Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Apr 1.
Published in final edited form as: Diabetes Res Clin Pract. 2022 Mar 14;186:109828. doi: 10.1016/j.diabres.2022.109828

Friedreich’s Ataxia Related Diabetes: Epidemiology and Management Practices

Jaclyn Tamaroff 1, Anna DeDio 1, Kristin Wade 1, McKenzie Wells 3, Courtney Park 3, Karla Leavens 1,2, Christian Rummey 4, Andrea Kelly 1,2, David R Lynch 3,5, Shana E McCormack 1,2
PMCID: PMC9075677  NIHMSID: NIHMS1793942  PMID: 35301072

Abstract

Aims:

Friedreich’s Ataxia (FRDA) is a progressive neuromuscular disorder typically caused by GAA triplet repeat expansions in both frataxin gene alleles. FRDA can be complicated by diabetes mellitus (DM). The objective of this study was to describe the prevalence of, risk factors for, and management practices of FRDA-related DM.

Methods:

FACOMS, a prospective, multi-site natural history study, includes 1,104 individuals. Extracted data included the presence of DM and other co-morbidities, genetic diagnosis, and markers of disease severity. We performed detailed medical record review and a survey for the subset of individuals with FRDA-related DM followed at one FACOMS site, Children’s Hospital of Philadelphia.

Results:

FRDA-related DM was reported by 8.7% of individuals. Age, severe disease, and FRDA cardiac complications were positively associated with DM risk. FRDA-related DM was generally well-controlled, as reflected by HbA1c, though diabetic ketoacidosis did occur. Insulin is the mainstay of treatment (64-74% overall); in adults, metformin use was common and newer glucose-lowering agents were used rarely.

Conclusions:

Clinical factors identify individuals at increased risk for FRDA-related DM. Future studies should test strategies for FRDA-related DM screening and management, in particular the potential role for novel glucose-lowering therapies in preventing or delaying FRDA-related cardiac disease.

Keywords: Friedreich’s Ataxia, Mitochondrial diabetes

1. Introduction:

Friedreich’s Ataxia (FRDA) is a progressive neuromuscular disorder affecting 1 in 50,000 individuals. FRDA arises in an autosomal recessive inheritance pattern, primarily due to a GAA triplet repeat expansion of both frataxin gene alleles.1 The length of the GAA expansion on the least-affected allele determines residual expression of the frataxin protein. Thus, the longer the length of the shorter GAA repeat expansion, the more rapidly progressive the disease is likely to be.2 A minority of individuals, approximately 4%, have a GAA triplet repeat expansion on one allele and a point mutation on the other.3 These genetic variants in frataxin lead to decreased amounts of frataxin, a protein critical for iron-sulfur cluster assembly. Frataxin deficiency in turn leads to tissue-specific deficits in multiple cellular processes, most notably mitochondrial respiration.4

FRDA is a multi-system disease, with ataxia being the most overt clinical feature. Cardiomyopathy is common and a frequent cause of premature death.5 FRDA can also be complicated by diabetes mellitus (DM).1,6 Friedreich’s ataxia related diabetes (FRDA-related DM) has been reported in 5 to 40% of children and adults with FRDA, with the largest prior study reporting a prevalence of 6.4% (52/811) 7,8 Impaired fasting glucose or impaired glucose tolerance have been estimated to occur in nearly half of adults with FRDA (49% in a study with n=41).6 Also, individuals with FRDA-related DM have worse functional status compared to individuals with FRDA without DM.8 Despite the clinical relevance of FRDA-related DM, no evidence-based clinical practice guidelines for screening or management currently exist.

Previous studies have suggested that older age, non-ambulatory status, and longer GAA repeat lengths are all risk factors for FRDA-related DM.8 Despite these insights, the mechanisms contributing to FRDA-related DM pathogenesis remain incompletely understood. Both impaired pancreatic insulin secretion and peripheral insulin resistance have been invoked, perhaps each the manifestations of tissue-specific mitochondrial impairment related to frataxin deficiency.9 Also, few previous investigations have focused on response to treatment in individuals with FRDA-related DM or relationship to other co-morbidities such as cardiac disease. Given these knowledge gaps, the most appropriate approaches for screening and management of FRDA-related DM across the lifespan remain unknown. The availability of a large, well-described FRDA natural history cohort has enhanced our ability to explore these important questions.

The objective of this study was therefore to describe the prevalence, diagnosis, and management of FRDA-related DM. To accomplish this goal, we leveraged the FRDA Clinical Outcomes Measures Study (FACOMS), an international, multi-center, prospective natural history study of FRDA.8,10 To enrich our understanding, we obtained additional information on the individuals with FRDA-related DM followed at our FACOMS site. Using this approach, we were able to provide more detail regarding screening practices, complications such as diabetic ketoacidosis (DKA), and the use of diabetes technology and new glucose lowering agents.

2. Subjects:

Participants had a genetic diagnosis of FRDA and were enrolled in FACOMS, a prospective, multi-site, longitudinal, observational study with annual visits. FACOMS enrolled 1,104 participants children and adults between October 2003 and April 2020 across 15 international sites. Nearly all (n=1,099) answered the question regarding the presence or absence of DM (Figure 1). Three analyses were conducted involving these participants, including a retrospective review of FACOMS data (natural history cohort), an electronic medical record (EMR) review of the subset with DM followed at the Children’s Hospital of Philadelphia (CHOP) (CHOP focused EMR review), and a patient survey (CHOP patient survey cohort). Subjects could be included in one, two, or all three analyses (Figure 1).

Figure 1:

Figure 1:

Open in a new tab

Participants included in each cohort.

