What question did this study set out to answer?

To assess underweight as a diverse nutritional state in young Japanese women using various analytical methods.

April 7, 2026Nutrients2 citationsOpen Access

Underweight in Young Adult Women as a Dynamic Nutritional State: Evidence from Four Complementary Longitudinal Methods

KIKatsumi Iizuka HMHitomi Matsuura KYKotone Yanagi

Key Points

To assess underweight as a diverse nutritional state in young Japanese women using various analytical methods.
Retrospective analysis of 883 women aged 20–29 years over 6.1 years.
Comparison of year-to-year transitions, state occupancy, Aalen–Johansen estimator, and time-weighted average.
Bidirectional flow analysis to quantify inflow/outflow rates and BMI distribution.
Discrepancies of 31 points between methods in underweight classification.
Inflow rates from ages 22–27 exceeded outflow rates (35.7% vs. 20.7% per year).
Four distinct phenotypes of underweight women were identified based on time spent underweight.

Abstract

Background: Underweight (BMI < 18.5 kg/m2) remains prevalent among young Japanese women but lacks standardized measurement approaches. We compared four analytical methods and identified discrepancies. Methods: A retrospective analysis of 883 underweight women aged 20–29 years followed for 6.1 ± 4.2 years was performed. We compared (1) year-to-year transitions, (2) state occupancy, (3) the Aalen–Johansen estimator, and (4) time-weighted average. We performed bidirectional flow analysis quantifying inflow/outflow rates, BMI distribution analysis, and time-weighted classification. Results: Methods 1 and 4 showed 31-point discrepancies (78.1% vs. 47.1% in women). In bidirectional flow, inflow exceeded outflow at ages 22–27 (35.7%/yr vs. 20.7%/yr, outflow/inflow ratio: 0.58), balanced at ages 27–37 (ratio: 1.02) and showed outflow-dominant pattern at ages 37–47 (ratio: 4.92). BMI clustered at 18.0–19.0 kg/m2 (42.7%); 69.4% crossed the threshold once. Time-weighted classification revealed four phenotypes: persistent (≥75% time underweight; 40.1%, BMI: 17.54 kg/m2), moderate (50–74%; 17.6%, BMI: 18.40 kg/m2), intermittent (25–49%; 17.6%, BMI: 18.97 kg/m2), and transient (<25%; 24.8%, BMI: 19.49 kg/m2). The moderate + intermittent group showed yo-yo phenotypes (35.2%). Conclusions: Underweight in young Japanese women should be viewed as a heterogeneous dynamic nutritional state. The methodological discrepancy, threshold crossing, and phenotypic classification show that BMI-defined underweight comprises distinct patterns. Cross-sectional data evaluation may lead to incorrect assessments. Future research examining relationships between longitudinal low body weight subgroups and clinical outcomes could identify at-risk populations within the underweight group.

KI fragen

Bookmark

View Full Paper