CEHF’s “Morgellons Arthritis” Article Does Not Say What Its Sources Say

A closer reading of CEHF’s citations reveals a four-day mouse experiment described as months of joint swelling, an animal study blurred into a human PTLDS diagnosis, and a 6 percent figure whose denominator changes as it moves through the Morgellons literature.

On August 13, 2026, the Charles E. Holman Morgellons Disease Foundation published an official statement warning Morgellons patients about unnamed critics whom it accused of failing to engage properly with peer-reviewed literature and of “nitpicking” researchers. The Foundation urged patients to rely on credible sources and peer-reviewed information while presenting its own work as rooted in scientific integrity.

That is a reasonable standard, and it should apply equally to CEHF. A standing article on the Foundation’s website, Understanding Morgellons Arthritis (archived May 18, 2026), provides an unusually useful test because its claims are specific and its scientific citations can be traced back to the original studies. The article, bylined to Mark Wilcox, was published in December 2022 and last modified February 20, 2023.

The stakes are greater than they would be for an informal blog post. After explaining what CEHF calls “Morgellons arthritis,” describing its proposed infectious mechanism and invoking peer-reviewed research as support, the same page asks readers to donate money so the Foundation can support research into treatments for Morgellons arthritis. That makes it particularly important to distinguish what the cited papers actually demonstrate from what later authors and CEHF say they demonstrate.

Following those citations all the way back to the original tables and experimental methods reveals several substantial discrepancies. Some appear on the CEHF page itself. Others developed earlier, as findings were restated and broadened within the peer-reviewed Morgellons literature before CEHF repeated them. Either way, a claim does not become accurate merely because a previous paper made the same generalization.

Where is the research establishing “Morgellons arthritis”?

The central problem appears before examining any individual experiment. CEHF describes a condition it calls “Morgellons arthritis,” telling patients that bacteria reach large joints, invade joint tissues, produce severe swelling and may eventually impair walking. Yet the scientific papers cited to explain this process do not directly characterize inflammatory arthritis in a Morgellons cohort.

The Embers study concerns experimental Borrelia burgdorferi infection and antibiotic treatment in rhesus macaques. The Jutras study concerns Lyme arthritis, persistent B. burgdorferi peptidoglycan in human synovial fluid and an acute mouse experiment. These studies may be relevant to hypotheses about musculoskeletal disease in Morgellons patients, particularly if an association with Borrelia is eventually established more broadly, but they are not studies demonstrating a distinct Morgellons-associated arthritis.

There is evidence that musculoskeletal complaints deserve serious investigation. The 2018 Fesler, Middelveen and Stricker study reported musculoskeletal symptoms in 93 percent of its 60 Morgellons patients, along with fatigue, insomnia and cognitive complaints. That makes musculoskeletal disease a legitimate research question rather than something that should be dismissed.

But “musculoskeletal symptoms” are not synonymous with objectively diagnosed inflammatory arthritis. Joint pain, muscle pain, stiffness, tendon complaints and true synovitis are clinically different findings. Fesler and colleagues did not establish the prevalence, pathology or diagnostic characteristics of an entity called Morgellons arthritis.

Mayne’s Australian cohort makes the arthritis claim harder to sustain

A 2014 Australian study by Peter Mayne is particularly important because it examined Morgellons findings and joint complaints within the same large clinical cohort. Mayne analyzed 500 consecutive patients who had either been referred or presented themselves because of concern about Lyme disease. His Table 2 reports Morgellons demonstrated by direct skin microscopy in 29 of those 500 patients, or 6 percent.

The same table also records joint findings, and reading it requires attention to which denominator is in use. Table 2 lists small-joint pain as 57 of 153—37 percent—where 153 is the subset for which that parameter was tabulated. The narrative reports the figure differently: “11% of patients presented with small joint pain.” Fifty-seven of 500 is approximately 11 percent, and Mayne appears to apply the same conversion to the row immediately above: the table lists dermopathy of any type as 55 of 153, or 36 percent, while the narrative states that skin problems were reported in 11 percent of the cohort. Mayne does not state either numerator in that narrative passage, so the correspondence is an inference from the arithmetic rather than an explicitly stated identity. But the pattern occurs in consecutive sentences, and the table percentages and narrative percentages plainly use different denominators.

Large-joint pain was recorded in only two of 153 patients, or 1 percent. The table’s monoarthritis row—labelled “documented illness in the United States”—records two cases in the full 500-patient cohort. Separately, Mayne writes that the two patients with large-joint pain had returned from living in the United States, where they had contracted and undergone treatment for their infection.