All studies were approved by the Institutional Review Board, and informed consent and/or assent were obtained, as appropriate. Interim analyses of this cohort have been reported previously,10,11 including interim analysis of FRDA-related DM in smaller cohorts.7,8

3. Materials and Methods:

Three analyses were conducted including “natural history cohort”, “CHOP focused EMR review”, and “CHOP patient survey cohort”.

3.1. Natural history cohort:

Participants reported whether they had DM (n=1,099 who reported DM status of 1,104 participants) during prospective, longitudinal, observational study visits as part of FACOMS.

Diabetes & other comorbidities:

Participants were considered to have DM if they reported DM at any visit. Self-reported hemoglobin A1c (HbA1c) and use of insulin were collected at each visit for the subset with FRDA-related DM. Implausible HbA1c values, less than 4.0% (20mmol/mol) and greater than 15.0% (140mmol/mol), were excluded. Overall DM control was determined by mean of the median within-participant HbA1c.

Genetic diagnosis, including GAA repeat length and the presence of a point mutation was collected. In individuals with two GAA repeat expansions, GAA repeat length on the allele with fewer triplet expansions was included as this correlates to disease severity.2 For analyses including GAA repeat lengths, only individuals with GAA expansions on both alleles were included. Age at most recent visit, age of onset of FRDA symptoms, and age of DM diagnosis were collected. Self-reported co-morbidities were collected including scoliosis and cardiac disease (hypertrophic cardiomyopathy, dilated cardiomyopathy, or arrythmia).

Measurements of FRDA disease severity:

At each visit, the modified Friedreich’s Ataxia Rating scale (mFARS) score was calculated.12,13 The mFARS score is a measure of disease severity, with higher scores indicating more severe disease. The mFARS from the most recent FACOMS visit was used in this analysis because this version of the FARS score most directly assesses functional ability.12

3.2. CHOP focused EMR review:

Of the 449 individuals with FRDA enrolled at the CHOP FACOMS site, 44 reported having DM, and their information was abstracted.

Electronic medical record abstraction:

Age and BMI at DM diagnosis, evaluation of DM auto-antibodies, DKA history, and medication use were obtained via review of the EMR for those with FRDA-related DM and evaluable data at CHOP (Figure 1).

3.3. CHOP patient survey cohort:

The patient survey was distributed to 473 unique e-mail addresses as part of a larger CHOP-focused survey. 133 individuals responded, 11 of whom had DM.

Data collection:

Survey data were collected prospectively using a HIPAA-compliant secure REDCap (Research Electronic Data Capture) database.14 Individuals reported height and weight. Details about DM included how they were diagnosed with DM, if they had antibodies tested, the type of DM with which they were told they had, family history of DM, history of ketones, and history of hospitalizations for DM. Medication use was reported including dose and type of insulin, use of diabetes technology, difficulty with glucose monitoring or insulin injection, use of other antidiabetic agents and associated side effects of these agents. (Supplementary material).

DM specific quality of life (QOL):

Age-appropriate validated questionnaires were used to assess DM specific QOL. Adults completed the Diabetes Distress Scale (DDS).15 The DDS asks 17 questions related to DM distress and participants rate their answers from 1, “not a problem”, to 6, “a very serious problem”. Individuals aged <18 years completed the Problem Areas in Diabetes Child/Teen (PAID-C/T) scales which ask questions related to DM distress.16,17

3.4. Overall statistical analyses:

The three cohorts were described using standard summary statistics. For the natural history cohort, Wilcoxon rank sum test was used to compare continuous variables and Chi squared test was used to compare categorical variables between participants with versus without DM. Candidate clinical risk factors for FRDA-related DM including age, sex, age of FRDA symptom onset, the presence of scoliosis, the presence of cardiac disease, and the presence of a point mutation or GAA repeat length on the least-affected allele were evaluated first using univariable and then using multivariable logistic regression analyses. Analyses were performed using STATA v16.1 (StataCorp LP, College Station, TX). A two-sided p value < 0.05 was considered statistically significant.

4. Results:

4.1. Natural history cohort:

Participant characteristics.

In FACOMS (N=1,104), self-reported FRDA symptom onset occurred at a median of 11 years of age (IQR 7-16). Participants enrolled in FACOMS across the life span, with median age of enrollment in young adulthood, and were followed typically for several years (median, 4 visits per person; IQR 2-7, range 1-17, with 76% of individuals who had at least 2 visits, and 61% at least 3 visits), The median age at the most recent FACOMS visit was 27 years (IQR 19-39) (Table 1). Similar to prior reports, 94.9% of individuals with FRDA in FACOMS have two GAA triplet repeat expansions and 5.1% have a point mutation. In this cohort, 85% had scoliosis, and 68% had cardiac disease related to FRDA (dilated cardiomyopathy, hypertrophic cardiomyopathy, and/or the presence of an arrythmia).

Table 1:

Characteristics of study participants.

Characteristic Natural History (n=1,104 with and without DM) CHOP focused EMR review (n=35, all with DM) CHOP Patient Survey (n=11, all with DM)
Age, y 27 (19-39) 33 (21-39) 37 (22-47)
Adults (% ≥18, n) 77% (847) 83% (29) 82% (9)
Sex (% female, n) 51% (563) 46% (16) 36% (4)
Genetics
 % with point mutation (n) 5% (56) 23% (8) 46% (5)
 GAA repeat length, nt 687 (500-800) (n=990) 898 (690-932) (n=26) 817 (500-901) (n=6)
Age of FRDA symptom onset, y 11 (7-16) 9 (7-14) 8 (3-16)
FRDA duration, y 14.5 (8.9-22.5) 21.7 (14-30.2) 20.5 (13.7-30.2)
mFARS score 56 (43-69) 72 (50-80) 70.5 (49-77.5)
BMI
 BMI (Z-score), age <18y (n) −0.3 (−1.3-0.6) (n=166) 0.6 (−0.7-1.7) (n=6) −0.7 (−3.2-1.9) (n=2)
 BMI (kg/m2), age ≥18y (n) 22.6 (19.6-26.3) (n=177) 25.5 (21.4-29.2) (n=26) 22.3 (21.1-27.3) (n=9)
Open in a new tab

