Brittle Diabetes: A Structured Diagnostic Approach to Unexplained Glycemic Variability

Moving from a Judgmental Label to Clinical Clarity

Author Perspective: A comprehensive review for postgraduate internal medicine trainees

Dr Neeraj Manikath , claude.ai

Abstract

The term "brittle diabetes" has historically been used to describe patients with wide, unpredictable glycemic excursions that seem resistant to standard therapeutic interventions. Rather than representing a distinct pathophysiological entity, this pattern reflects a clinical syndrome with multiple identifiable underlying causes. This review provides a systematic approach to diagnosing and managing these challenging cases, emphasizing the importance of ruling out factitious disorder, gastroparesis, hormonal comorbidities, and psychosocial barriers before labeling a patient as having truly "brittle" diabetes.

Keywords: Glycemic variability, brittle diabetes, factitious hypoglycemia, gastroparesis, diagnostic approach

Introduction

The concept of "brittle diabetes" emerged in the 1940s to describe patients with diabetes experiencing severe, unpredictable glucose fluctuations that disrupted daily life and resisted conventional management. The term itself carries a judgmental undertone, often implying that the patient is somehow "difficult" or "noncompliant." However, modern diabetology recognizes that extreme glycemic variability nearly always has an identifiable cause—whether medical, pharmacological, or psychosocial.

The prevalence of true brittle diabetes, once thought to affect 1-3% of insulin-dependent patients, has declined dramatically with advances in insulin analogs, continuous glucose monitoring (CGM), and insulin pump technology. What remains are cases where clinicians have not yet identified the root cause of instability. This article provides a structured diagnostic framework to systematically approach these patients, validating both the patient's experience and the clinician's frustration while offering a clear path forward.

The Clinical Phenotype: Recognizing the Pattern

Pearl #1: Before diving into complex workups, ensure you're truly dealing with problematic glycemic variability and not just normal physiological variation captured by CGM.

Patients with genuinely concerning glycemic instability typically present with:

Glucose swings exceeding 200-300 mg/dL within hours without clear precipitants
Recurrent severe hypoglycemia requiring third-party assistance
Unpredictable glucose responses to identical meals and insulin doses
Coefficient of variation (CV) >36% on CGM data
Frequent emergency department visits or hospitalizations for hypo- or hyperglycemia

Oyster: The widespread adoption of CGM has made us aware of glucose variability that was previously invisible. A CV of 25-36% may cause anxiety but doesn't necessarily represent pathological instability requiring aggressive investigation.

Rule Out Factitious First: The Elephant in the Room

The Uncomfortable Truth

Factitious hypoglycemia and fabricated hyperglycemia account for a significant minority of "brittle diabetes" cases, particularly in healthcare workers or patients with secondary gain from illness. While uncomfortable to consider, failure to identify factitious disorder leads to unnecessary investigations, inappropriate treatments, and perpetuation of illness behavior.

Diagnostic Approach

The C-peptide Gold Standard: When a patient presents with documented hypoglycemia (glucose <55 mg/dL with symptoms):

Simultaneously measure: glucose, insulin, C-peptide, proinsulin, and beta-hydroxybutyrate
Pearl #2: In exogenous insulin administration (factitious), you'll see:
- Low C-peptide (<0.6 ng/mL)
- High insulin levels
- Low or suppressed beta-hydroxybutyrate
- Insulin:C-peptide ratio >1 (normally <1)

Hack: Consider screening sulfonylurea and meglitinide levels if C-peptide is inappropriately elevated during hypoglycemia, as these agents can also be used factitiously.

Clinical Clues to Factitious Disorder

Healthcare profession or access to medications
Hypoglycemia occurring only when patient is alone
Resistance to observed insulin administration or supervised care
Dramatic symptomatic presentation inconsistent with glucose readings
Psychiatric comorbidity or history of other factitious disorders

Management Considerations: This diagnosis requires sensitivity and psychiatric consultation. Direct confrontation rarely helps; instead, frame the conversation around "stress-induced insulin sensitivity" or "treatment adjustments" while engaging mental health professionals.

References:

Grunberger G, et al. Factitious hypoglycemia due to surreptitious administration of insulin. Ann Intern Med. 1988;108(2):252-257.
Service FJ. Classification of hypoglycemic disorders. Endocrinol Metab Clin North Am. 1999;28(3):501-517.