Mayne’s broader description is even more important. Early in the paper he contrasts Australian presentations with North American Lyme disease, where mono- or pauciarticular arthritis with extensive swelling is well recognized, and states that this arthritic presentation was not seen in the Australian borreliosis series.

This does not prove that Morgellons patients cannot develop Lyme arthritis or another inflammatory arthropathy. Mayne did not cross-tabulate the 29 patients with microscopy-demonstrated Morgellons against the patients with large-joint findings. We therefore cannot say whether any of the 29 Morgellons patients were among the two patients reporting large-joint pain. That missing cross-tabulation is precisely why the paper cannot characterize arthritis specifically among its Morgellons patients.

What it does show is that one of the major cohorts later invoked in support of the recurring 6 percent Morgellons figure did not demonstrate the widespread North American-style large-joint presentation described in CEHF’s article. In a 500-patient cohort where Morgellons and joint findings were both recorded, large-joint disease was rare, and Mayne explicitly stated that the characteristic North American arthritic presentation was not seen in the Australian series.

That makes direct research even more necessary. If “Morgellons arthritis” is meant to describe a real and recognizable manifestation, investigators should define a filament-confirmed Morgellons cohort, document objective synovitis, characterize which joints are involved, exclude competing causes, study imaging or synovial samples where appropriate, and determine whether arthritis correlates with Borrelia markers or disease activity. The research cited by CEHF has not yet done that.

Follow the 6 percent claim back to the denominator

CEHF tells readers that roughly 6 percent of Lyme patients develop Morgellons disease. Importantly, CEHF did not invent that generalization. It already appears in the peer-reviewed Morgellons literature. A 2020 classification-and-staging paper by Middelveen and colleagues states that two separate cohort studies demonstrated that Morgellons affects approximately 6 percent of Lyme disease patients.

Tracing that statement back to the original studies, however, shows that the two 6 percent figures do not have the same denominator.

The North American study by Fesler and colleagues examined 1,000 seropositive Lyme disease patients recruited from a single San Francisco medical practice specializing in the diagnosis and treatment of tick-borne diseases. Sixty met the investigators’ Morgellons criteria. In that study, 6 percent genuinely refers to 60 of 1,000 seropositive Lyme patients.

The Australian Mayne study reported something different. Its cohort consisted of 500 consecutive patients presenting because of concern about Lyme disease. Of those 500, 296 were considered clinically diagnosed with Lyme borreliosis and 273 tested positive for the disease, with Mayne explicitly noting that the laboratory-positive group was not simply a subset of the clinical group. Overall, 450 of the 500 patients—90 percent—had either clinical evidence for or laboratory proof of borrelial infection.

Mayne himself described his work as the first report of Morgellons incidence in a “large borreliosis population cohort.” With 450 of 500 patients considered to have clinical or laboratory evidence of borrelial infection, the later characterization of this as a Lyme or borreliosis cohort has a reasonable basis in Mayne’s own description of the study population.

But that is not the same thing as saying Mayne found Morgellons in 6 percent of seropositive Lyme patients.

Mayne’s Table 2 lists “Morgellons evidenced by direct skin microscopy” in 29 of the entire 500-patient cohort. His narrative likewise says that Morgellons was demonstrated by microscopy in 6 percent of the patients. The denominator used for that reported 6 percent is therefore 500.

The 2018 Fesler paper subsequently describes Mayne as having identified Morgellons in approximately 6 percent of seropositive Lyme patients. But Mayne’s original paper does not provide the information necessary to calculate such a percentage. It never reports how many of the 29 Morgellons patients belonged to the 273-patient laboratory-positive group.

That is not a semantic objection. A seropositive-subgroup percentage requires both a denominator and the number of Morgellons cases within that denominator. Mayne provides the denominator of 273 laboratory-positive patients, but he does not provide the corresponding Morgellons numerator. Consequently, the paper cannot yield a figure of “6 percent of seropositive Lyme patients.”

By 2020, Middelveen and colleagues summarized Mayne and Fesler together as two cohort studies showing Morgellons in approximately 6 percent of Lyme disease patients. CEHF appears to be repeating that published interpretation rather than independently creating it. But following the citation chain back to Mayne’s original Table 2 shows that one of the two percentages did not originate as a seropositive-subgroup estimate.

