Blood glucose — mg/dL vs mmol/L and where the 18.02 factor comes from

Blood glucose is a mass concentration in the US (milligrams per decilitre) and a molar concentration in most other countries (millimoles per litre). Both describe the same physical quantity — the conversion factor is fixed by glucose's molar mass (180.16 g/mol) and the dL-to-L unit step (÷10), giving mg/dL × 0.05551 = mmol/L, or equivalently mmol/L × 18.016 = mg/dL. This is unit-specific to glucose; cholesterol uses a different factor (38.67), triglycerides another (88.57).

The exact factors

Glucose: mg/dL × 0.05551 = mmol/L (or ÷ 18.016) > Glucose: mmol/L × 18.016 = mg/dL > Reference: glucose molar mass 180.156 g/mol; 1 dL = 0.1 L > HbA1c: DCCT % × 10.929 - 23.5 = IFCC mmol/mol

Standard clinical thresholds — both units

| Category | mg/dL | mmol/L | |---|---|---| | Hypoglycaemia (severe) | <54 | <3.0 | | Hypoglycaemia (mild) | 54-69 | 3.0-3.8 | | Fasting normal | 70-99 | 3.9-5.5 | | Fasting pre-diabetes | 100-125 | 5.6-6.9 | | Fasting diabetes (Dx) | ≥126 | ≥7.0 | | Post-meal target (2h) | <140 | <7.8 | | Post-meal impaired | 140-199 | 7.8-11.0 | | Post-meal diabetes (Dx) | ≥200 | ≥11.1 | | Random symptomatic | ≥200 | ≥11.1 |

Worked example — CGM alarm setup

A Libre 3 CGM shipped in a country using mmol/L, brought to the US. Default alarms: low 3.9 mmol/L (= 70 mg/dL), high 13.9 mmol/L (= 250 mg/dL). A US user without conversion might set 'low = 4' thinking it is 4 mg/dL — the app rejects it, but a manual notebook log noting 'BG 5' is ambiguous without the unit. Every CGM lets you switch units in settings; do it once and stay consistent.

HbA1c — the other unit ladder

HbA1c has its own pair: DCCT (%) used in the US, IFCC (mmol/mol) used in most of Europe. Conversion: IFCC = 10.929 × (DCCT − 2.15). Both describe fraction of glycated haemoglobin:

| DCCT (%) | IFCC (mmol/mol) | eAG mg/dL | eAG mmol/L | Interpretation | |---|---|---|---|---| | 5.0 | 31 | 97 | 5.4 | Normal | | 5.7 | 39 | 117 | 6.5 | Pre-diabetes start | | 6.5 | 48 | 140 | 7.8 | Diabetes threshold | | 7.0 | 53 | 154 | 8.6 | Good control target | | 8.0 | 64 | 183 | 10.2 | Elevated | | 10.0 | 86 | 240 | 13.4 | Poor control |

eAG (estimated average glucose) is derived from A1c by the ADAG formula: eAG (mg/dL) = 28.7 × A1c(%) − 46.7. It maps the long-term marker back to daily-reading units.

Where the 5.55 boundary shows up

Fasting glucose 100 mg/dL = 5.55 mmol/L is not clinically special — the pre-diabetes cutoff (100 mg/dL, 5.6 mmol/L) is defined in the US as 100 and rounded up to 5.6 in mmol/L countries. Both are meant to be the same threshold; the rounding difference is trivial in clinical decisions but visible on lab reports.

Fingerstick vs venous vs CGM

The three measurement methods do not agree exactly. Venous plasma (lab draw) is the reference. Fingerstick capillary reads ~5–7 mg/dL higher after meals due to arterial-side timing. CGM sensors read interstitial fluid, lagging blood glucose by 5–15 minutes when values are moving. All three units convert with the same 18.016 factor — the discrepancies are physiological, not unit issues.

Sports and endurance context

Athletes measuring during long events see values outside the clinical ranges: post-race 40–60 mg/dL (2.2–3.3 mmol/L) is common and asymptomatic in trained individuals; carbohydrate feeding pushes back up to 120–160 mg/dL (6.7–8.9 mmol/L). CGMs in non-diabetic athletes routinely alarm on both ends — the clinical alarm thresholds assume sedentary metabolism.

The five-second recap

mg/dL to mmol/L: divide by 18. mmol/L to mg/dL: multiply by 18. The pre-diabetes fasting cutoff is 100 mg/dL = 5.6 mmol/L; diabetes diagnostic is 126 mg/dL = 7.0 mmol/L. Cholesterol uses a different factor — do not reuse the 18.