerted to concerns regarding some of the figures in this paper. Nature is investigating these concerns, and a further editorial response will follow as soon as possible. In the meantime, readers are advised to use caution when using results reported therein.”  This seems to be taking a long time, but others may have a different view.

Print Friendly, PDF & Email


  1. Roger Blakely

    So it isn’t my Diet Pepsi that’s making me stupid? I can stop worrying about the aluminum and aspartame?

  2. Terry Flynn

    Excellent article. Whilst not equivalent, similar thought went into “n-of-1” trials which were a thing back during my PhD before the turn of the millennium. Essentially the patient is “conceptually/philosophically” the sample. If you are able to produce reproducible theory-based outcomes based on “stimulus x” then you may be onto something.

    However, you must, essentially, administer stimuli (groups of stimulus) in a particular way (typically using statistical design theory) if you are to get real insights into the question of whether “A causes B”.

    Some of the article goes above my pay grade but if I understand correctly, it contributes to a branch of health and medicine that could be revolutionary….. Though it has a hurdle to cross in that it typically is expensive to conduct.

  3. Susan the other

    Reading Michio Kaku’s new book on quantum computing. I’d say the future looks good for personalized medicine in all its variations. It will never simplify the complexity but it will specify the details. Funny now that we know a little science, that things do not get simpler, they just get more and more complex.

  4. Laura in So Cal

    My Mom has Alzheimer’s, and it is horrible and heartbraking. I have read before that the amyloid plague might be a reaction to Alzheimers vs. the cause of it. I always like seeing more research.

    For many diseases that we don’t really understand from dementia to autism, each patient is n=1. We are currently trying to find medication to deal with psychosis and what works for some patients doesn’t work for others.

  5. Objective Ace

    a single case is challenging widely-held views about how Alzheimers afflicts the brain, through amyloid plaques

    Hasnt this hypothesis been debunked and the data it was based on proven fraudulent for awhile now?

    Its more a matter of getting the research and drug companies who have invested billions into it to acknowledge and change course, which additional data like this is certainly welcome

    1. les online

      The year i turned 40 i began using a bronchi dilater (asthma puffer) for the developing regular bouts of asthma…Just after turning 40 i began to have occassional spelling problems – word spelling being a previous talent…A couple of years later i began to suspect more than a corelation, and later read that the (long-term) use of such medication can impair memory…
      Initially the diagnosis was asthma, then chronic airways limitation (CAL), and subsequently, chronic obstructive pulmonary disease (COPD)…With the ‘progression’ has been the steady deterioration of memory…
      at 75, my memory retrieval system is faulty – i can sense the name of the person, meaning of a word, etc, but the ‘connection’ is very weak…My usual excuse for ‘memory failure’ is “it’ll come to me – in a week.”
      As for Genetics, much of it is just as fraudulent as Virology !

  6. ChrisPacific

    N=1 is enough when it’s a counterexample. If you are trying to formulate a theory of causation, like the theory of gravity, it’s expected to hold all of the time.

    For example, a functioning heart is necessary for us to live. We know this because when people’s hearts stop beating, they die (in the absence of medical intervention). They don’t die at a 99.9% confidence interval. We don’t see one out of every one thousand heart attack victims leading a happy normal long life without a beating heart. They just die. If we were to find even one person remaining alive and functional without a heartbeat, then that would be an indication that something outside our understanding was going on that merited further investigation. It would not just be a rare outlier to be dismissed out of hand.

Comments are closed.