The distinction is important because it illustrates how scientific claims can broaden through repetition even within peer-reviewed literature. Mayne reported 29 of 500 selected clinical patients. A later paper restated that as approximately 6 percent of seropositive Lyme patients. A subsequent paper grouped it with a separate North American cohort and described both as demonstrating that Morgellons affects approximately 6 percent of Lyme disease patients. CEHF then repeated the broader formulation for patients.

The proper response is not to accuse CEHF of inventing the number. It is to follow the citation chain back to the original denominator.

Two cohorts, but not two completely independent research programs

The similarity between the Australian and North American percentages remains noteworthy. A comparable observation in two geographically different clinical populations deserves more attention than a single isolated result, and it would be unfair to pretend that the Australian cohort contributes nothing.

At the same time, these were not two completely disconnected research programs. Mayne’s acknowledgments specifically thank Marianne Middelveen for reviewing the manuscript. His references on Morgellons include Middelveen and Stricker publications, and Mayne himself had previously co-authored a Morgellons filament paper with Middelveen, D.G. Kahn and Raphael Stricker.

Mayne also disclosed that from 2014 he served, without compensation, as a medical adviser to Australian Biologics, one of the principal laboratories used for Borrelia testing in the study. The study period ran from October 2010 through February 2014, so the disclosed advisory role began around the time data collection concluded rather than throughout the study period. The relationship does not invalidate the results, just as manuscript review does not make two datasets identical, but both facts are relevant context when later papers describe the Australian and North American studies in terms that may sound like wholly independent replication by unrelated research groups.

The fairest conclusion is therefore narrower than either extreme. Two selected clinical cohorts in different countries produced superficially similar 6 percent Morgellons figures. One figure was 60 of 1,000 seropositive Lyme patients. The other was 29 of an entire 500-patient cohort that Mayne himself regarded as predominantly borreliosis-related. The similarity is a research signal worth testing independently, but it does not establish a population incidence or prove that approximately 6 percent of Lyme patients generally “develop” Morgellons.

The demographic claim does not resolve the causal question

CEHF’s 6 percent sentence also states that middle-aged Caucasian women appear to be affected more frequently. The demographic observation is worth noting, but it does not provide the etiological support the surrounding infectious narrative might lead readers to assume.

The 2012 Kaiser Permanente/CDC investigation by Pearson and colleagues found a strikingly similar demographic profile despite reaching a very different conclusion about cause. Its case-patients had a median age of 52, and 77 percent were female and 77 percent Caucasian. Yet investigators did not identify a common underlying medical condition or infectious source in the population they studied.

This was a CDC-funded investigation conducted within Kaiser Permanente Northern California, with CDC epidemiologists involved in study design, data collection and analysis, the decision to publish and preparation of the manuscript. I have previously discussed important limitations of that investigation in Why Morgellons Studies Disagree, including questions about its case definition and the relatively small number of participants whose nonbiopsy material underwent dedicated analysis.

Those limitations matter, but they do not make the demographic comparison disappear. A Borrelia-oriented Morgellons literature and a major investigation that failed to identify a common infection both describe cohorts dominated by middle-aged white women. The demographic pattern therefore does not distinguish between the competing etiological models and cannot, on its own, establish a tick-borne cause.

Association becomes established tick transmission

CEHF does not present the infectious model as one hypothesis among competing interpretations. It tells readers that Morgellons is a tick-borne illness caused by the same bacteria responsible for Lyme disease and that Morgellons spreads through black-legged tick bites. Those statements are considerably stronger than saying that multiple Morgellons studies have reported associations with Borrelia.

The distinction matters because even supportive Morgellons papers frequently use the language of association. The Fesler study describes B. burgdorferi infection as associated with Morgellons and notes the limited epidemiological evidence addressing that relationship. Its authors called for additional research into the clinical and genetic determinants involved.

The Pearson investigation, meanwhile, found no common infectious cause in its study population. Its methods and case definition have been criticized by Morgellons researchers and advocates, and those criticisms deserve serious consideration. But positive association studies alongside a contested negative investigation describe a literature that remains unresolved. They do not make epidemiological proof of tick transmission unnecessary.

Future research may establish a causal Borrelia pathway convincingly. Until that happens, an organization committed to peer-reviewed science should distinguish between the model it considers best supported and a mechanism that has actually been demonstrated at the population level.

The Jutras citation is where the CEHF article breaks down

The clearest factual problem in CEHF’s article involves a 2019 Proceedings of the National Academy of Sciences study by Brandon Jutras and colleagues examining B. burgdorferi peptidoglycan. CEHF describes mice receiving bacterial cell-wall material, developing severe swelling, failing to clear spirochetal particles from their joints and remaining swollen for “several months.”

