Table 5. Odds ratio (95% CIs) of hip fractures for quartiles of seafood stratified by gender, source of controls, and fracture sites in Guangzhou, China.
Quartiles of dietary energy-adjusted intake | p-trend | p-intera -ction | |||||
Variable | Pair N | 1 (referent) | 2 | 3 | 4 (highest) | ||
Freshwater fish | |||||||
Gender | 0.189 | ||||||
Male | 148 | 1.00 | 1.09(0.19–6.38) | 1.82(0.29–11.38) | 2.30(0.41–12.82) | 0.270 | |
Female | 433 | 1.00 | 0.82(0.47–1.43) | 1.08(0.60–1.95) | 0.52(0.28–0.97) | 0.212 | |
Source of controls | 0.143 | ||||||
Community | 398 | 1.00 | 1.07(0.56–2.03) | 1.76(0.90–3.45) | 1.21(0.60–2.46) | 0.510 | |
Hospital | 183 | 1.00 | 1.35(0.64–2.87) | 1.01(0.47–2.19) | 0.60(0.26–1.38) | 0.215 | |
Sea fish | |||||||
Gender | 0.052 | ||||||
Male | 148 | 1.00 | 0.24(0.06–1.02) | 0.03(0.01–0.19) | 0.10(0.02–0.47) | 0.015 | |
Female | 433 | 1.00 | 0.86(0.51–1.48) | 0.72(0.40–1.29) | 0.36(0.18–0.69) | 0.003 | |
Source of controls | 0.768 | ||||||
Community | 398 | 1.00 | 0.47(0.25–0.90) | 0.34(0.17–0.68) | 0.29(0.14–0.58) | <0.001 | |
Hospital | 183 | 1.00 | 0.75(0.35–1.60) | 0.30(0.13–0.68) | 0.18(0.06–0.54) | <0.001 | |
Mollusca and shellfish | |||||||
Gender | 0.540 | ||||||
Male | 148 | 1.00 | 0.51(0.09–2.74) | 0.89(0.17–4.61) | 0.08(0.01–1.37) | 0.252 | |
Female | 433 | 1.00 | 1.00(0.59–1.69) | 0.47(0.26–0.87) | 0.59(0.32–1.06) | 0.016 | |
Source of controls | 0.403 | ||||||
Community | 398 | 1.00 | 0.37(0.20–0.69) | 0.39(0.20–0.77) | 0.59(0.31–1.12) | 0.039 | |
Hospital | 183 | 1.00 | 0.54(0.25–1.17) | 0.34(0.15–0.79) | 0.25(0.09–0.68) | 0.002 | |
Total fish | |||||||
Gender | 0.864 | ||||||
Male | 148 | 1.00 | 1.24(0.29–5.41) | 2.10(0.41–10.73) | 0.41(0.07–2.43) | 0.665 | |
Female | 433 | 1.00 | 1.06(0.61–1.86) | 1.18(0.65–2.12) | 0.42(0.22–0.78) | 0.037 | |
Source of controls | 0.389 | ||||||
Community | 398 | 1.00 | 0.86(0.50–1.66) | 1.19(0.58–2.44) | 0.57(0.28–1.15) | 0.200 | |
Hospital | 183 | 1.00 | 0.95(0.47–1.91) | 1.36(0.63–2.91) | 0.25(0.09–0.70) | 0.075 |
Odds ratios (95% CI): from multivariate conditional logistic regression models. Covariates adjusted for: see ORII in Table 3. For women, years since menopause, oral contraceptive user, estrogen user were further adjusted for by the stepwise forward method.