Age, sex and ethnicity changes in creatine kinase and sex- and ethnicity-specific reference intervals of creatine kinase

Creatine kinase (CK) reference intervals (RIs) derived from predominantly White cohorts do not account for well-documented ethnic differences in CK. This misclassification may lead to unnecessary investigations or inappropriate treatment. We aimed to characterise age-, sex- and ethnicity-related variation in CK and establish RIs for a multi-ethnic population. Deidentified CK results of primary-care requests between 2008 and 2024 from a laboratory in the West Midlands (UK) were partitioned into six groups (male and female of White, Asian, Black ethnicities). Indirect RIs were derived from 58,096 individuals (33,879 female; 24,217 male) with the refineR algorithm. Derived RIs were verified in two independent datasets. Median CK and upper reference limits (URLs) were highest in Black, intermediate in Asian, and lowest in White groups. Female URLs for ≥13 years were 389 IU/L (Black), 188 IU/L (Asian) and 170 IU/L (White); male URLs were 757 IU/L, 357 IU/L and 314 IU/L, respectively. Applying the United Kingdom Pathology Harmony sex-specific RIs to all ethnicities classified 22.6–30.3% of results from Black individuals as high, versus 5.1–7.7% of White individuals. Using the new ethnicity-specific RIs reduced high-result rates in Black men and women. In individuals of mixed ethnicity, CK levels aligned more closely with the parent group with higher CK. After excluding outliers, the 97.5th percentile CK values for female and male children <13 years of Black ethnicities were 1.22 and 1.28 times higher, respectively, than those of White ethnicities. Adoption of CK RIs stratified by sex and ethnicity will allow more precise interpretation of the results, supporting equitable clinical decision-making and appropriate use of healthcare resources. • First validated sex- and ethnicity-specific creatine kinase (CK) reference intervals for the UK derived by indirect method from 58,096 individuals across White, Asian, and Black ethnic groups. • Individuals of Black and Asian groups show higher CK URLs than White counterparts • Applying non-ethnic reference intervals misclassifies up to 30% of results from individuals of Black ethnicities as elevated. • Ethnicity-specific reference intervals improve interpretation of CK results and support equitable clinical care.