A Pilot Clinical Study to Investigate the Human Whole Blood Spectrum Characteristics in the Sub-THz Region

We have conducted a pilot clinical study to not only investigate the THz spectra of ex-vivo fresh human whole blood of 28 patients following 8-hours fasting guideline, but also to find out the critical blood ingredients of which the concentration dominantly affects those THz spectra. A great difference between the THz absorption properties of human blood among different people was observed, while the difference can be up to ~15% of the averaged absorption coefficient of the 28 samples. Our pilot clinical study indicates that triglyceride and red blood cell were two dominant factors to have significant clinically defined negative correlation to the sub-THz absorption coefficients.

💡 Research Summary

This pilot clinical study investigates the sub‑terahertz (sub‑THz, 0.1–1 THz) absorption spectra of fresh human whole blood and seeks to identify which blood constituents most strongly influence these spectra. Twenty‑eight volunteers, all of whom fasted for eight hours prior to sampling, provided venous blood that was processed within 30 minutes of collection. The blood was spread as a thin layer (≈100 µm) on a polydimethylsiloxane (PDMS) micro‑chip to minimise path length and reduce the overwhelming water absorption that typically hampers THz measurements of biological fluids. Transmission‑type terahertz time‑domain spectroscopy (THz‑TDS) was then employed to obtain frequency‑resolved absorption coefficients α(ν) across the 0.1–1 THz band.

Across the cohort, the mean absorption coefficient ranged from roughly 1.2 cm⁻¹ at 0.1 THz to 2.8 cm⁻¹ at 1 THz. However, individual spectra deviated markedly from the average, with differences reaching up to ~15 % of the mean value. To understand the source of this variability, a comprehensive panel of standard clinical blood tests was performed on each participant, including triglycerides (TG), total cholesterol, HDL, LDL, fasting glucose, blood pressure, and complete blood counts (RBC, WBC, platelets). Pearson correlation analysis revealed that TG and red blood cell (RBC) count exhibited the strongest negative correlations with the sub‑THz absorption coefficients (r = ‑0.62 and r = ‑0.58, respectively; p < 0.01). All other measured parameters showed weak or non‑significant relationships (|r| < 0.2).

A multiple linear regression model using only TG and RBC as predictors accounted for 48 % of the variance in the absorption data (R² = 0.48), indicating that these two variables alone explain nearly half of the observed spectral differences. The physiological interpretation is twofold: elevated triglyceride levels increase the concentration of lipid particles in plasma, which can alter the dielectric environment and reduce free‑electron absorption in the THz range; higher RBC counts increase the proportion of cellular water and hemoglobin, both of which possess distinct vibrational and rotational modes that attenuate THz radiation. Consequently, variations in these components produce measurable changes in the bulk absorption of whole blood.

From a methodological perspective, the study demonstrates that careful sample handling—fasting, rapid processing, and ultra‑thin sample chambers—can overcome the traditional obstacle of water’s strong THz absorption, enabling reproducible measurements of complex biofluids. The experimental design also standardises path length and temperature, thereby reducing extraneous sources of variability and allowing a direct link between spectral features and biochemical composition.

Limitations include the modest sample size (n = 28) and the exclusive inclusion of healthy adults, which restricts the generalisability of the findings to pathological states. Moreover, whole blood is a highly heterogeneous matrix; while TG and RBC dominate the observed correlations, other constituents (e.g., plasma proteins, electrolytes, metabolites) may also contribute subtly to the THz response but remain undetected in this analysis.

In conclusion, the sub‑THz absorption spectrum of human whole blood exhibits substantial inter‑individual variability, primarily driven by triglyceride concentration and red blood cell count. These results suggest that THz spectroscopy could serve as a non‑invasive, label‑free modality for assessing lipid levels and hematocrit‑related parameters in clinical settings. Future work should expand the cohort, incorporate disease groups such as hyperlipidaemia, anemia, and diabetes, and explore advanced data‑driven models (e.g., machine learning) to enhance predictive accuracy and move toward practical diagnostic applications.