Circulating fatty acids as biomarkers of dairy fat intake: data from the lifelines biobank and cohort study

Figures & data

Figure 1. Flow diagram of the participants in the study.

Table 1. Baseline characteristics of the Lifelines population.

Baseline Characteristics	N = 769	Males N = 404	Females N = 365	p-value
General
Age visit 1, years	53.0 ± 15.5	53.3 ± 15.2	52.7 ± 16.0	0.60
BMI, kg/m^2	26.0 ± 4.0	26.3 ± 3.4	25.8 ± 4.6	0.08
Smoking, yes (%)	115 (15.0%)	68 (16.8%)	47 (12.9%)	0.13
SBP, mmHg	126.5 ± 17.0	129.5 ± 15.1	123.1 ± 18.2	<0.001
DBP, mmHg	73.7 ± 9.7	76.6 ± 9.8	70.5 ± 8.4	<0.001
Laboratory measurements
Total Cholesterol, mmol/L	5.2 ± 1.0	5.1 ± 1.0	5.3 ± 1.1	0.03
HDL cholesterol, mmol/L	1.5 ± 0.4	1.3 ± 0.3	1.7 ± 0.4	<0.001
LDL cholesterol, mmol/L	3.3 ± 0.9	3.3 ± 0.9	3.3 ± 1.0	0.27
Total Triglycerides, mmol/L	1.0 (0.8–1.5)	1.2 (0.9–1.6)	1.0 (0.7–1.3)	<0.001
Glucose, mmol/L	5.1 ± 0.7	5.2 ± 0.6	5.0 ± 0.7	0.003
HbA1c, %	5.6 ± 0.4	5.6 ± 0.4	5.6 ± 0.4	0.99
Creatinin, umol/L	76.0 ± 13.6	83.6 ± 12.3	67.6 ± 9.3	<0.001
C-reative protein (hs-CRP), mg/L	1.0 (0.5–2.2)	0.9 (0.4–1.9)	1.1 (0.5–2.9)	0.05
General health
Diabetes Type 2, yes (%)	22 (2.9%)	12 (3.0%)	10 (2.8%)	0.86
Hypertension, yes (%)	200 (26.0%)	96 (23.8%)	104 (28.6%)	0.21
High cholesterol, yes (%)	165 (21.5%)	95 (23.5%)	70 (19.3%)	0.35
Dietary intake (FFQ)
Energy intake, Kcal/day	1971 ± 624	2206 ± 665	1711 ± 449	<0.001
Total carbohydrates, g/d (En%)	228.8 ± 76.7 (46%)	254.4 ± 80.7 (46%)	200.4 ± 60.6 (47%)	<0.001
Total protein, g/d (En%)	74.8 ± 20.9 (15%)	81.5 ± 22.6 (15%)	67.3 ± 15.8 (16%)	<0.001
Total fat, g/d (En%)	77.2 ± 30.0 (35%)	87.1 ± 32.7 (36%)	66.2 ± 22.0 (35%)	<0.001
Total Meat, g/d	73.5 ± 37.5	83.8 ± 40.1	62.0 ± 30.7	<0.001
Total Meat fat, g/d	10.2 ± 5.8	11.9 ± 6.3	8.3 ± 4.5	<0.001
Dairy intake (FFQ)
Total Dairy intake, g/d	322 (209–447)	315 (202–444)	324 (216–447)	0.99
Total Dairy fat intake, g/d	12.3 (8.4–17.4)	13.5 (9.0–18.3)	11.3 (8.0–15.6)	<0.001
Circulating fatty acids in plasma TG
C14:0, mol%	1.87 (1.40–2.48)	1.98 (1.49–2.68)	1.74 (1.35–2.35)	<0.001
C15:0, mol%	0.30 (0.25–0.35)	0.29 (0.24–0.35)	0.30 (0.25–0.35)	0.27
C17:0, mol%	0.49 (0.41–0.59)	0.48 (0.39–0.58)	0.52 (0.44–0.62)	<0.001
T-C16:1(n-7), mol%	0.026 (0.018–0.036)	0.025 (0.017–0.036)	0.026 (0.019–0.037)	0.05
T-C18:1(n-7), mol%	0.098 (0.064–0.141)	0.099 (0.067–0.147)	0.096 (0.070–0.138)	0.30
CLA, mol%	0.077 (0.050–0.095)	0.068 (0.049–0.094)	0.071 (0.050–0.097)	0.24
Circulating fatty acids in plasma PL
C14:0, mol%	0.49 (0.41–0.57)	0.47 (0.40–0.55)	0.50 (0.43–0.58)	<0.001
C15:0, mol%	0.28 (0.24–0.33)	0.28 (0.24–0.33)	0.29 (0.25–0.33)	0.09
C17:0, mol%	0.40 (0.36–0.44)	0.39 (0.35–0.44)	0.40 (0.36–0.45)	0.08
T-C16:1(n-7), mol%	0.020 (0.014–0.025)	0.019 (0.013–0.024)	0.020 (0.015–0.025)	0.002
T-C18:1(n-7), mol%	0.091 (0.068–0.115)	0.086 (0.063–0.111)	0.094 (0.072–0.119)	0.002
CLA, mol%	0.033 (0.022–0.043)	0.030 (0.020–0.041)	0.032 (0.017–0.044)	0.01