Descriptive data are presented as median (IQR) for continuous variables and percentage (n) for categorical variables. BMI Z-scores for individuals under 18 years old were generated based on CDC 2000 reference values. GAA repeat length indicates length, in nucleotides (nt) on the least affected allele and excludes those with point mutations. Natural history cohort: age, FRDA duration, mFARS, BMI are based on the most recent natural history study visit. CHOP focused EMR review: age, FRDA duration, mFARS are based on the most recent natural history study visit. BMI is based on most recent available BMI on chart review. CHOP patient survey cohort: mFARS and FRDA duration are based on the most recent natural history study visit and age and BMI are based on survey self-report. For individuals that height was not documented via survey or was implausible, height was obtained from chart review.

Diabetes diagnosis and glycemic control.

Nine percent (96/1,099 of survey respondents) reported having DM at any FACOMS visit. DM was usually diagnosed in early adulthood (median, 27 years; IQR 17-41; n=84/96 with self-reported DM diagnosis details). Sixty-nine of 96 individuals reported at least one HbA1c value. Overall, DM appeared to be adequately controlled according to non-FRDA specific clinical standards, with mean self-reported HbA1c of 7.1% (SD 1.1) (54mmol/mol) (SD 12). Nearly three-quarters of individuals with DM (64/86 individuals who responded) reported using at least some insulin, a proportion similar to what we found in 2017, in this larger cohort.8 Most of these individuals were on insulin as of their most recent visit (n=54), with a minority no longer on insulin (n=4) or not reporting insulin use at the most recent visit (n=6).

Clinical factors associated with DM, univariable and multivariable analyses.

We performed univariable analyses first. We did not detect any sex-specific differences in DM risk. Older age was associated increased risk of DM (OR 1.03, 95% CI: 1.02-1.04, p=<0.001). The median age of those with DM was 36 years (IQR, 25-53) as compared to 26 years in those without DM (IQR, 18-38), p=<0.0001 for difference by Wilcoxon rank sum test. Only 17/96 (18%) of individuals with FRDA-related DM had height and weight measurements at their most recent FACOMS visit, as compared to (334/1104=31%) of the whole cohort (p=0.01 for difference in measurement availability). Also, mFARS score was lower in those with BMI measurements (mean mFARS 50, 95%CI 48-52) compared to those without available measurements (mean mFARS 58, 95% CI 57-59, p=<0.001 for difference), consistent with our prior study.18 Of children and adults with available measurements, those who had overweight or obesity (n=83) according to age-appropriate BMI thresholds, had an increased risk of DM compared to those who were normal weight (n=208), OR 3.4 (95% CI: 1.2-9.5, p=0.018). With respect to genetics, individuals with point mutations were more likely to have DM than individuals with two GAA expansions (OR 3.11, 95% CI 1.58-6.12, p=0.001). When we restricted the analysis to individuals who had two GAA expansions, we found a nominal, but not statistically significant, association between longer GAA repeat length on the least affected allele and DM diagnosis (OR 1.10, 95% CI 0.99-1.2, p=0.08). Individuals with DM had slightly longer GAA repeat lengths than individuals without DM (728 vs. 680 nucleotides, p=0.04 by Wilcoxon rank sum test). When stratified by age of onset (0-7, 8-14, 15-24, >24 years), GAA repeat length was associated with diabetes in those with age of onset 8-14 years (OR 1.003, CI: 1.00-1.01, p=0.007). We did not detect an association between likelihood of DM and age of FRDA symptom onset.

When we examined the potential association between DM and other clinical features of FRDA, we found that having more severe disease (mFARS 4th vs. 1st quartile, OR 10.3, 95% CI 4.4-24.5, p=<0.001) was associated with increased likelihood of DM. With respect to comorbidities, we did not detect an association between scoliosis and DM. In contrast, hypertrophic cardiomyopathy, the presence of an arrythmia, and the presence of any cardiac disease were each associated with increased likelihood of DM (p=0.02, 0.001, 0.007, respectively).

Next, we performed a multivariable regression analysis including the risk factors identified in univariable analyses. We focused on individuals with GAA triplet expansions on both alleles (n=934) given heterogeneity of individuals with point mutations, and included sex as a biologically relevant covariate since there are sex differences in the rates of “common” forms of DM.19 Overall (Table 2), we found that older age (OR 1.05, 95% CI 1.02-1.07, p<0.001), more severe disease (mFARS 4th vs 1st quartile, OR 3.9, 95% CI 1.3-11.4, p=0.012), and cardiac disease (OR 2.9, 95% CI 1.4-6.2, p=0.005) were associated with greater risk of DM. We did not detect an independent association between longer GAA repeat length and increased risk of DM (Table 2). We did not detect an association between age of FRDA symptom onset and DM in a sensitivity analysis (results not shown). We performed an additional multivariable regression analysis including both individuals with and without point mutations (n=989), and confirmed that the presence of a point mutation on one allele (as compared to two GAA repeat expansions) was independently associated with a significantly greater risk of DM even after accounting for age, sex, and the presence of cardiac disease (Supplementary Table 1).

Table 2:

Multivariable logistic regression assessing clinical risk factors associated with diabetes.

