Does Long Term Energy & Carbohydrate Availability Affect Lumbar Spine Bone Gains In Female Adolescents?

In the short term, low energy and carbohydrate availability (EA, CA) may decrease bone formation markers in adults, yet long term EA & CA patterns with fat mass and lumbar spine bone mineral density (LSBMD, g/cm2) gains are unclear in adolescent girls. We have shown that peak velocity for LSBMD gains in active girls occurs circum-menarche (gynecological age, GYNAGE = PRE -2.5 yrs to POST +2.5 yrs). PURPOSE: To test circum-menarcheal EA & CA as factors in fat & LSBMD gains for active girls within the PREPOST peak growth window. METHODS: Annual measures provided PRE height (HT, cm) & PREPOST DXA data [LSBMD (g/cm2); Whole Body fat mass (FAT, kg), non-bone lean mass (nbLM, kg), PREPOSTnbLM (mean kg), fat mass index (FMI, kg/m2)]. Semi-annual questionnaires yielded PREPOST interval means for physical activity (PA, h/day), and diet intakes [Harvard YAQ: Energy (E, kcal/d), Carbohydrate (C, g/d)]. We calculated CA = [C / PREPOSTnbLM], EA = [((PA*300 kcal/h)-E) / PREPOSTnbLM], EA tertiles (LO, MED, HI), and annualized LSBMD gain (LSBMDgain, (g/cm2/yr) for the PREPOST INTERVAL. Multiple linear regression tested predictors of POSTFMI, POSTLSBMD, & LSBMDgain. Models included: PREGYNAGE, PREHT, PREFMI, EA/tertile, and CA. POSTFMI & POSTLSBMD models included INTERVAL (yrs). Bone models added PRELSBMD & weight-bearing PA (interval mean h/wk). RESULTS: This sample of gymnasts & non-gymnasts engaged in varied physical activities (n = 45, means: PREGYNAGE = -1.76 yrs, POSTGYNAGE = +1.69 yrs, INTERVAL = 3.49 yrs). In the POSTFMI model (R2 = 51.7%), key factors were: PREFMI (B = +0.629, p < 0.001), EA (B = -0.133, p < 0.001), CA (B = +0.738, p = 0.018). In the POSTLSBMD model (R2 = 74.9%), key factors were: PRELSBMD (B = +0.880, p < 0.001), PREFMI (B = +0.008, p = 0.013), with a weak EA tertile trend (B-0.33, p = 0.081). In the LSBMDgain model (R2 = 29.9%), key factors were: PREFMI (B = +0.003, p = 0.005), EA tertile (B = -0.014, p = 0.023), with a weak CA trend (B = +0.004, p = 0.059). PRE vs. POST correlations were high for LSBMD & PA (r > +0.84, p < 0.001). CONCLUSION: In active girls, peak adolescent LSBMD gains are a function of baseline PRE FMI & PRE LSBMD. Adequate fat reserves may blunt EA & CA influence, and bone gains may track with consistent PRE-POST exercise. Future work will assess more detailed exercise and nutrient profiles with a polygenic risk score for BMD. Supported by: NIH/NIAMS R01 54145; NIH/NIAMS R0346713; OREF (97-015); UW-Madison; SUNY Upstate Medical University