TABLE 3.
The association between tertiles of δ13C and δ15N values, types of dietary protein intake, and type 2 diabetes: the EPIC-Norfolk Study1
| Tertiles of stable isotopes and dietary protein types2 |
|||||
| 1 | 2 | 3 | Effect per tertile | P-trend | |
| δ13C | |||||
| Cases/subcohort | 205/245 | 126/234 | 145/239 | — | |
| Median (‰) | −23.1 | −22.8 | −22.5 | — | |
| Model 1 | 1 | 0.66 (0.53, 0.83) | 0.60 (0.48, 0.75) | 0.77 (0.69, 0.86) | <0.001 |
| Model 2 | 1 | 0.64 (0.50, 0.80) | 0.60 (0.48, 0.76) | 0.77 (0.68, 0.86) | <0.001 |
| Model 3 | 1 | 0.66 (0.52, 0.84) | 0.55 (0.43, 0.69) | 0.74 (0.65, 0.83) | <0.001 |
| Model 4 | 1 | 0.68 (0.54, 0.87) | 0.55 (0.43, 0.70) | 0.75 (0.65, 0.83) | <0.001 |
| δ15N | |||||
| Cases/subcohort | 131/248 | 170/235 | 175/235 | — | |
| Median (‰) | 9.9 | 10.3 | 10.7 | — | |
| Model 1 | 1 | 1.27 (1.01, 1.61) | 1.23 (0.97, 1.55) | 1.10 (0.98, 1.23) | 0.10 |
| Model 2 | 1 | 1.32 (1.05, 1.67) | 1.22 (0.96, 1.54) | 1.09 (0.98, 1.23) | 0.12 |
| Model 3 | 1 | 1.52 (1.19, 1.95) | 1.55 (1.22, 1.99) | 1.23 (1.09, 1.38) | 0.001 |
| Model 4 | 1 | 1.53 (1.19, 1.96) | 1.54 (1.21, 1.97) | 1.22 (1.09, 1.38) | 0.001 |
| Fish protein | |||||
| Cases/subcohort | 179/240 | 140/240 | 157/238 | — | |
| Median (g/d) | 3.3 | 5.9 | 10.4 | — | |
| Model 1 | 1 | 0.75 (0.59, 0.93) | 0.95 (0.76, 1.18) | 0.97 (0.86, 1.08) | 0.56 |
| Model 2 | 1 | 0.74 (0.59, 0.92) | 1.00 (0.80, 1.25) | 0.99 (0.88, 1.11) | 0.85 |
| Model 3 | 1 | 0.63 (0.50, 0.80) | 0.73 (0.57, 0.92) | 0.84 (0.74, 0.95) | 0.006 |
| Model 4 | 1 | 0.63 (0.50, 0.79) | 0.72 (0.57, 0.92) | 0.84 (0.74, 0.95) | 0.005 |
| Terrestrial animal protein 3 | |||||
| Cases/subcohort | 133/239 | 152/240 | 191/239 | — | |
| Median (g/d) | 29.7 | 42.6 | 58.5 | — | |
| Model 1 | 1 | 1.05 (0.82, 1.33) | 1.46 (1.16, 1.83) | 1.22 (1.09, 1.37) | 0.001 |
| Model 2 | 1 | 1.14 (0.89, 1.46) | 1.65 (1.27, 2.13) | 1.29 (1.14, 1.47) | <0.001 |
| Model 3 | 1 | 0.87 (0.68, 1.12) | 1.56 (1.19, 2.05) | 1.27 (1.10, 1.46) | 0.001 |
| Model 4 | 1 | 0.87 (0.68, 1.12) | 1.58 (1.20, 2.07) | 1.28 (1.11, 1.47) | 0.001 |
| Meat protein | |||||
| Cases/subcohort | 118/239 | 174/240 | 184/239 | — | |
| Median (g/d) | 12.5 | 23.7 | 36.2 | — | |
| Model 1 | 1 | 1.28 (1.00, 1.63) | 1.40 (1.10, 1.77) | 1.17 (1.05, 1.32) | 0.006 |
| Model 2 | 1 | 1.29 (1.01, 1.65) | 1.43 (1.10, 1.85) | 1.19 (1.05, 1.35) | 0.008 |
| Model 3 | 1 | 1.01 (0.78, 1.29) | 1.35 (1.04, 1.75) | 1.17 (1.02, 1.34) | 0.021 |
| Model 4 | 1 | 1.02 (0.80, 1.32) | 1.37 (1.06, 1.78) | 1.18 (1.03, 1.35) | 0.014 |
| Dairy protein | |||||
| Cases/subcohort | 167/239 | 155/240 | 154/239 | — | |
| Median (g/d) | 10.8 | 16.8 | 23.8 | — | |
| Model 1 | 1 | 0.91 (0.72, 1.14) | 1.07 (0.85, 1.35) | 1.03 (0.92, 1.16) | 0.56 |
| Model 2 | 1 | 0.92 (0.73, 1.16) | 1.25 (0.98, 1.58) | 1.11 (0.98, 1.25) | 0.10 |
| Model 3 | 1 | 0.87 (0.69, 1.09) | 1.07 (0.84, 1.37) | 1.03 (0.91, 1.17) | 0.66 |
| Vegetable protein | |||||
| Cases/subcohort | 158/239 | 158/240 | 160/239 | — | |
| Median (g/d) | 3.6 | 5.5 | 8.7 | — | |
| Model 1 | 1 | 1.05 (0.84, 1.32) | 1.17 (0.93, 1.46) | 1.08 (0.96, 1.21) | 0.18 |
| Model 2 | 1 | 1.07 (0.85, 1.34) | 1.33 (1.05, 1.67) | 1.15 (1.02, 1.29) | 0.018 |
| Model 3 | 1 | 0.97 (0.77, 1.23) | 1.04 (0.82, 1.32) | 1.02 (0.91, 1.15) | 0.75 |
| Cereal protein | |||||
| Cases/subcohort | 167/239 | 148/240 | 161/239 | — | |
| Median (g/d) | 4.5 | 8.4 | 14.4 | — | |
| Model 1 | 1 | 0.78 (0.62, 0.99) | 0.78 (0.62, 0.97) | 0.88 (0.79, 0.99) | 0.030 |
| Model 2 | 1 | 0.87 (0.69, 1.10) | 0.95 (0.74, 1.21) | 0.97 (0.86, 1.10) | 0.67 |
| Model 3 | 1 | 0.94 (0.74, 1.19) | 0.97 (0.75, 1.25) | 0.98 (0.86, 1.12) | 0.80 |
HRs (95% CIs) and P-trend values across tertiles of δ13C and δ15N and dietary protein types were estimated by Prentice-weighted Cox regression, adjusted for the following: model 1 [sex, family history of diabetes, smoking, education level, physical activity (results for δ13C values and δ15N values mutually adjusted for each other)], model 2 (model 1 + total energy intake, alcohol intake, and plasma vitamin C), model 3 (model 2 + BMI and waist circumference), and model 4 (model 3 + terrestrial animal protein for δ13C, δ15N, and fish protein and fish protein for meat protein and terrestrial animal protein). EPIC, European Prospective Investigation into Cancer and Nutrition; δ13C, carbon stable-isotope ratio (13C/12C ratio); δ15N, nitrogen stable-isotope ratio (15N/14N ratio).
Total cases: n = 476 and subcohort = 718 (includes 16 cases). Median values of isotope and protein tertiles are based on the subcohort only.
Terrestrial animal protein is the sum of meat, dairy, and egg protein.