The mouse experiment did not show that.

The animals were never infected with B. burgdorferi. Researchers intravenously administered purified, sonicated B. burgdorferi peptidoglycan to BALB/c mice to test whether the cell-wall material itself could trigger inflammation. No living spirochetes were introduced, there was no infection to treat, and the mouse experiment was not designed to determine whether animals could clear spirochetes from their joints.

The injected bacterial material did provoke inflammation. The mice developed bilateral ankle edema and inflammatory changes characterized by the authors as acute tenosynovitis. But this was an acute experiment measured over hours rather than months.

Arthritis scores were assessed at 24, 48, 72 and 96 hours. Half of the animals in each group were euthanized on day two for examination and the remaining animals were euthanized on day four. By 96 hours, every experimental mouse had reached the study’s protocol endpoint.

The experiment therefore could not have demonstrated that those same mice remained swollen for several months. There were no study animals remaining months later in which such a finding could have been observed.

The experiment also did not determine that the mice were unable to clear “spirochete particles” from their joints. No living spirochetes had been introduced into the mice, and clearance was not what the mouse experiment measured. The persistence finding came from a separate component of the Jutras paper involving human Lyme arthritis patients.

Jutras and colleagues detected B. burgdorferi peptidoglycan in 32 of 34 synovial-fluid samples from patients with Lyme arthritis, including patients who had undergone antibiotic treatment. The investigators proposed persistent bacterial antigen as one mechanism that could contribute to continuing synovial inflammation after treatment.

The paper therefore contains two separate findings answering two different questions. The human samples supplied evidence that peptidoglycan could persist in Lyme-arthritis synovial fluid. The mouse experiment demonstrated that purified peptidoglycan could provoke an acute inflammatory response.

CEHF combines those findings into one narrative in which the mice themselves fail to clear bacterial material and remain swollen for months. The human persistence result has been welded onto the acute mouse experiment, producing an account the experimental design could not have generated.

This is not a subtle disagreement over how to interpret a biomarker. The chronology alone resolves the months-long-swelling claim. A four-day mouse experiment cannot demonstrate several months of subsequent swelling in the same animals.

The same Jutras paper complicates CEHF’s persistent-infection explanation

Immediately before discussing the Jutras research, CEHF explains continuing symptoms by saying that the underlying B. burgdorferi infection persists and therefore inflammation persists. Yet the paper it then cites provides a reason not to treat persistent inflammation as automatically equivalent to persistent viable infection.

Jutras and colleagues detected peptidoglycan in treated Lyme arthritis patients and proposed that persistent bacterial antigen could contribute to synovitis after antibiotic therapy, including after antibiotics eradicate the pathogen. That does not prove viable Borrelia can never persist following treatment, nor was the study designed to settle that larger controversy. It demonstrates the narrower point that persistent inflammation and persistent bacterial material are not themselves proof of an ongoing viable infection.

https://www.youtube.com/watch?v=LfNe3WSMB4s

That distinction has obvious biological importance. Persistent organisms, retained antigen, immune dysregulation and residual tissue injury are different mechanisms with different implications. A patient-facing article should explain those distinctions rather than treating continuing inflammation as straightforward evidence that the infection itself must still be active.

The Embers macaque experiment is not a human PTLDS diagnosis

CEHF also cites a 2017 study by Monica Embers and colleagues examining post-treatment persistence in rhesus macaques. Ten animals were infected with B. burgdorferi through tick feeding, and five received doxycycline for 28 days beginning approximately four months after infection. Investigators subsequently looked for evidence of persistence using xenodiagnosis, molecular methods, immunostaining, mouse bioassays and an in-vivo culture approach.

The article’s timing claim is not the strongest problem. Xenodiagnosis was performed approximately three months and again seven to eight months after treatment. The problematic step is CEHF’s movement from an experimental persistence result to language implying a PTLDS diagnosis.

These were rhesus macaques in an experimental infection model, not human patients being clinically diagnosed with post-treatment Lyme disease syndrome. The study investigated a biological question potentially relevant to PTLDS: whether organisms or evidence of organisms could persist after antibiotic treatment. That is not equivalent to diagnosing the experimental animals with the human syndrome.

