Table 3.

Results from multiple linear regression models that evaluated the effect of type of protein intake on specific blood biomarkers (Results are presented as β-coefficients±SE, along with their 95% confidence intervals (CI) (n=490)

Parameter	Cystatin C, mg/dL		Haemoglobin, g/dL		Haematocrit, %
	β±SE	95% CI	β±SE	95% CI	β±SE	95% CI
Model 1: Animal protein intake (haem), % of total energy	-1.44±0.15	(-0.010, 0.001)	0.09±0.02	(0.04, 0.13)	0.27±0.07	(0.14, 0.40)
Model 2: #1+Age and sex	-0.0002±0.003	(-0.005, 0.005)	0.04±0.02	(0.005, 0.08)	0.15±0.05	(0.04, 0.25)
Model 3: #2+BMI, smoking, and physical activity	-0.0007±0.003	(-0.006, 0.004)	0.04±0.02	(0.004, 0.08)	0.14±0.05	(0.04, 0.24)
Model 4: #3+Alcohol and MedDietScore	-0.001±0.003	(-0.006, 0.004)	0.04±0.02	(0.006, 0.08)	0.15±0.05	(0.04, 0.25)

	Triglycerides, mg/dL†		Iron, μg/dL†		Haemoglobin, g/dL		Haematocrit, %
	β±SE	95% CI	β±SE	95% CI	β±SE	95% CI	β±SE	95% CI
Model 1: Animal protein intake (non-haem), % of total energy	-0.03±0.03	(-0.08, 0.02)	-0.05±0.02	(-0.09, -0.008)	-0.32±0.07	(-0.44, -0.19)	-0.93±0.18	(-1.3, -0.57)
Model 2: #1+Age and sex	-0.03±0.02	(-0.08, 0.02)	-0.03±0.02	(-0.07, 0.01)	-0.08±0.05	(-0.19, 0.02)	-0.27±0.15	(-0.57, 0.02)
Model 3: #2+BMI, smoking and physical activity	-0.02±0.02	(-0.06, 0.03)	-0.03±0.02	(-0.07, 0.01)	-0.07±0.05	(-0.18, 0.03)	-0.24±0.15	(-0.53, 0.05)
Model 4: #3+Alcohol and MedDietScore	-0.02±0.02	(-0.06, 0.03)	-0.02±0.02	(-0.06, 0.02)	-0.08±0.05	(-0.18, 0.03)	-0.25±0.15	(-0.55, 0.04)

	Platelets, 10^3/μL
	β±SE	95% CI
Model 1: Plant protein intake, % of total energy	-12±4.0	(-20, -4.1)
Model 2: #1+Age and sex	-11±4.0	(-18, -2.9)
Model 3: #2+BMI, smoking and physical activity	-9.0±3.9	(-17, -1.3)
Model 4: #3+Alcohol and MedDietScore	-6.4±4.4	(-15, 2.2)

^†Triglycerides and blood iron concentrations were log-transformed; Standard error