Letter to the President of the American Academy of Family Physicians (AAFP) on misdiagnosis (T1D misdiagnosed as T2D)

Many of the adults who are misdiagnosed as having Type 2 diabetes, when they actually have Type 1 diabetes, are misdiagnosed by family physicians. Unfortunately, there is little awareness in the medical community of adult-onset Type 1 diabetes, and the American Academy of Family Physicians guidelines Diabetes Mellitus: Screening and Diagnosis (2016) present incorrect information about adult-onset Type 1 diabetes. Below is a letter that I wrote in 2018 to the President of the American Academy of Family Physicians and the authors of Diabetes Mellitus: Screening and Diagnosis.

Dear Dr. Munger:

The vision of the American Academy of Family Physicians (AAFP) is to transform health care to achieve optimal health for everyone.  The mission of AAFP is to improve the health of patients, families, and communities by serving the needs of members with professionalism and creativity.

Those are admirable vision and mission statements, but they are not being applied to people with adult-onset Type 1 diabetes, who are too often incorrectly diagnosed by Family Physicians who are provided incorrect information and inadequate guidelines by AAFP.

First, let’s look at what respected endocrinologists say about the epidemic of adult-onset Type 1 diabetes being misdiagnosed as Type 2 diabetes, which is an altogether different disease:

• Dr. Irl Hirsch, a professor of medicine at the University of Washington and former chairman of the Professional Practice Committee of the American Diabetes Association, has described the misdiagnosis of adult-onset Type 1 diabetes as an “epidemic.” He blames a lack of awareness and insufficient medical training about diabetes as the reason why so many patients fall through the gaps. The problem is particularly frustrating because there is a simple blood test that can check for antibodies associated with Type 1 diabetes, which could easily be used in diagnosing patients.  Dr. Hirsch quickly orders autoantibody tests when he is unsure of the correct diagnosis (Footnote 1). In Dr. Hirsch’s busy clinic, they see approximately one new person per week who is misdiagnosed (diagnosed as Type 2 when the person actually has Type 1).
• Dr. Robin Goland, co-director of the Naomi Berrie Diabetes Center at Columbia University Medical Center in New York, says "Most of my [adult-onset Type 1 patients] have been misdiagnosed as having Type 2.  Once the right diagnosis is made the patient feels much, much better, but they are distrustful of doctors and who could blame them (Footnote 2)?"
• Dr. Regina Castro, an endocrinologist at the Mayo Clinic, estimates that anywhere between 10 to 30 percent of adults diagnosed with Type 2 diabetes each year may in fact have Type 1 diabetes.  Exactly how many adults with Type 1 diabetes are misdiagnosed each year in the United States is hard to track—in  2015, the year for which data is most recently available, 1.5 million adults were diagnosed with diabetes, which is how, even taking the conservative end of Castro’s estimate, one gets to the possibility that tens of thousands if not hundreds of thousands go misdiagnosed each year.  “It is under-recognized and more prevalent than we think,” says Dr. Castro (Footnote 3).
• Dr. Steve Edelman, the founder and director of Taking Control of Your Diabetes (TCOYD), had this to say about misdiagnosis, “Latent autoimmune diabetes in adults (LADA) is the most misdiagnosed area in diabetes. We [Steve Edelman, Jeremy Pettus, Tricia Santos] are primarily adult endocrinologists and see tons of folks who are misdiagnosed.” (Personal communication)

Probably the most notable person to be misdiagnosed is UK Prime Minister Theresa May.  At the age of 56, Prime Minister May was misdiagnosed as having Type 2 diabetes; about 6 months after the initial misdiagnosis, she was correctly diagnosed as having Type 1 diabetes (Footnote 4).

Next, let’s look at what the AAFP guidelines Diabetes Mellitus: Screening and Diagnosis (2016) say about Type 1 diabetes, and then see how the guidelines stand up against the evidence-based facts.

AAFP 2016: Type 1 diabetes is caused by autoimmune destruction of the islet cells of the pancreas, and onset is typically in childhood.

Fact:  Type 1 diabetes is not a childhood disease.  The incidence of autoimmune diabetes is about equal in almost all age groups; adult-onset Type 1 diabetes is more common than childhood-onset Type 1 diabetes (Footnotes 5 & 6).

AAFP 2016:  Type 1 diabetes patients typically present with an acute onset of symptoms.