Characteristic Likelihood of Diabetes, OR (95% CI) N=934
Current age (years) 1.05 (1.02-1.07) ***
Female sex (vs. reference=male) 1.24 (0.75-2.05)
Clinical disease severity, as reflected by mFARS score: median (IQR)
  Quartile 1: 34 (28-38) (reference)		 -
  Quartile 2: 49 (47-53) 1.31 (0.42-4.03)
  Quartile 3: 62 (60-66) 2.21 (0.78-6.31)
  Quartile 4: 76 (73-83) 3.93 (1.35-11.44) *
GAA repeat length on the shorter allele (per 100 nucleotides) 1.12 (0.97-1.30)
Presence of any cardiac disease (vs. reference=no cardiac disease) 2.93 (1.39-6.17) **
Open in a new tab

Multivariable logistic regression assessing the independent associations of clinical risk factors with the presence of diabetes in FRDA, excluding individuals with point mutations. Statistically significant OR are indicated by bold text with p values.

*

p<0.05,

**

p<0.01,

***

p<0.001.

Because age of FRDA symptom onset is a critical parameter for FRDA disease severity, reanalysis of risk factors was completed by stratifying participants into those with age of onset less than 15 years and those with age of onset greater than or equal to 15 years (Supplementary Table 2).20 The association with the presence of cardiac disease was attenuated in those with age of onset < 15 years and the association with mFARS score was no longer significant in those with age of onset ≥15 years.

4.2. CHOP-focused EMR review:

Participant characteristics.

Ten percent (44 of 489 individuals, including children and adults) of the CHOP FACOMS cohort reported having FRDA-related DM (Figure 1). Of these, 35 (out of 44) had sufficient EMR data to be included. The 9 excluded had limited EMR notes (n=3), were deceased (n=5, excluded due to lack of current data), and/or the note in the EMR indicated no DM diagnosis (n=1) (Table 1).

Diabetes diagnosis and glycemic control.

Similar to the overall cohort, median age of DM diagnosis was 26 years (IQR 15-31). Two participants were diagnosed with DM before they were diagnosed with FRDA. One was diagnosed with DM at age 12 years (with negative GAD65, IA-2, and insulin anti-pancreatic auto-antibodies), and FRDA symptom onset was at 22 years. This individual is treated with insulin (0.8units/kg/day) and has never had DKA. The other individual was diagnosed with DM at 6 years (with negative islet cell auto-antibodies), and soon after was diagnosed with FRDA via genetic testing in the setting of a family history, with additional clinical symptoms of FRDA appearing 1-2 years later. Of the 35 participants with FRDA-related DM, none of the individuals who had at least one pancreatic auto-antibody checked (n=10/35) had any evidence of auto-immunity.

With respect to treatment, similar to the overall FACOMS cohort, 69% (24/35) were treated with insulin. Metformin (dimethylbiguanide) use was common as well, noted in 43% (15/35) of individuals. Two individuals have notes documenting stopping metformin due to side effects (GI upset, fatigue). Forty percent (14/35) used at least one other class of medications including sulfonylureas, GLP-1R agonists, thiazolidinediones, DPP-4 inhibitors, bile acid sequestrants, amylin analogs, and/or combination medications (Figure 2). Only one individual had no recorded use of glucose lowering medication (DM managed with diet alone). Overall, median HbA1c was 7.0% (53 mmol/mol) on most recent measurement with IQR 6.0-8.3% (42-67 mmol/mol), range 5.1-9.6% (32-81 mmol/mol), in n=28/35 with available HbA1c measurements.

Figure 2: Medication use in FRDA-related DM.

Figure 2:

Open in a new tab

Medication use in the CHOP limited EMR review. Other medications in the CHOP limited EMR review cohort included bile acid sequestrants and amylin analogues. The natural history cohort had a similar percent (74%, 64/86) of individuals who reported having used insulin. The survey data also demonstrated a similar pattern to the EMR review, with 7/11 on insulin, 7/11 currently or previously on metformin, 2/11 on GLP-1R agonists, and 1/11 on metformin/pioglitazone.

Diabetes complications.

Two (out of 35) had episodes of clinically relevant ketosis recorded in the EMR. In both individuals, ketosis occurred after the initial diagnoses of FRDA-related DM. One individual was diagnosed with DM based on incidental hyperglycemia and confirmatory OGTT at 25 years, then treated with sitagliptin, but self-discontinued this medication within a year. This individual presented 2 years later with hyperglycemia and ketosis (BG 480mg/dL, β-hydroxybutyrate 6.7mmol/L, pH 7.32). The other individual had DM (diagnosed at approximately 18 years) that was adequately controlled on basal-bolus insulin for over 10 years (HbA1c range 5.9-8.1% (41-65mmol/mol)), then presented in DKA (BG 253mg/dL, β-hydroxybutyrate 5.76mmol/L, bicarbonate 14mmol/L) due to poor oral intake and insufficient insulin use in the setting of an infection. Neither had evidence of pancreatic auto-immunity (GAD65, islet cell, and ZNT8 antibodies were negative).

4.3. CHOP focused survey:

Participant characteristics.

The survey was distributed via email to 473 e-mail addresses of FACOMS participants (or parents of participants) who are followed at CHOP. One hundred and thirty-three individuals, both adults and children, with FRDA completed the survey, 11 of whom reported a history of DM. Nine individuals were also included in the “CHOP focused EMR review” (Figure 1).

Diabetes diagnosis and glycemic control.

All were diagnosed based on screening studies. In 5, random, i.e., non-fasting, blood draws were sufficient to demonstrate DM (hyperglycemia ≥200 mg/dL and/or HbA1c ≥6.5%), in 4, the diagnosis was made based on fasting blood glucose, and 2 underwent OGTTs. Participants reported that they either did not have pancreatic auto-antibodies checked (n=6) or they did not know if pancreatic auto-antibodies had been checked (n=5). With respect to etiology of DM, they reported being told they had T1D (n=2), T2D (n=5), or FRDA-related DM (n=4). Family history of DM was common, with 8/11 reporting at least one first-degree relative with DM. Seven individuals reported receiving care for DM from an endocrinologist, with the balance from their primary care physician.