The "Gut" Check: Gastroparesis as a Primary Driver

Pathophysiology and Prevalence

Gastroparesis affects up to 40% of patients with longstanding type 1 diabetes and 30% of those with type 2 diabetes. The delayed and erratic gastric emptying creates a fundamental mismatch between insulin action timing and nutrient absorption, resulting in postprandial hypoglycemia followed by late hyperglycemia.

Pearl #3: Gastroparesis should be your first thought in any patient with:

Postprandial hypoglycemia 2-3 hours after meals
Unexplained hyperglycemia 4-6 hours post-meal
Nausea, early satiety, or bloating
Wide glycemic excursions despite careful carbohydrate counting

Diagnostic Workup

Standard Approach:

Gastric emptying scintigraphy (4-hour study) remains the gold standard
- Abnormal if >60% retention at 2 hours or >10% at 4 hours
Oyster: Acute hyperglycemia itself delays gastric emptying, so ensure glucose <275 mg/dL before testing to avoid false positives

Emerging Technology:

Wireless motility capsules provide ambulatory assessment
CGM pattern recognition: Look for the "double hump" pattern—early dip, then late spike

Management Strategies

Pharmacological:

Metoclopramide (10 mg before meals): First-line but limited by tardive dyskinesia risk
Domperidone (10-20 mg QID): Effective but requires FDA exemption in United States
Erythromycin (40-250 mg TID): Develops tachyphylaxis within weeks
Prucalopride: Emerging data for diabetic gastroparesis

Hack: Use ultra-rapid acting insulins (Fiasp, Lyumjev) or inhaled insulin (Afrezza) given mid-meal or even post-meal in severe gastroparesis to better match delayed absorption.

Nutritional Modifications:

Small, frequent meals with lower fat and fiber content
Liquid or pureed consistency during symptomatic periods
Consider gastric electrical stimulation for refractory cases

References: 3. Camilleri M, et al. Clinical guideline: management of gastroparesis. Am J Gastroenterol. 2013;108(1):18-37. 4. Bharucha AE, et al. Diabetic gastroparesis. Endocr Rev. 2019;40(5):1318-1352.

The "Hormonal Rollercoaster": Endocrine and Autoimmune Comorbidities

Celiac Disease: The Silent Saboteur

Prevalence: 4-9% of type 1 diabetes patients have celiac disease, compared to 1% in the general population.

Mechanism of Glycemic Instability:

Villous atrophy causes erratic carbohydrate absorption
Malabsorption leads to unpredictable postprandial glucose patterns
Untreated celiac increases insulin resistance paradoxically

Screening and Diagnosis:

Tissue transglutaminase IgA antibody (tTG-IgA) with total IgA level
Pearl #4: Check IgA level simultaneously—IgA deficiency occurs in 2-3% of celiac patients and causes false-negative tTG results
If positive, confirm with small bowel biopsy before dietary changes

Clinical Impact: Studies show gluten-free diet improves glycemic control within 3-6 months, reducing insulin requirements by 15-30%.

References: 5. Barker JM, et al. Clinical characteristics of celiac disease in type 1 diabetes: a prospective study. Pediatr Diabetes. 2005;6(2):69-73. 6. Pham-Short A, et al. Screening for celiac disease in type 1 diabetes: a systematic review. Pediatrics. 2015;136(1):e170-e176.

Addison's Disease: The Great Masquerader

Prevalence: Occurs in 0.5-2% of type 1 diabetes patients as part of autoimmune polyglandular syndrome.

Clinical Presentation:

Recurrent unexplained hypoglycemia (reduced gluconeogenesis and cortisol-mediated insulin antagonism)
Decreasing insulin requirements over time
Hyperpigmentation, salt craving, fatigue, hypotension
Hyponatremia and hyperkalemia

Diagnostic Approach:

8 AM cortisol <3 μg/dL is highly suggestive
ACTH stimulation test: Cortisol <18 μg/dL at 60 minutes confirms diagnosis
Check ACTH level (elevated in primary adrenal insufficiency)
Hack: In patients with recurrent hypoglycemia despite insulin dose reduction, check a morning cortisol before pursuing expensive investigations

Management: Hydrocortisone replacement (15-25 mg daily in divided doses) dramatically stabilizes glucose control within days to weeks.