Fact:  Onset of Type 1 diabetes can be rapid or slowly progressive, with those with slowly progressive Type 1 diabetes far outnumbering those with rapid onset.  The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus’s definition of Type 1 diabetes clearly encompasses all autoimmune diabetes, regardless of age (“Type 1 diabetes results from a cellular-mediated autoimmune destruction of the beta-cells of the pancreas. In Type 1 diabetes, the rate of beta-cell destruction is quite variable, being rapid in some individuals (mainly infants and children) and slow in others (mainly adults)”).  From The Type 1 Diabetes Sourcebook (ADA/JDRF 2013), “Adult [Type 1] patients can vary greatly at presentation, from a more acute picture, with DKA and marked hyperglycemia, to a more gradual course such as is often seen in latent autoimmune diabetes in adults (LADA).” 

AAFP 2016:  Some patients with latent autoimmune diabetes in the adult or Type 2 diabetes may have certain autoantibodies present making these tests less specific.

Fact:  By definition, those with LADA are autoantibody positive, and by definition of ADA and WHO, LADA is Type 1 diabetes.  Also by definition of ADA and WHO, "Although the specific etiologies of [Type 2] diabetes are not known, autoimmune destruction of beta-cells does not occur."   In other words, if a person has been diagnosed with Type 2 diabetes but is autoantibody positive, the person has been misdiagnosed and in fact has Type 1a diabetes.  This makes autoantibody testing more specific, not less.

AAFP 2016:  Despite these concerns [AAFP:  that autoantibodies are not specific to Type 1 diabetes], the American Association of Clinical Endocrinologists recommend routine confirmation of Type 1 diabetes using autoantibody testing.

Fact:  As established above, AAFP’s arguments that autoantibody tests are “less specific” are not true.  In fact, autoantibody testing is used to establish the correct diagnosis for people with Type 1 diabetes, and is highly predictive.  Diabetes researcher R. David Leslie MD says, "The best way to identify autoimmune diabetes is to assess diabetes-associated autoantibodies [GAD, IAA, IA-2, ZnT8], which represent the only relevant categorical trait."  An American Diabetes Association position statement affirms, “Consider measurement of pancreatic autoantibodies to confirm the diagnosis of Type 1 diabetes (Footnote 7).”

AAFP 2016:  Additional testing to determine the etiology of diabetes is not recommended.  Additional research is required to determine whether further testing [autoantibody testing] to classify the etiology of diabetes improves patient outcomes.  In the meantime, additional testing is not routinely recommended.

Fact:  There is an epidemic of misdiagnosis (T1D misdiagnosed as T2D), and misdiagnosis can result in the rapid onset of diabetic complications and even death due to diabetic ketoacidosis (DKA).  The Type 1 Diabetes Sourcebook (ADA/JDRF 2013) makes it clear that Type 1 and Type 2 diabetes are fundamentally different diseases (Footnote 8).  Without a doubt, a correct diagnosis would improve patient outcomes.  I challenge you to name even one disease where an incorrect diagnosis improves patient care or outcomes; I don’t believe there is one.    

For a patient with symptoms of diabetes, their first source of care probably is a family practice physician/primary care doctor.  Sadly, many of those doctors rely on the incorrect assumption that an adult with an elevated blood glucose must have Type 2 diabetes; autoantibody testing is a way to get beyond that assumption and actually provide a correct diagnosis.  If a person is positive for any one autoantibody and has a fasting blood glucose above 125 mg/dl, by definition the person has Type 1 autoimmune diabetes.  An article in Diabetes Spectrum, whose authors are associated with a university pharmacy college and who routinely encounter misdiagnosed patients, states, “It is imperative to establish distinct practice guidelines for the diagnosis and treatment of LADA [adult-onset Type 1 diabetes] and for providers to recognize this clinical scenario as one that requires special testing (autoantibody testing) to establish a proper diagnosis and thus improve patient safety and treatment efficacy.  Incorrect diagnosis can delay proper treatment (insulin therapy), exposing patients to potential adverse effects from ineffective Type 2 drugs, slowing progress toward normoglycemia, and ultimately increasing the risk of long-term complications (Footnote 9).”

In an editorial in American Family Physician, Jeff Unger MD states, “Family physicians care for most patients in the United States with Type 2 diabetes and, therefore, should be aware that approximately 10% of these patients have LADA [slowly progressive Type 1 diabetes] (Footnote 10).” Dr. Irl Hirsch states that “in the primary care setting, adult-onset Type 1 diabetes is not on the physician’s radar or in their bandwidth; they see so many Type 2s and have so little time.”  For primary care doctors who believe they have never misdiagnosed a person, I would suggest they consider what often happens in these situations: the frustrated patient seeks a second opinion, finally receives an accurate diagnosis and appropriate care, and never returns to the primary care provider who misdiagnosed them in the first place.  In this situation, the patient loses trust in the healthcare system, and the doctor never knows the original diagnosis was wrong.