Seven individuals (64%) reported current insulin use. The median total daily dose of insulin was dichotomous, with 3/7 individuals using < 0.2 units/kg/day (basal only) and 4/7 requiring > 0.75 units/kg/day (basal and at least one bolus). In terms of technology, five reported using insulin pens and two used insulin pumps, one of which was on a closed-loop system. Many individuals reported difficulty administering their own insulin with 3/7 needing assistance for all steps required to administer insulin with either a syringe and vial or insulin pen.

In terms of non-insulin anti-diabetic agents, 64% (7/11) patients are currently or were previously taking metformin. One discontinued it because of apparent lack of efficacy. Three out of seven reported gastrointestinal side effects of metformin (abdominal pain, vomiting, or diarrhea), but none reported elevated lactate, B12 deficiency, or other side effects. Two individuals (18%) reported using GLP-1R agonists, and both denied having side effects. Two individuals reported being on combination medications; one was taking metformin/pioglitazone, and reported no side effects, and the other did not specify. None reported currently or previously taking other classes of medications (Figure 2).

Similar to the larger cohort, these individuals (n=10 with available data) are overall well-controlled with average reported most recent HbA1c of 7.2%. SD 1.2 (55mmol/mol, SD 13).

Monitoring:

With respect to self-monitoring of blood glucose, all 11 of the participants surveyed reported checking glucoses, and 4/11 had used a continuous glucose monitor. Five reported that they were not able to check their own glucoses. Two experienced hypoglycemia severe enough to require glucagon administration. Six individuals reported ketone monitoring; four reported they had ketones that were small or greater and two had been hospitalized for DKA.

Diabetes-specific quality of life.

Respondents under 18 years old were invited to complete the Problem Areas in Diabetes Child/Teen (PAID-C/T) scales.16,17 On the PAID-C scale, the respondent reported no problems in eight of 11 parameters and a “medium problem” in three categories. The individual who completed PAID-T reported no problems in all 14 parameters. The nine adults completed the Diabetes Distress Scale (DDS).15 The median score on the diabetes distress scale was 22 (range 17-56). All individuals except one (who had a DDS score of 56) had scores under 30.

5. Discussion:

This study expands upon our group’s prior work on FRDA-related DM, published in 2017.8 At that time, the FACOMS natural history cohort included 52 individuals with FRDA-related DM (52/811, 6.4%).8 Now, FACOMS includes 96 individuals with FRDA-related DM, 8.7% of the N=1,099 participants with available data. In contrast, in the U.S. population, in individuals aged 18-44 years, a range that is similar to the inter-quartile range for the FACOMS cohort, the prevalence of DM was 3% (95%CI: 2.6-3.6) in 2013-2016.21

The larger sample size has enabled us to generate new insights. We focused on FRDA-related DM risk factors that are readily assessed in clinical practice. We envision these findings will lead to improved, evidence-supported strategies for screening.

Advancing age is a well-established risk factor for “common” type 2 DM21, and appears to increase risk for FRDA-related DM as well.8 Individuals with clinically severe disease are more likely to have a combination of more genetically severe disease, longer disease duration, chronic illness-related stress, and more prolonged inactivity, all of which could plausibly contribute to risk for FRDA-related DM.22 In our stratified models, the association between mFARS score and risk for FRDA-related DM appears driven by those with age of FRDA symptom onset <15 years. We speculate that in those with earlier onset disease, who tend to have more rapidly progressive and severe disease, mFARS is thus more tightly coupled to the incidence of comorbidities.

In terms of genetics, while GAA repeat length was not significantly associated with DM diagnosis, when the analysis was restricted to individuals with age of onset between 8-14 years, there is an association. We hypothesize that this cohort will have the most genotype-phenotype correlation. Those with younger age of onset are heterogenous and may have severe phenotypes or be diagnosed early due to family history, while those with age of onset over 15 have milder FRDA, and DM may be more likely to be multifactorial. Having a point mutation in the frataxin gene was also associated with increased risk of DM. Other studies have made this observation, with one (N=158) measuring a prevalence of DM among those with point mutations (n=16/158) of 40%, as compared to 4.3% in those without point mutations.7 Previously, we did not identify an association between point mutations and FRDA-related DM, which may have been because of smaller sample size and heterogeneity between individuals with point mutations.8 For example, some point mutations confer milder clinical disease.3 The overall low prevalence of point mutations in FRDA makes the DM prevalence estimates in this subgroup likely imprecise. Investigating the in vitro effects of specific point mutations in pancreatic β-cell model systems6 may help determine the extent to which point mutations impact glucose-stimulated insulin secretion, since pancreatic β-cell failure is a critical component of FRDA-related DM pathogenesis.

Finally, we found that cardiac disease (hypertrophic cardiomyopathy, dilated cardiomyopathy, and/or arrhythmia) was associated with increased risk of FRDA-related DM. We identified this association likely related to the larger sample size and longer duration of follow-up in the current FACOMS cohort.8 Of note, this association was attenuated in individuals with age of FRDA with symptom onset <15 years, perhaps because the prevalence of cardiomyopathy was already high in this sub-group, such that we could not detect the impact of variation in cardiomyopathy status on DM.

