Table 4

Associations between low-carbohydrate, high-protein (LCHP) score and incident all-cause and site-specific cancer in Västerbotten Intervention Programme participants in a subgroup with reduced follow-up until 2002
Cancer type Sex LCHP score1 n Cases Model 12,3 HR (95% CI) p-trend5 Model 22,4 HR (95% CI) p-trend5
All cancer sites Men low 94 1 1
n = 531 medium 118 1.20 (0.91-1.58) 1.35 (1.01-1.79)
high 55 1.05 (0.75-1.46) 0.599 1.25 (0.86-1.80) 0.093
Women low 82 1 1
medium 101 1.10 (0.81-1.47) 1.15 (0.84-1.56)
high 81 1.29 (0.95-1.76) 0.049 1.39 (0.98-1.96) 0.020
Prostate cancer Men low 47 1 1
n = 106 medium 40 0.86 (0.56-1.32) 0.98 (0.63-1.51)
high 19 0.80 (0.47-1.38) 0.261 1.01 (0.57-1.80) 0.871
Breast cancer Women low 29 1 1
n = 91 medium 33 0.98 (0.59-1.62) 1.05 (0.62-1.78)
high 29 1.24 (0.74-2.09) 0.210 1.38 (0.77-2.46) 0.100
Colorectum All low 17 1 1
n = 57 medium 24 1.38 (0.74-2.58) 1.48 (0.77-2.81)
high 16 1.58 (0.79-3.13) 0.320 1.76 (0.83-3.73) 0.245

1 LCHP scores were calculated separately for FFQ version and sex, and categorized into roughly equally sized groups: low (2–8 points), medium (9–13 points) and high (14–20 points).

2 Hazard ratios were determined by Cox regression analyses.

3 Including age strata.

4 Further adjusted for obesity, sedentary lifestyle, lack of postsecondary education, current smoking, and intake of energy, alcohol, and saturated fat.

5 p-trend were calculated per 1-point increase in LCHP score.

Nilsson et al.

Nilsson et al. Nutrition Journal 2013 12:58   doi:10.1186/1475-2891-12-58

Open Data