Reference range for serum neurofilament light chain: Findings from healthy Thai adults

Serum neurofilament light chain is a notable biomarker for detecting axonal injury and has shown significant potential for clinical applications. Establishing a reference interval and cut-off level is a critical step towards implementing a serum neurofilament light chain in routine clinical practice. In this study, we aimed to establish a reference range of serum neurofilament light chains for the Thai population. Blood samples were collected from healthy Thai adults without a history of neurological diseases and screened at the Siriraj Hospital. The relationship between age, sex and log₁₀-transformed serum neurofilament light chain levels was analysed using linear regression. A crude reference interval was calculated as the 2.5-97.5th percentile values. An age-normative percentile curve for serum neurofilament light chain was derived using the generalized additive model for location, scale and shape. A total of 223 subjects (96 males and 127 females) aged 18-70 years were recruited. Male sex (P = 0.008) and older age (P < 0.001) were significantly associated with higher serum neurofilament light chain levels. A median of the observed serum neurofilament light chain values was 5.8 pg/ml (95% confidence interval 5.4-6.2), ranging from 1.0 to 18.4 pg/ml, with a crude reference interval of 2.3-15.9 pg/ml. The 2.5-97.5th percentile intervals for serum neurofilament light chain by age group were as follows: 20-29 years (n = 57): 1.7-8.7 pg/ml; 30-39 years (n = 58): 2.5-10.6 pg/ml; 40-49 years (n = 59): 3.5-14.3 pg/ml; 50-59 years (n = 37): 4.7-15.8 pg/ml and 60-69 years (n = 12): 4.2-18.2 pg/ml. The age-normative serum neurofilament light chain curve predicted the 97.5th percentile of 8.2, 9.9, 11.7, 14.6 and 19.9 pg/ml for ages 20, 30, 40, 50 and 60, respectively. This study is the first to establish reference values for serum neurofilament light chains in Thailand. The age-normative upper reference curve is closely aligned with observed values and those previously reported in other studies, providing a robust framework for clinical implementation. However, further validation in larger cohorts and among individuals with neurological diseases is warranted.