One possible explanation for the association between cardiomyopathy and FRDA-related DM is that individuals with severe FRDA are more likely to develop both co-morbidities. Alternatively, DM may contribute to the pathogenesis of cardiac disease. In adults without FRDA, DM is a risk factor for non-ischemic cardiomyopathy and heart failure.23 In adults without FRDA, impaired glucose tolerance is prevalent in heart failure and correlates with heart failure severity.24 Finally, the chronic stress of cardiomyopathy could contribute to abnormal glucose metabolism. The complex relationship between FRDA-related cardiac disease and glycemia is under investigation. This work is additionally motivated by the availability of new glucose-lowering agents (specifically SLGT2 inhibitors) that have clear benefits in non-FRDA related cardiac disease, and could be repurposed for FRDA.25 We did not identify any patients currently taking SGLT2 inhibitors.

Our study also provides valuable information about presentation, management practices, and impact of FRDA-related DM. In our EMR review, 2 out of 35 individuals with FRDA-related DM had episodes of clinically significant ketosis, consistent with previous reports and confirming β-cell dysfunction as a component of the pathophysiology of FRDA-related DM.26 None of the individuals who had been tested for the presence of pancreatic auto-antibodies (N=10) had evidence of auto-immunity. However, at least one prior report has described the presence of pancreatic auto-antibodies in individuals with both FRDA and DM.26 We propose that in individuals whose clinical presentation is consistent with insulin deficiency (e.g., DKA, low c-peptide, pediatric presentation), testing for pancreatic auto-antibodies be considered. We were not able to assess for the presence of microvascular complications in individuals with FRDA-related DM. Clinically, we are aware that dedicated screening for DM-related complications may be inconsistent, and in the case of DM-related neuropathy, challenging to evaluate in the context of FRDA. Further investigation is needed to assess the risk for microvascular complications in individuals with FRDA-related DM. If in the future, individuals with FRDA-related DM survive longer with the benefit of FRDA-focused therapies, the risk for DM complications may become even more relevant.Eight of 11 individuals with FRDA-related DM surveyed reported having at least one first-degree relative with DM (regardless of FRDA status), perhaps more than might be expected than with “common” type 2 diabetes27, though the number of respondents is small. The potential contribution of genetic background/variants associated with Type 2 DM to FRDA-related DM risk is another possible avenue of future research.

Notably, many individuals with FRDA reported being told they had either Type 1 or Type 2 DM. Use of this terminology illustrates that FRDA-related DM is still underappreciated and poorly understood. While we expect a background prevalence of autoimmune DM, the absence of pancreatic auto-antibodies in young individuals with FRDA-related DM suggests that it is not the same as Type 1 DM. Also, FRDA-related DM occurs at younger ages and at a lower BMI than Type 2 DM. Taken together, these observations support that FRDA-related DM is an entity whose pathogenesis may share some features of other forms of DM, but also is likely a reflection of frataxin deficiency both in the pancreatic β cell and in peripheral tissues.

With respect to management, most individuals with FRDA-related DM (64-74% in each of the three cohorts) were receiving insulin. Additional survey data helped better understand insulin use. Two of the 11 survey respondents had experienced hypoglycemia severe enough to require glucagon. To avoid severe and/or undetected hypoglycemia, we recommend that continuous glucose monitoring be considered, in particular for individuals with limited dexterity in whom frequent fingersticks can be difficult. Of the seven individuals in the survey who reported taking insulin, four were taking doses exceeding 0.75 units/kg/day (likely full replacement and/or reflecting insulin resistance), while the balance had low requirements, potentially suggesting a role for alternative glucose lowering therapies in a subset of patients.

With respect to other glucose-lowering therapies, metformin use was common and appeared well-tolerated. Risk for lactic acidosis and reports of mitochondrial respiratory chain complex I inhibition with metformin8,28 have appropriately led to caution with its use in individuals with mitochondrial forms of diabetes.29 Given this experience in FRDA, we propose in individuals with clinical evidence of insulin resistance, using the least effective dose and avoiding use in any patient with a personal history of lactic acidosis. We also advise providing education to discontinue the medication in the setting of serious illness and dehydration.

Newer anti-diabetic agents are being used in individuals with FRDA-related DM, particularly GLP-1R agonists (17% in EMR review cohort). In individuals with “common” T2DM, the use of SGLT-2 inhibitors is increasing30 though these agents do not appear to yet be used in those with FRDA-related DM. Given the benefits of SGTL2 inhibitors on heart failure in the general population,25 the effects of SGLT2 inhibitors on glycemic control and cardiac outcomes in FRDA is worthy of investigation. However, the risk of euglycemic DKA with SGLT2 inhibitor use and the known DKA risk with FRDA-related DM must be acknowledged, and the safety of SGLT2 inhibitors will need to be carefully studied and monitored.31 While optimal management of FRDA-related DM remains unknown, “N of 1” trials, in which alternative therapies are tested systematically in a single individual, may represent an opportunity to evaluate anti-diabetic medications in specific individuals. DM is well suited to studies using the “N of 1” format as it is a chronic disease with well-established biomarkers (blood sugar, time in range, and HbA1c), many medications that do not have a prolonged carry-over effect, and the effectiveness of many interventions can be assessed quickly.32 While individual “N of 1” trials may not be generalizable to all individuals with FRDA-related DM, they could provide valuable insights into management strategies tailored to affected individuals.

We found that individuals with FRDA-related DM reported less adverse impact of DM on quality of life as compared to other forms of DM.15 We speculate that for individuals not taking insulin or taking basal insulin only, DM management may not pose a substantial burden relative to other challenges associated with living with FRDA. While these results are encouraging from a quality-of-life standpoint, it remains possible that improved management of FRDA-related DM (or even pre-DM) could help to mitigate other long-term sequelae of FRDA.