The results were also more complicated than a simple statement that persistent viable organisms had been cleanly demonstrated. Embers and colleagues interpreted the combined findings as evidence of persistent, intact and metabolically active B. burgdorferi, but conventional cultures did not yield motile spirochetes. The authors reported no statistically significant differences between treated and untreated animals in the numbers testing positive by PCR, xenodiagnosis or RT-PCR, and exactly seven spirochetes were identified across approximately 100 tissue sections from treated and untreated animals.

Other findings, including transcription detected using an in-vivo heart-tissue culture procedure and immunofluorescent detection in some specimens, were interpreted by the investigators as support for viability. These findings make the experiment scientifically interesting, but they also demonstrate why reducing it to a simple proof of persistent infection in human PTLDS goes beyond what the experiment established.

The work was formally challenged in the peer-reviewed literature. In 2018, Gary Wormser and colleagues published Critical Analysis of a Doxycycline Treatment Trial of Rhesus Macaques Infected with Borrelia burgdorferi, addressing the Embers experiment and a companion histopathology paper by Crossland and colleagues. The critique argued that viable organisms had not been convincingly demonstrated in treated primates and raised questions about whether the doxycycline exposure achieved in the model was comparable to human treatment.

Acknowledging that criticism does not require accepting it as the final word. The defensible conclusion is narrower: a small nonhuman-primate study produced findings its authors interpreted as post-treatment persistence; aspects of its methodology and interpretation were challenged by other Lyme researchers; and the significance of those findings for human PTLDS remains contested.

Why the fundraising appeal matters

The donation request changes the context in which these scientific claims should be evaluated. There is nothing inherently inappropriate about CEHF raising money for Morgellons research. Direct investigation of objective joint abnormalities in carefully defined Morgellons patients could answer important questions that existing studies have not answered.

The issue is the sequence. The page first presents “Morgellons arthritis” and its proposed mechanisms with substantial certainty, uses scientific citations to give that explanation authority, and then asks affected patients to finance research into treatments for the condition. A clearer article would distinguish between what has already been demonstrated and what CEHF believes deserves future investigation before asking readers to fund that investigation.

The strongest case for funding research into joint disease in Morgellons patients would not be to imply that “Morgellons arthritis” has already been characterized. It would be to point out how little direct research exists and explain why that gap urgently needs to be filled.

Following the citation chain is not “nitpicking”

CEHF’s August 13 statement argues that Morgellons patients should be wary of people who fail to engage seriously with peer-reviewed literature. On that basic principle, there should be little disagreement. But engaging with peer-reviewed literature requires more than seeing that a paper has been published and repeating the interpretation attached to it by another author.

It requires opening the study, reading the methods and following the denominator back to the table. It means determining whether a percentage describes all 500 selected clinical patients, a 273-patient laboratory-positive subgroup or a representative Lyme disease population. It also means noticing when the same figure appears as 37 percent of a tabulated subset in a table and 11 percent of the cohort in the narrative.

It means distinguishing a human synovial-fluid result from an acute mouse experiment and a four-day observation period from several months. It means distinguishing persistent antigen from persistent viable infection and an experimental macaque persistence model from a human PTLDS diagnosis.

It also requires following citations backward when one peer-reviewed paper summarizes an earlier study. The Mayne-to-Fesler-to-Middelveen chain provides an unusually clear example. Mayne reported microscopy-demonstrated Morgellons in 29 of 500 patients. He also considered the cohort overwhelmingly borreliosis-related, with 450 patients having clinical or laboratory evidence of infection. But he did not report how many of the 29 Morgellons patients were among the 273 who tested positive.

Fesler subsequently described Mayne as finding Morgellons in approximately 6 percent of seropositive Lyme patients. That specific calculation cannot be reconstructed from Mayne’s paper because the necessary subgroup numerator is absent. Middelveen and colleagues then summarized Mayne and Fesler as two studies demonstrating that Morgellons occurs in approximately 6 percent of Lyme disease patients. CEHF later repeated the broader formulation.

This is precisely why peer review should not end scrutiny of a citation chain. A secondary paper can restate a primary paper imprecisely, and later publications can repeat the restatement until it begins to look like the original result. Returning to the source table is not hostility toward science. It is how the claim is checked.

The same standard should apply when evidence cuts against the interpretation we favor. It means acknowledging the Pearson investigation’s failure to identify a common infectious source while also discussing its methodological limitations. It means taking the Embers persistence findings seriously while acknowledging negative and equivocal components and the published critique. It means recognizing that musculoskeletal complaints are common in some Morgellons cohorts while refusing to equate those complaints automatically with a defined inflammatory arthritis.

