2016 年 62 巻 5 号 p. 317-321
Carotenoids in fruit and vegetables are important for health, yet determining dietary intake is challenging. This study aimed to establish the validity and reliability of a portable field Resonance Raman Spectroscope (RRS) in reflecting human carotenoid status. A diagnostic accuracy study involving 81 healthy adults was conducted. The RRS was the index test. Serum carotenoids (β-carotene, lycopene, lutein, and zeaxanthin) and consumption of fruit and vegetables were primary and secondary reference standards respectively. Data were collected in two seasons. Validity was determined by the correlation between the RRS score and the two reference standards, and by diagnostic statistics comparing dichotomised RRS scores and serum β-carotene. Reliability was assessed by intra-class correlation from repeated observations within subjects and within repeated measurements using three devices. The RRS score was significantly correlated with the individual and summed serum carotenoids (r range 0.45 to 0.78; p always <0.001), and with fruit and vegetable intake (season one: r=0.38, p=0.016; season two: r=0.42, p<0.001). Sensitivity: 87.5%; specificity: 75.5%; positive and negative predictive values: 35.0% and 97.6% respectively. Within- and between-device reliability was high (r=0.98, p=0.004 and r=0.97, p=0.009 respectively). The RRS field model achieved criterion validity for assessing carotenoid status and fruit and vegetable intake, and also demonstrated reliability. It thus holds promise for the screening of carotenoid status and fruit and vegetable intake.