References: 7. Erichsen MM, et al. Clinical, immunological, and genetic features of autoimmune primary adrenal insufficiency. J Clin Endocrinol Metab. 2009;94(12):4882-4890. 8. Hahner S, et al. Epidemiology of adrenal crisis in chronic adrenal insufficiency. Eur J Endocrinol. 2015;172(3):277-285.

Thyroid Disorders

Hyperthyroidism:

Accelerates insulin clearance and increases hepatic glucose production
Increases insulin resistance
Look for tachycardia, tremor, weight loss, heat intolerance

Hypothyroidism:

Delays insulin absorption from subcutaneous sites
Slows gastric emptying
Can present with hypoglycemia due to reduced clearance

Pearl #5: Check TSH and free T4 annually in all type 1 diabetes patients as part of routine screening, and whenever glycemic control deteriorates unexpectedly.

Menstrual Cycle Effects: The Overlooked Variable

Physiological Changes:

Luteal phase (days 14-28): Progesterone increases insulin resistance by 15-30%
Follicular phase: Improved insulin sensitivity
This creates predictable but significant glucose variability in premenopausal women

Management Strategies:

Pattern recognition using CGM and menstrual tracking apps
Increase basal insulin by 10-20% during luteal phase
Adjust insulin:carbohydrate ratios accordingly
Consider extended-cycle oral contraceptives to reduce hormonal fluctuations

References: 9. Brown SA, et al. Effects of the menstrual cycle on glucose control in women with type 1 diabetes. Diabetes Care. 2015;38(Suppl 1):S1-S2. 10. Trout KK, et al. Menstrual cycle effects on insulin sensitivity in women with type 1 diabetes. Diabetes Care. 2007;30(10):2376-2378.

Social and Psychological Factors: The Hidden Curriculum

Eating Disorders: A Dangerous Intersection

Epidemiology:

Eating disorders affect 20-30% of young women with type 1 diabetes
"Diabulimia" (insulin omission for weight control) occurs in 10-15% of females with type 1 diabetes
Binge eating disorder is increasingly recognized in type 2 diabetes

Clinical Recognition:

Erratic glucose control in otherwise knowledgeable patients
Unexplained high HbA1c despite reported adherence
Weight concerns and body image issues
Recurrent DKA in type 1 diabetes
Oyster: Male patients with eating disorders are often missed—maintain high clinical suspicion regardless of gender

Diagnostic Approach:

Diabetes Eating Problem Survey-Revised (DEPS-R) screening tool
SCOFF questionnaire for general eating disorder screening
Look for discordance between reported insulin doses and prescription refill patterns
Check insulin pump or pen download data for intentional under-dosing

Management: Requires multidisciplinary care including:

Endocrinology
Psychiatry or psychology specialized in eating disorders
Dietitian with diabetes expertise
Family involvement when appropriate

Hack: Frame conversations around "getting the most out of your insulin" rather than accusatory questioning about omission. This therapeutic alliance is crucial for engagement.

References: 11. Young V, et al. Eating problems in adolescents with type 1 diabetes. Diabetes Care. 2013;36(6):1580-1586. 12. Goebel-Fabbri AE, et al. Insulin restriction and associated morbidity and mortality in women with type 1 diabetes. Diabetes Care. 2008;31(3):415-419.

Financial Barriers and Insulin Insecurity

The Scope of the Problem:

25-30% of Americans with diabetes report cost-related medication non-adherence
Insulin rationing leads to erratic dosing patterns and apparent "brittleness"
Patients may use expired insulin, share insulin, or take inappropriate doses to stretch supplies

Clinical Clues:

Inconsistent glucose patterns without clear cause
Patient appears knowledgeable but control is poor
Reluctance to discuss insulin dosing specifics
Missing appointments due to inability to afford copays

Pearl #6: Directly but compassionately ask: "Are there times when you've had to skip or reduce insulin doses because of cost?" This simple question often reveals the underlying issue.