The adult presentation of Type 1 diabetes does not present a challenge for the diabetes classification system—it is simply Type 1 diabetes.  The problem is that some in the medical community cannot let go of the myth that “Type 1 diabetes is a childhood disease,” and many in the medical profession have a difficult time coping with the fact that the majority of new-onset Type 1a diabetes is actually seen in adults.  If it were simply accepted that new-onset Type 1 diabetes occurs at all ages and is most commonly seen in adults (Footnote 11), then there undoubtedly would be fewer misdiagnosed cases.  However, we do know that scientific communities can be surprisingly resistant to new ideas or data that do not fit the accepted model, in this case the “juvenile diabetes” model.

Medical doctors already know how to effectively treat Type 1 diabetes in children and teenagers; that excellence in care should also be applied to adults with new-onset Type 1 diabetes.  When a child is diagnosed with Type 1 diabetes, the medical community springs to action on the child’s behalf, because Type 1 diabetes is a serious, life-threatening disease.  Kids who are diagnosed with Type 1 diabetes are shown great compassion, and the disease is acknowledged to be profoundly life-altering; adults deserve that same consideration.  When people with adult-onset Type 1 diabetes are finally correctly diagnosed and correctly treated with exogenous insulin (Footnote 12), they express great relief, and are able to reclaim their lives.

I would encourage the American Academy of Family Physicians to be at the forefront of addressing this epidemic of misdiagnoses—that would truly transform health care and achieve optimal health in alignment with your vision and mission.

Footnote 1: Slate magazine, March 6, 2018.  “Type 1 Diabetes is No Longer Just for Kids.”  Amy Mackinnon, author.

Footnote 2: Wall Street Journal, August 7, 2012.  “Wrong Call:  The Trouble Diagnosing Diabetes.”

Footnote 3: Slate magazine, March 6, 2018.  “Type 1 Diabetes is No Longer Just for Kids.”  Amy Mackinnon, author.

Footnote 4: July 13, 2016 Medscape article “New UK Prime Minister Brings Spotlight to Type 1 Diabetes.” Simon Heller, MD, and Irl B. Hirsch, MD, authors.

Footnote 5: S.R. Merger, R.D. Leslie, and B.O. Boehm.  “The broad clinical phenotype of Type 1 diabetes at presentation.”  Diabetic Medicine 2012.

Footnote 6: Miriam E Tucker, “Half of All Type 1 Diabetes Develops after 30 Years of Age.”  Medscape, September 20, 2016.  [Note that this study found that 50% of people diagnosed with Type 1 diabetes were diagnosed older than 30 years (and the study subjects only went up to 60 years, so greater than 50% are diagnosed over the age of 30)]. The Medscape article is based on data presented September 16, 2016, at the European Association for the Study of Diabetes (EASD) 2016 Annual Meeting by Dr Nicholas JM Thomas, of the Institute of Biomedical and Clinical Science, University of Exeter Medical School, United Kingdom (later published as Frequency and phenotype of type 1 diabetes in the first six decades of life: a cross-sectional, genetically stratified survival analysis from UK Biobank. Lancet Diabetes Endocrinol. 2018 February; 6(2): 122-129).

Footnote 7: “Type 1 Diabetes Through the Life Span: A Position Statement of the American Diabetes Association.”  Diabetes Care, June 16, 2014.  Jan L. Chiang, M. Sue Kirkman, Lori M.B. Laffel, and Anne L. Peters.

Footnote 8: “The pathophysiology of the two diseases [T1D and T2D] differ on a basic pathophysiologic level such that T1D is marked by insulinopenia while T2D is characterized by obesity, hyperinsulinemia, insulin resistance, and relative insulinopenia.” Page 104, The Type 1 Diabetes Sourcebook.

Footnote 9: “Recognizing and Appropriately Treating Latent Autoimmune Diabetes in Adults (LADA)” (Diabetes Spectrum 2016 Nov; 29(4):249-252).

Footnote 10: The first medical journal article that described the 10% of Type 2s who are autoantibody positive and in fact have Type 1 diabetes was “Clinical and pathogenic significance of pancreatic-islet-cell antibodies in diabetics treated with oral hypoglycaemic agents.”  The Lancet, Volume 309, No. 8020, p1025-1027, 14 May 1977.  Numerous subsequent studies, including the UKPDS, have confirmed the 10% figure.

Footnote 11: Miriam E Tucker, “Half of All Type 1 Diabetes Develops after 30 Years of Age.”  Medscape, September 20, 2016. 

Footnote 12: The Type 1 Diabetes Sourcebook (ADA/JDRF 2013) states, “For those presenting acutely as well as those presenting more indolently, starting insulin is the mainstay of therapy.”