Table 3.
Summary of significant age, gender, vertebra and race effects on the micro structural variables, grouped by distribution properties. Females, T12 and blacks are the positive directions in the Gender (G), Position (P) and Race (R) rows, respectively. The blank entries are those that were removed from the model during the backward stepwise analysis described in Methods. The entries that are not blank show the final reduced status of the model with “NS” indicating the nonsignificant factors. The final reduced model contains nonsignificant factors only when significant interactions are present. The results are separated between genders, positions and/or races if significant interactions were found. The results of such analyses are shown below the main model in the same cell. (Eg. Average Tb.Th is smaller in T12 than in L1. Further analysis of the interaction between age and gender suggested that Tb.Th.Av decreases with age in female and that T12 has less Tb.Th.Av than L1 in females.)
| BV/TV | Tb.Th | Tb.N | Tb.Sp | ||
|---|---|---|---|---|---|
| NS | ↓ p<0.0001 | ↑ p<0.0001 | |||
| NS | |||||
| ↓ p<0.05 | |||||
| Race (B) | |||||
| A*G | p<0.009 | ||||
| A*P | |||||
| A*R | |||||
| G*P | |||||
| G*R | |||||
| P*R | |||||
| A*G*P | |||||
| A*G*R | |||||
| A*P*R | |||||
| G*P*R | |||||
| A*G*P*R | |||||
| F: Age ↓ p<0.002 | |||||
| F: P(T12) ↓ p<0.03 | |||||
| see Fig. 3 | |||||
| SD | Age | NS | ↑ p<0.0003 | ||
| G (F) | ↑ p<0.009 | ↑ p<0.01 | NS | ||
| P (T12) | ↑ p=0.0154 | ↑ p<0.04 | |||
| Race (B) | NS | ||||
| A*G | NS | ||||
| A*P | NS | ||||
| A*R | NS | ||||
| G*P | p<0.04 | ||||
| G*R | NS | ||||
| P*R | NS | ||||
| A*G*P | NS | ||||
| A*G*R | p<0.03 | ||||
| A*P*R | NS | ||||
| G*P*R | NS | ||||
| A*G*P*R | p<0.05 | ||||
| M: P(T12) ↑ p<0.006 | |||||
| BFL: Age ↑ p<0.05 | |||||
| WFT: : Age ↓ p<0.03 | |||||
| BMT: Age ↑ p<0.03 | |||||
| see Fig. 4 | |||||
| CV | Age | ↑ p<0.004 | ↑ p<0.008 | ↑ p<0.03 | |
| G (F) | ↑ p<0.04 | ↑ p=0.0136 | NS | NS | |
| P (T12) | NS | ↑ p=0.0140 | ↑ p<0.007 | ||
| Race (B) | NS | NS | NS | ||
| A*G | NS | NS | NS | ||
| A*P | NS | NS | NS | ||
| A*R | NS | NS | NS | ||
| G*P | NS | p<0.04 | NS | ||
| G*R | NS | NS | NS | ||
| P*R | NS | NS | NS | ||
| A*G*P | NS | ||||
| A*G*R | p=0.0195 | p<0.03 | p<0.007 | ||
| A*P*R | NS | ||||
| G*P*R | NS | ||||
| A*G*P*R | p<0.04 | ||||
| B: Age ↑ p<0.003 | BFL: Age ↑ p<0.008 | BF: Age ↑ p<0.009 | |||
| BF: Age ↑ p<0.003 | BFT: Age ↑ p<0.04 | ||||
| B: G(F) ↑ p<0.04 |
An adjustment for multiple tests, taking into account the correlations between all parameters, suggested that a p-value of 0.0196 be considered as significant for an α of 0.05.