Solutions:

Patient assistance programs from insulin manufacturers
340B pharmacy programs for eligible patients
Generic insulin options (though less physiologic)
Sliding scale fee clinics
Connection with social work services

References: 13. Herkert D, et al. Cost-related insulin underuse among patients with diabetes. JAMA Intern Med. 2019;179(1):112-114. 14. Litchman ML, et al. State of the science on insulin affordability. Curr Diab Rep. 2019;19(12):152.

Diabetes Distress and Burnout

Definition: Diabetes distress represents the emotional burden of living with diabetes, distinct from clinical depression but equally impactful on self-care and glycemic outcomes.

Manifestations:

Disengagement from diabetes care
Erratic blood glucose monitoring
Inconsistent insulin dosing
Apparent "brittleness" due to behavioral variability

Assessment Tools:

Problem Areas in Diabetes (PAID) scale
Diabetes Distress Scale (DDS)
Hack: A single question—"How burned out are you feeling about your diabetes?"—correlates well with formal assessment

Intervention:

Acknowledge and validate the difficulty of diabetes management
Simplify regimens when possible
Consider technology (CGM, automated insulin delivery) to reduce burden
Referral to diabetes-specialized mental health professionals

References: 15. Fisher L, et al. Diabetes distress but not clinical depression or depressive symptoms is associated with glycemic control in both cross-sectional and longitudinal analyses. Diabetes Care. 2010;33(1):23-28. 16. Polonsky WH, et al. Assessing psychosocial distress in diabetes. Diabetes Care. 2005;28(3):626-631.

A Systematic Workup Protocol: The 5-Point Checklist

When confronted with apparent brittle diabetes, proceed systematically through this framework:

Point 1: Verify True Instability

Review CGM data for ≥14 days
Calculate coefficient of variation (target <36%)
Document frequency of severe hypoglycemia (<54 mg/dL) and hyperglycemia (>250 mg/dL)
Assess time in range (target 70-180 mg/dL for >70% of time)
Hack: Use the "AGP report" (Ambulatory Glucose Profile) from CGM devices for standardized variability assessment

Point 2: Rule Out Factitious Disorder

During next documented hypoglycemia episode:

✓ Simultaneous glucose, insulin, C-peptide, beta-hydroxybutyrate
✓ If C-peptide elevated: sulfonylurea and meglitinide screen
✓ Review insulin vial/pen disposal patterns
✓ Consider observed insulin administration for pattern assessment

Point 3: Evaluate Gastrointestinal Function

✓ Clinical assessment for gastroparesis symptoms
✓ Gastric emptying study if symptoms present
✓ CGM pattern analysis for delayed postprandial peaks
✓ Trial of metoclopramide if high clinical suspicion

Point 4: Screen for Hormonal and Autoimmune Comorbidities

First-tier screening (all patients):

✓ TSH, free T4
✓ Tissue transglutaminase IgA + total IgA
✓ 8 AM cortisol if recurrent hypoglycemia
✓ Document menstrual cycle relationship in premenopausal women

Second-tier (based on clinical suspicion):

✓ ACTH stimulation test if cortisol <10 μg/dL
✓ Growth hormone/IGF-1 if acromegaly suspected
✓ 21-hydroxylase antibodies if Addison's suspected

Point 5: Address Psychosocial and Economic Factors

✓ Screen for eating disorders (DEPS-R)
✓ Assess diabetes distress (PAID or DDS)
✓ Directly address insulin affordability
✓ Review medication regimen complexity
✓ Evaluate family/social support systems
✓ Consider referral to social work and/or diabetes psychology

Pearl #7: Don't pursue all investigations simultaneously. The systematic approach not only improves diagnostic yield but also demonstrates to the patient that their concerns are being taken seriously.

Technology-Based Solutions for Persistent Variability

After addressing underlying causes, technology can dramatically reduce variability:

Continuous Glucose Monitoring

Real-time alerts prevent severe excursions
Pattern recognition guides insulin adjustments
Coefficient of variation improved by 20-30% in most studies

Automated Insulin Delivery (AID) Systems

"Hybrid closed-loop" systems (e.g., Tandem Control-IQ, Omnipod 5, Medtronic 780G)
Time in range improves to 70-80% in previously unstable patients
Reduces fear of hypoglycemia and diabetes distress
Hack: Even patients with identified underlying causes benefit from AID technology while addressing root problems

Emerging Technologies

Ultra-rapid insulins (Fiasp, Lyumjev) reduce postprandial excursions
Dual-hormone systems (insulin + glucagon) in development
Artificial intelligence-based decision support systems

References: 17. Brown SA, et al. Six-month randomized, multicenter trial of closed-loop control in type 1 diabetes. N Engl J Med. 2019;381(18):1707-1717. 18. Bergenstal RM, et al. Safety of a hybrid closed-loop insulin delivery system in patients with type 1 diabetes. JAMA. 2016;316(13):1407-1408.

