Bone mineral density and estimated height loss based on patients' recalls

Height loss has been shown to be an indicator of incident vertebral fractures. However, the relationship between height loss and bone mineral density (BMD) in different skeletal regions, as well as the power of human memory in estimation of height loss across the life span, has not yet been established. Given that the variation in BMD between populations is substantially less than the variation in fracture risk, we studied the relationship between height loss based on patient's recalls and BMD in Iranian men and women of all ages.

Randomized clustered sampling from all regions of Tehran was performed to recruit the study population. Participants were asked about their maximum recalled previously measured height, if they were confident. In the 457 participants included, the difference between the participants' maximum recalled and current measured height was calculated.

L1-L4 lumbar BMD, femoral neck BMD, and young adjusted T-scores were significantly lower in the group of participants with estimated height reduction of greater than 5 cm. In simple linear regression analysis, height loss was a significant predictor of femoral neck T-score (standardized beta coefficient=-0.15; p0.003) and L1-L4 lumbar T-score (beta=-0.08; p0.048). After adjustment for age, gender, and weight, height loss remained a significant predictor for femoral neck T-score (beta=-0.078; p0 .043). In multivariate models for lumbar T-score, height loss was an independent predictor only in participants equal to or younger than 50 years of age (beta=-0.144; p0.033).

Higher estimated height loss according to patients' recalls was an indicator of lower BMD in our sample. Especially in the femoral neck region, this factor might be considered as a substitute case-finding tool for low BMD. Considering relatively young nature of our study group and biological differences between populations, our findings need to be validated in future prospective studies.