These are not trivial distinctions. They determine what the evidence actually establishes.

CEHF should correct the article

The Charles E. Holman Morgellons Disease Foundation has asked the Morgellons community to value rigorous research and peer-reviewed evidence. Its Understanding Morgellons Arthritis article should be judged according to the same standard.

The article describes “Morgellons arthritis” without citing research that directly characterizes inflammatory arthritis in a Morgellons cohort. The recurring 6 percent claim also becomes broader as it passes through the literature, with Mayne’s original whole-cohort percentage eventually restated as a seropositive-subgroup percentage that his paper does not provide enough information to calculate.

The article states Borrelia causation and tick transmission with greater certainty than the contested epidemiological literature warrants. It invokes an experimental macaque persistence study using language associated with human PTLDS while omitting the formal published criticism of that work.

Most seriously, CEHF’s description of the Jutras paper is incompatible with the experimental design. The mice were never infected with spirochetes. The mouse experiment did not test whether they could clear spirochetal particles. The persistent-antigen finding came from human Lyme arthritis samples. The experimental animals reached their protocol endpoints by 96 hours, yet CEHF tells readers that their swelling continued for several months.

Morgellons science does not need claims stronger than the underlying evidence. There are legitimate findings worth pursuing: musculoskeletal symptoms are common in some studied Morgellons patients, Borrelia has been associated with Morgellons in several publications, Lyme arthritis offers plausible mechanisms worth directly testing, persistent bacterial antigens are biologically interesting, and the question of post-treatment Borrelia persistence remains contested and worthy of continued research.

If every known manifestation of Lyme disease is simply imported into Morgellons after first assuming that Morgellons is Lyme-related, however, the reasoning becomes circular. Lyme arthritis becomes “Morgellons arthritis,” and by the same logic Lyme carditis could become “Morgellons carditis” or neuroborreliosis “Morgellons neuroborreliosis” without those Morgellons-specific entities ever being directly characterized.

The better approach is empirical. Determine whether rigorously defined Morgellons patients have objective inflammatory arthritis. Determine how often it occurs, which joints are affected and whether it resembles established Lyme arthritis. Test whether joint findings correlate with Borrelia evidence, persistent peptidoglycan, skin disease severity or other measurable biological markers. Compare affected patients with appropriate controls.

Those studies could materially advance Morgellons research. They are also more scientifically valuable than treating their conclusions as known before the studies have been performed.

Scientific integrity is not demonstrated by deciding in advance who should or should not be trusted. It is demonstrated by representing evidence faithfully, distinguishing primary data from later interpretation, acknowledging uncertainty and contradictory findings, and correcting claims when tracing a citation back to the original study shows that the source does not say what we have been told it says.

CEHF has asked the Morgellons community to apply that standard to its critics. The Foundation should apply it to its own website.

---

Primary sources discussed

• Charles E. Holman Morgellons Disease Foundation — Understanding Morgellons Arthritis (live page) — Wayback Machine snapshot, May 18, 2026
• Charles E. Holman Morgellons Disease Foundation — Official Statement, August 13, 2026
• Fesler MC, Middelveen MJ, Stricker RB. Clinical evaluation of Morgellons disease in a cohort of North American patients. Dermatology Reports. 2018.
• Mayne PJ. Clinical determinants of Lyme borreliosis, babesiosis, bartonellosis, anaplasmosis, and ehrlichiosis in an Australian cohort. International Journal of General Medicine. 2014.
• Middelveen MJ, et al. Classification and Staging of Morgellons Disease: Lessons from Syphilis. Clinical, Cosmetic and Investigational Dermatology. 2020.
• Jutras BL, et al. Borrelia burgdorferi peptidoglycan is a persistent antigen in patients with Lyme arthritis. Proceedings of the National Academy of Sciences. 2019.
• Embers ME, et al. Variable manifestations, diverse seroreactivity and post-treatment persistence in non-human primates exposed to Borrelia burgdorferi by tick feeding. PLOS ONE. 2017.
• Wormser GP, et al. Critical Analysis of a Doxycycline Treatment Trial of Rhesus Macaques Infected with Borrelia burgdorferi. Diagnostic Microbiology and Infectious Disease. 2018.
• Pearson ML, et al. Clinical, Epidemiologic, Histopathologic and Molecular Features of an Unexplained Dermopathy. PLOS ONE. 2012.
• MorgellonsSurvey.org — Why Morgellons Studies Disagree