Case-Based Learning: Putting It All Together

Case 1: The Healthcare Worker

Presentation: 28-year-old female nurse with type 1 diabetes for 12 years presenting with recurrent severe hypoglycemia requiring paramedic assistance 3-4 times monthly. Episodes occur both at work and home, glucose as low as 25 mg/dL, yet recovers rapidly. CGM shows dramatic variability (CV 48%).

Workup: During one episode, simultaneous labs showed glucose 32 mg/dL, insulin 156 μIU/mL (markedly elevated), C-peptide <0.2 ng/mL (suppressed), and beta-hydroxybutyrate 0.4 mmol/L (low).

Diagnosis: Factitious hypoglycemia from exogenous insulin administration.

Management: Psychiatry consultation diagnosed borderline personality disorder with illness fabrication for attention. Structured diabetes care with observed insulin administration and intensive psychotherapy led to resolution.

Case 2: The Frustrated Patient

Presentation: 35-year-old male with type 1 diabetes for 20 years, HbA1c ranging 7.8-11.2% despite apparent good knowledge and insulin pump therapy. Frequent postprandial hypoglycemia followed by rebound hyperglycemia.

Workup: Gastroparesis symptoms (early satiety, bloating) led to gastric emptying study showing 45% retention at 4 hours. Celiac screening positive for tTG-IgA; biopsy confirmed Marsh 3b villous atrophy.

Diagnosis: Gastroparesis and celiac disease.

Management: Gluten-free diet and metoclopramide 10 mg before meals. Switched to ultra-rapid insulin analogs taken mid-meal. Three months later: HbA1c 7.1%, CV reduced from 42% to 28%, and patient-reported dramatic improvement in quality of life.

Case 3: The Adolescent Mystery

Presentation: 16-year-old female with type 1 diabetes for 5 years, previously well-controlled (HbA1c 6.8-7.2%), now presenting with erratic control (HbA1c 10.5%) and two DKA admissions. Parents frustrated, describing her as "non-compliant."

Workup: DEPS-R screening strongly positive for eating disorder symptoms. Pump download revealed consistent under-bolusing for meals and frequent basal rate reductions. Patient eventually disclosed insulin omission for weight control.

Diagnosis: Diabulimia (insulin omission eating disorder).

Management: Multidisciplinary team approach with diabetes psychology, family therapy, and endocrinology. Gradual restoration of appropriate insulin dosing with intensive psychological support. Eighteen months later: HbA1c 7.4%, healthy relationship with food and insulin, attending college successfully.

Clinical Pearls Summary

Always verify true instability before extensive workup—CGM has made us aware of normal physiological variation
C-peptide with simultaneous insulin during hypoglycemia is the key to diagnosing factitious disorder
Gastroparesis should be suspected in any patient with postprandial hypoglycemia followed by late hyperglycemia
Check total IgA when screening for celiac disease to avoid false-negatives from IgA deficiency
Morning cortisol is an inexpensive screen for Addison's disease in patients with recurrent hypoglycemia
Ask directly about insulin affordability—patients rarely volunteer this information
Systematic approach demonstrates respect for patient's experience and builds therapeutic alliance

Oysters (Clinical Traps) to Avoid

Assuming CGM-detected variability always requires intervention (CV 25-36% may be normal)
Missing factitious disorder due to reluctance to consider the diagnosis
Performing gastric emptying studies during hyperglycemia (causes false positives)
Overlooking menstrual cycle effects in premenopausal women with unexplained variability
Attributing glycemic instability to "non-compliance" without assessing for eating disorders or financial barriers
Failing to recognize male patients with eating disorders (often missed demographic)
Pursuing exotic diagnoses before addressing basic psychosocial factors