This study has several limitations. First, the presence of DM is self-reported in FACOMS. Screening practices are inconsistent, thus there may be individuals with undiagnosed DM and/or whose DM diagnosis was made using varying criteria. OGTT may be more sensitive than HbA1c in diagnosing diabetes in FRDA,9 as well as in other populations,33 and therefore, HbA1c screening alone may not detect all cases of DM. In the future, updated screening guidelines may increase the consistency of reporting FRDA-related DM. Second, our detailed review of individuals with FRDA-related DM was limited to smaller cohorts, which allowed us to perform a more detailed abstraction, but may limit generalizability. BMI could be an important contribution to DM risk, and in those with available anthropometric measurements, BMI in the overweight or obese range was associated with increased risk of DM. Consistent with our prior study,18 those with available measurements had less severe FRDA. We also found individuals with FRDA-related DM were less likely to have measurements made. While the association between excess adiposity, as reflected by higher BMI, and DM is well-established in the general population, and this association appears be recapitulated in those with milder FRDA, this result may not be generalizable to individuals with more severe FRDA, in whom other factors (e.g., residual pancreatic cell function) may be more important for DM risk, and in whom BMI may be less reflective of excess adiposity (Vasquez-Trincado; personal communication). Finally, we focused on FRDA-related DM, and not pre-DM, the burden of which may be substantial and may also have other health impacts, particularly for the heart, as noted in other populations.9,24 Studies focused on the pathophysiology and impact of FRDA-related hyperglycemia will enrich our understanding.

In summary, FRDA-related DM is associated with advancing age, having a point mutation in frataxin, clinical disease severity, and the presence of cardiac disease. Clinicians who care for individuals with FRDA should bear these risk factors in mind when providing education and performing screening. Longitudinal studies are needed to evaluate interventions to prevent or delay FRDA-related DM and to optimize its treatment. Long-term studies are also needed to evaluate the potential contribution of FRDA-related DM to other clinically important outcomes in FRDA, particularly cardiac disease, and the exciting potential therapeutic relevance of SGLT2 inhibitors for both FRDA-related glucose metabolism and cardiomyopathy. In the future, we envision that evidence-supported screening, monitoring and management will improve health outcomes in individuals with FRDA-related DM.

Supplementary Material

1
Supplementary material2

Highlights:

  • Friedreich’s Ataxia diabetes (FRDA-related DM) was reported in 8.7% of those with FRDA.

  • Risk factors for FRDA-related DM include age, severe disease, and cardiac disease.

  • FRDA-related DM is distinct from common forms of diabetes.

  • Currently, insulin and metformin are the most common treatments for FRDA-related DM.

  • Newer anti-diabetic agents have not been used widely in FRDA-related DM.

Acknowledgements:

This study would not have been possible without the participation of individuals with FRDA in FACOMS and support from the clinical and research teams. This work was supported by Grants from the Friedreich’s Ataxia Research Alliance. We would also like to acknowledge Sara Pinney, MD, MTR for guidance in creating the EMR data collection forms and Kenneth Rodenheiser, BSN, RN, CDCES, CPT for assistance with survey creation.

Funding Sources:

J.T. was supported by NIH grant T32 DK063688 from the National Institute of Diabetes and Digestive and Kidney Disease and GRT-00000432 Friedreich’s Ataxia Research Alliance Post-Doctoral Fellowship Grant from the Friedreich’s Ataxia Research Alliance and is supported by the National Institutes of Health F32DK128970. D.R.L. and S.E.M are supported by the Friedreich’s Ataxia Research Alliance.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Declarations of interest: None

