Healthy Premenopausal Women with Low HDL (less than 50 mg/dL) Have Significantly Lower Trabecular Bone Volume and Bone Formation Rate

Presentation Number: LB-OR01-1
Date of Presentation: March 5th, 2015

Adi Cohen*1, David W Dempster2, Hua Zhou2, Robert R Recker3, Joan M. Lappe3, Serge Cremers4, Ralph Müller5, Alexander Zwahlen5, Mafo Kamanda-Kosseh1, Mariana Bucovsky1, Emily M. Stein1, Tom Nickolas1 and Elizabeth Shane1
1Columbia University, New York, NY, 2Helen Hayes Hospital, West Haverstraw, NY, 3Creighton University, Omaha, NE, 4Columbia University, College of Physicians and Surgeons, New York, NY, 5ETH Zurich, Institute for Biomechanics


Excess visceral adipose tissue (VAT) is associated with low BMD and poor bone structure. We recently reported that healthy premenopausal women with high DXA trunk fat, a surrogate for VAT, had significantly lower trabecular (Tb) bone volume fraction (BV/TV), Tb number and thickness, lower bone formation rate (BFR) on bone biopsies and serum bone formation markers. In postmenopausal women, serum LDL is inversely and HDL directly related to BMD. We therefore investigated relationships between lipids, BMD, bone microstructure and remodeling in premenopausal women.

In 40 healthy premenopausal women (37 ± 8 yrs; BMI 20-39 kg/m2) with normal BMD by DXA and no history of fracture, we measured body composition by DXA; Tb BV/TV, microstructure and BFR/BS on tetracycline-labeled bone biopsies; and serum IGF-1, sclerostin, bone turnover markers (BTM), HDL, TG, and calculated LDL on fasting morning serum.

As expected, LDL (±SD: 100 ± 25 mg/dL) and TG (87 ± 53 mg/dL) were directly and HDL (50 ± 15 mg/dL) inversely related to BMI and DXA trunk fat. HDL was directly related to Tb BV/TV (r=0.40; p=0.01) and BFR (0.40; p=0.01). We compared 23 women with HDL <50 mg/dL, the metabolic syndrome cutoff, to 17 with normal HDL (63 ± 11 mg/dL). BMI (27.8 ± 4.9 vs 23.0 ± 2.3 kg/m2) and %DXA trunk fat (36 ± 8 vs 27 ± 8) were higher in the low HDL group (both p<0.001). Age, BMD, HOMAIR, serum glucose, PTH, 25-OHD and reported exercise did not differ. The low HDL group had significantly lower BV/TV (21.0 ± 7.3 vs 27.3 ± 6.8%; p=0.009) and Tb thickness (148 ± 36 vs 179 ± 32 µm; p=0.008), and lower BFR (0.006 ± 0.004 vs 0.012 ± 0.008 mm2/mm/yr; p=0.01) and serum markers of resorption (C-telopeptide, 214 ± 78 vs 386 ± 241 pg/mL; p=0.01) and formation (osteocalcin, 13 ± 5 vs 19 ± 10 ng/mL, p=0.04; and PINP, 42 ± 11 vs 53 ± 23 µg/L, 0.07). The low HDL group also had significantly lower IGF-1 (168 ± 40 vs 206 ± 65 ng/mL; p=0.03) and higher sclerostin (0.50 ± 0.19 vs 0.36 ± 0.10 ng/mL; p=0.01), both potential mediators of the detrimental relationship between fat and bone. Relationships between HDL and Tb BV/TV and BFR were unaffected by controlling for reported exercise but no longer significant after controlling for DXA trunk fat.  Relationships between HDL and Tb BV/TV remained significant after controlling for serum IGF-1 (0.38; p=0.018) and sclerostin (0.33; p=0.046). Relationships between HDL and BFR remained significant after controlling for serum IGF-1 (0.36; p=0.029) but were attenuated (0.28; p=0.093) by controlling for sclerostin.  

In summary, healthy premenopausal women with HDL levels consistent with the definition of metabolic syndrome (< 50 mg/dL) were distinguished by significantly lower Tb bone formation and bone volume on bone biopsies. We conclude that HDL is an easily measured marker of excess VAT that may identify young women at risk for poor bone health for studies of the skeletal effects of interventions that promote weight loss or increase exercise.


Nothing to Disclose: AC, DWD, HZ, RRR, JML, SC, RM, AZ, MK, MB, EMS, TN, ES