Hacks for Clinical Practice

The "mid-meal bolus": For gastroparesis patients, give ultra-rapid insulin halfway through the meal or even after eating
The "luteal phase adjustment": Increase basal insulin by 15-20% on day 14-28 of menstrual cycle automatically
The "one question screen": "Are there times you've had to skip insulin because of cost?" opens doors to financial discussions
The "AGP report trick": Use the standardized Ambulatory Glucose Profile from CGM devices rather than raw data dumps
The "observed dose trial": One week of observed insulin administration can quickly identify factitious disorder or confirm appropriate dosing
The "burnout question": "On a scale of 1-10, how burned out are you about your diabetes?" correlates with formal distress scales
The "technology bridge": While investigating underlying causes, implement automated insulin delivery to improve quality of life immediately

Conclusion: From Label to Solution

The term "brittle diabetes" should be retired from modern diabetology. When clinicians encounter patients with extreme glycemic variability, a systematic diagnostic approach reveals identifiable, treatable causes in the vast majority of cases. By ruling out factitious disorders, identifying gastroparesis and hormonal comorbidities, and addressing psychosocial and economic barriers, we validate both the patient's frustrating experience and provide a clear path to improved outcomes.

The key paradigm shift is moving from viewing these patients as "difficult" or "non-compliant" to recognizing that unexplained glucose variability represents a diagnostic challenge requiring methodical investigation. This approach respects patient dignity, enhances the therapeutic alliance, and ultimately improves both glycemic control and quality of life.

As educators, we must teach our trainees not to accept "brittle" as a diagnosis but rather as a call to action—a clinical syndrome demanding our most thoughtful, systematic, and compassionate care.

Key References for Further Reading

Grunberger G, et al. Factitious hypoglycemia due to surreptitious administration of insulin. Ann Intern Med. 1988;108(2):252-257.
Service FJ. Classification of hypoglycemic disorders. Endocrinol Metab Clin North Am. 1999;28(3):501-517.
Camilleri M, et al. Clinical guideline: management of gastroparesis. Am J Gastroenterol. 2013;108(1):18-37.
Bharucha AE, et al. Diabetic gastroparesis. Endocr Rev. 2019;40(5):1318-1352.
Barker JM, et al. Clinical characteristics of celiac disease in type 1 diabetes. Pediatr Diabetes. 2005;6(2):69-73.
Pham-Short A, et al. Screening for celiac disease in type 1 diabetes: a systematic review. Pediatrics. 2015;136(1):e170-e176.
Erichsen MM, et al. Clinical, immunological, and genetic features of autoimmune primary adrenal insufficiency. J Clin Endocrinol Metab. 2009;94(12):4882-4890.
Hahner S, et al. Epidemiology of adrenal crisis in chronic adrenal insufficiency. Eur J Endocrinol. 2015;172(3):277-285.
Brown SA, et al. Effects of the menstrual cycle on glucose control in women with type 1 diabetes. Diabetes Care. 2015;38(Suppl 1):S1-S2.
Trout KK, et al. Menstrual cycle effects on insulin sensitivity in women with type 1 diabetes. Diabetes Care. 2007;30(10):2376-2378.
Young V, et al. Eating problems in adolescents with type 1 diabetes. Diabetes Care. 2013;36(6):1580-1586.
Goebel-Fabbri AE, et al. Insulin restriction and associated morbidity and mortality in women with type 1 diabetes. Diabetes Care. 2008;31(3):415-419.
Herkert D, et al. Cost-related insulin underuse among patients with diabetes. JAMA Intern Med. 2019;179(1):112-114.
Litchman ML, et al. State of the science on insulin affordability. Curr Diab Rep. 2019;19(12):152.
Fisher L, et al. Diabetes distress but not clinical depression is associated with glycemic control. Diabetes Care. 2010;33(1):23-28.
Polonsky WH, et al. Assessing psychosocial distress in diabetes. Diabetes Care. 2005;28(3):626-631.
Brown SA, et al. Six-month randomized, multicenter trial of closed-loop control in type 1 diabetes. N Engl J Med. 2019;381(18):1707-1717.
Bergenstal RM, et al. Safety of a hybrid closed-loop insulin delivery system. JAMA. 2016;316(13):1407-1408.