References

  • 1.Cossee M, et al. Evolution of the Friedreich’s ataxia trinucleotide repeat expansion: founder effect and premutations. Proc Natl Acad Sci U S A. 1997;94(14):7452–7. doi: 10.1073/pnas.94.14.7452. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Filla A, et al. The relationship between trinucleotide (GAA) repeat length and clinical features in Friedreich ataxia. American journal of human genetics. 1996;59(3):554–60. [PMC free article] [PubMed] [Google Scholar]
  • 3.Cossée M, et al. Friedreich’s ataxia: point mutations and clinical presentation of compound heterozygotes. Annals of neurology. 1999;45(2):200–6. doi: . [DOI] [PubMed] [Google Scholar]
  • 4.Fox NG, et al. Structure of the human frataxin-bound iron-sulfur cluster assembly complex provides insight into its activation mechanism. Nature communications. 2019;10(1):2210. doi: 10.1038/s41467-019-09989-y. Seungil Han are employees of Pfizer Inc. Nicolas G. Fox, Henry J. Bailey, Claire Strain-Damerell and Wyatt W. Yue declare no competing interests. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Tsou AY, et al. Mortality in Friedreich ataxia. Journal of the neurological sciences. 2011;307(1-2):46–9. doi: 10.1016/j.jns.2011.05.023. [DOI] [PubMed] [Google Scholar]
  • 6.Cnop M, et al. Central role and mechanisms of beta-cell dysfunction and death in friedreich ataxia-associated diabetes. Annals of neurology. 2012;72(6):971–82. doi: 10.1002/ana.23698. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Greeley NR, et al. Cross-sectional analysis of glucose metabolism in Friedreich ataxia. Journal of the neurological sciences. 2014;342(1-2):29–35. doi: 10.1016/j.jns.2014.04.015. [DOI] [PubMed] [Google Scholar]
  • 8.McCormick A, et al. Impact of diabetes in the Friedreich ataxia clinical outcome measures study. Annals of clinical and translational neurology. 2017;4(9):622–31. doi: 10.1002/acn3.439. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Azzi AS, et al. OGTT is recommended for glucose homeostasis assessments in Friedreich ataxia. Annals of clinical and translational neurology. 2019;6(1):161–6. doi: 10.1002/acn3.686. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Regner SR, et al. Friedreich ataxia clinical outcome measures: natural history evaluation in 410 participants. Journal of child neurology. 2012;27(9):1152–8. doi: 10.1177/0883073812448462. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Patel M, et al. Progression of Friedreich ataxia: quantitative characterization over 5 years. Annals of clinical and translational neurology. 2016;3(9):684–94. doi: 10.1002/acn3.332. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Rummey C, et al. Test-retest reliability of the Friedreich’s ataxia rating scale. Annals of clinical and translational neurology. 2020;7(9):1708–12. doi: 10.1002/acn3.51118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Rummey C, et al. Psychometric properties of the Friedreich Ataxia Rating Scale. Neurology Genetics. 2019;5(6):371. doi: 10.1212/nxg.0000000000000371. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Harris PA, et al. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. Journal of biomedical informatics. 2009;42(2):377–81. doi: 10.1016/j.jbi.2008.08.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Polonsky WH, et al. Assessing psychosocial distress in diabetes: development of the diabetes distress scale. Diabetes care. 2005;28(3):626–31. doi: 10.2337/diacare.28.3.626. [DOI] [PubMed] [Google Scholar]
  • 16.Evans MA, et al. Psychometric Properties of the Parent and Child Problem Areas in Diabetes Measures. J Pediatr Psychol. 2019;44(6):703–13. doi: 10.1093/jpepsy/jsz018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Shapiro JB, et al. Psychometric Properties of the Problem Areas in Diabetes: Teen and Parent of Teen Versions. J Pediatr Psychol. 2018;43(5):561–71. doi: 10.1093/jpepsy/jsx146. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Patel M, et al. Body Mass Index and Height in the Friedreich Ataxia Clinical Outcome Measures Study. Neurology Genetics. 2021;7(6):e638. doi: 10.1212/nxg.0000000000000638. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Huebschmann AG, et al. Sex differences in the burden of type 2 diabetes and cardiovascular risk across the life course. Diabetologia. 2019;62(10):1761–72. doi: 10.1007/s00125-019-4939-5.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Rummey C, et al. Scoliosis in Friedreich’s ataxia: longitudinal characterization in a large heterogeneous cohort. Annals of clinical and translational neurology. 2021;8(6):1239–50. doi: 10.1002/acn3.51352. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Centers for Disease Control and Prevention. National Diabetes Statistics Report, 2020. Atlanta, GA: Centers for Disease Control and Prevention, U.S. Dept of Health and Human Services; 2020. [Internet]. [Google Scholar]
  • 22.Rummey C, et al. Predictors of loss of ambulation in Friedreich’s ataxia. EClinicalMedicine. 2020;18:100213. doi: 10.1016/j.eclinm.2019.11.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Bertoni AG, et al. Diabetes and idiopathic cardiomyopathy: a nationwide case-control study. Diabetes care. 2003;26(10):2791–5. doi: 10.2337/diacare.26.10.2791. [DOI] [PubMed] [Google Scholar]
  • 24.Mamas MA, et al. Impaired glucose tolerance and insulin resistance in heart failure: underrecognized and undertreated? Journal of cardiac failure. 2010;16(9):761–8. doi: 10.1016/j.cardfail.2010.05.027. [DOI] [PubMed] [Google Scholar]
  • 25.McMurray JJV, et al. Dapagliflozin in Patients with Heart Failure and Reduced Ejection Fraction. The New England journal of medicine. 2019. doi: 10.1056/NEJMoa1911303. [DOI] [PubMed] [Google Scholar]
  • 26.Pappa A, et al. Diabetes mellitus in Friedreich Ataxia: A case series of 19 patients from the German-Austrian diabetes mellitus registry. Diabetes Res Clin Pract. 2018;141:229–36. doi: 10.1016/j.diabres.2018.05.008. [DOI] [PubMed] [Google Scholar]
  • 27.Owen K, et al. Increased Risk of Diabetes in First-Degree Relatives of Young-Onset Type 2 Diabetic Patients Compared With Relatives of Those Diagnosed Later. Diabetes care. 2002;25(3):636–7. doi: 10.2337/diacare.25.3.636. [DOI] [PubMed] [Google Scholar]
  • 28.Cameron AR, et al. Metformin selectively targets redox control of complex I energy transduction. Redox Biol. 2018;14:187–97. doi: 10.1016/j.redox.2017.08.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Yeung RO, et al. Management of mitochondrial diabetes in the era of novel therapies. J Diabetes Complications. 2020:107584. doi: 10.1016/j.jdiacomp.2020.107584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Kitten AK, et al. National ambulatory care non-insulin antidiabetic medication prescribing trends in the United States from 2009 to 2015. PLOS ONE. 2019;14(8):e0221174. doi: 10.1371/journal.pone.0221174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Peters AL, et al. Euglycemic Diabetic Ketoacidosis: A Potential Complication of Treatment With Sodium-Glucose Cotransporter 2 Inhibition. Diabetes care. 2015;38(9):1687–93. doi: 10.2337/dc15-0843. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Tsapas A, Matthews DR. N of 1 trials in diabetes: making individual therapeutic decisions. Diabetologia. 2008;51(6):921–5. doi: 10.1007/s00125-008-0983-2. [DOI] [PubMed] [Google Scholar]
  • 33.Vajravelu ME, Lee JM. Identifying Prediabetes and Type 2 Diabetes in Asymptomatic Youth: Should HbA1c Be Used as a Diagnostic Approach? Curr Diab Rep. 2018;18(7):43. doi: 10.1007/s11892-018-1012-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.9. Pharmacologic Approaches to Glycemic Treatment: Standards of Medical Care in Diabetes-2021. Diabetes care. 2021;44(Suppl 1):S111–s24. doi: 10.2337/dc21-S009. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1793942-supplement-1.pdf (114.6KB, pdf)
Supplementary material2
NIHMS1793942-supplement-Supplementary_material2.docx (25.7KB, docx)

RESOURCES