What tests should be performed to check insulin sensitivity in a patient suspected of having insulin resistance?

Investigations for Checking Insulin Sensitivity

The most practical approach to assess insulin sensitivity combines fasting plasma glucose (100-125 mg/dL indicates impaired fasting glucose), an oral glucose tolerance test (2-hour values of 140-199 mg/dL indicate impaired glucose tolerance), and HbA1c (5.7-6.4% suggests prediabetes), with fasting insulin levels (>15 mU/L directly confirms insulin resistance) providing additional confirmation when needed. 1, 2

Primary Diagnostic Tests

Glucose-Based Measurements

• Fasting Plasma Glucose (FPG): Values between 100-125 mg/dL indicate impaired fasting glucose, suggesting insulin resistance, with testing performed after at least 8 hours without caloric intake 3, 1, 2

• Oral Glucose Tolerance Test (OGTT): The 75-gram OGTT with 2-hour glucose values of 140-199 mg/dL indicates impaired glucose tolerance and is the most sensitive test for detecting early insulin resistance before fasting glucose becomes elevated 3, 1, 2

• Hemoglobin A1C: Values between 5.7-6.4% suggest prediabetes with underlying insulin resistance 3, 1, 2

Insulin-Based Measurements

• Fasting Insulin Levels: Normal is <15 mU/L, borderline high is 15-20 mU/L, and clearly elevated is >20 mU/L, with values >15 mU/L directly confirming insulin resistance 1, 2

• C-Peptide Measurement: Can help differentiate between endogenous insulin production and exogenous insulin administration; simultaneous elevation of insulin and C-peptide suggests endogenous hyperinsulinism indicating insulin resistance 3, 1

Test Sensitivity Ranking

From most to least sensitive for earliest detection of insulin resistance: 2

1. OGTT with 2-hour glucose measurement
2. Combined fasting insulin + glucose (calculated indices like HOMA-IR or QUICKI)
3. Combined fasting insulin + triglycerides
4. Fasting plasma glucose alone
5. HbA1c
6. Fasting insulin alone

Calculated Indices (Research/Epidemiological Use)

While not routinely recommended for clinical diagnosis, several validated indices can estimate insulin resistance: 3, 4

• HOMA-IR (Homeostasis Model Assessment): Calculated from fasting glucose and insulin, correlates well with direct insulin resistance measurements 4, 5, 6

• QUICKI (Quantitative Insulin Sensitivity Check Index): Derived from fasting glucose and insulin values 4

• Matsuda Index: Calculated from OGTT glucose and insulin values 4

Clinical Risk Assessment

Before ordering tests, identify high-risk individuals requiring evaluation: 3, 1

• BMI ≥25 kg/m² (or ≥23 kg/m² for Asian Americans) 3, 1, 2
• First-degree family history of type 2 diabetes 3, 1
• High-risk race/ethnicity (Native American, African American, Hispanic/Latino, Asian American, Pacific Islander) 3, 1
• Physical signs: acanthosis nigricans, hypertension, dyslipidemia, polycystic ovary syndrome 3, 1
• History of gestational diabetes 3

Critical Testing Considerations

Always perform testing in the fasting state (minimum 8 hours without caloric intake) to avoid postprandial variations that can confound results 1, 2

The OGTT is more sensitive than fasting glucose alone for detecting early insulin resistance but is poorly reproducible and more difficult to perform in practice compared to fasting measurements 3, 2

Stress hyperglycemia during acute illness can temporarily mimic insulin resistance and should not be used for diagnosis 1

Note medications affecting glucose metabolism (corticosteroids, atypical antipsychotics) when interpreting results, as these can alter test accuracy 3, 1

Practical Clinical Algorithm

1. Screen with fasting plasma glucose and HbA1c as initial tests due to convenience and reproducibility 3

2. Add OGTT when fasting glucose is borderline (100-109 mg/dL) or when clinical suspicion remains high despite normal fasting values, as OGTT can detect abnormalities before fasting glucose becomes elevated 3, 2

3. Measure fasting insulin levels when direct confirmation of insulin resistance is needed, particularly in patients with borderline glucose values or strong clinical features of insulin resistance 1, 2

4. Consider C-peptide measurement in ambiguous cases to confirm endogenous insulin production, especially when differentiating between insulin resistance syndromes and other causes of hyperinsulinemia 3, 1

What NOT to Do

Routine measurement of insulin or proinsulin is not recommended for general screening in asymptomatic individuals, as these assays are primarily useful for research purposes rather than routine clinical diagnosis 3

Community screening outside healthcare settings is not recommended because people with positive tests may not seek appropriate follow-up, and there may be failure to ensure repeat testing for those who test negative 3