Data are presented as mean ± SD, median (25th–75th percentile) or number (%). Differences between sex were tested by independent t-test, Mann-Whitney U test or Chi-square test.

BMI: body mass index; SBP: systolic blood pressure; DBP: diastolic blood pressure; HDL: high density lipoprotein; LDL: low density lipoprotein; HbA1C: haemoglobin A1C; FFQ: Food Frequency Questionnaire; Kcal: Kilocalories: TG: Triglycerides; PL: Phospholipids: C14:0: Myristic acid; C15:0: Pentadecanoic acid; C17:0: Heptadecanoic acid; T-C16:1(n-7): Trans-Palmitoleic acid; T-C18:1(n-7): Vaccenic acid; CLA: Cis-9: Trans-11 Conjugated linoleic acid.

Table 2. Univariate regression analyses and multivariate models to predict dairy fat with circulating fatty acids in plasma TG, plasma PL and a combination of the two in the Lifelines Cohort.

	Model 1			Model 2		Model 3
Plasma TG	Std β	R^2	p-value	Std β	p-value	Std β	p-value
TG C14:0	0.204	0.071	<0.001	0.049	0.27
TG C15:0	0.286	0.111	<0.001	0.133	0.01	0.162	0.001
TG C17:0	0.049	0.032	0.19	0.016	0.67
TG T-C16:1(n-7)	0.233	0.072	<0.001	0.031	0.54
TG T-C18:1(n-7)	0.292	0.115	<0.001	0.162	0.002	0.181	<0.001
TG CLA	0.215	0.076	<0.001	−0.004	0.94
				R^2 =	0.130	R^2 =	0.128
Plasma PL	Std β	R^2	p-value	Std β	p-value	Std β	p-value
PL C14:0	0.214	0.074	<0.001	0.073	0.07	0.074	0.06
PL C15:0	0.215	0.075	<0.001	0.104	0.02	0.085	0.03
PL C17:0	0.132	0.047	<0.001	−0.017	0.70
PL T-C16:1(n-7)	0.130	0.046	<0.001	−0.062	0.13
PL T-C18:1(n-7)	0.269	0.099	<0.001	0.181	<0.001	0.148	<0.001
PL CLA	0.272	0.103	<0.001	0.154	<0.001	0.151	<0.001
				R^2 =	0.146	R^2 =	0.143
Plasma TG + PL				Std β	p-value	Std β	p-value
TG C14:0				0.033	0.57
TG C15:0				0.088	0.15	0.111	0.04
TG C17:0				0.005	0.90
TG T-C16:1(n-7)				0.023	0.66
TG T-C18:1(n-7)				0.095	0.10	0.113	0.04
TG CLA				−0.092	0.07	−0.088	0.08
PL C14:0				0.023	0.64
PL C15:0				0.087	0.06	0.081	0.05
PL C17:0				−0.024	0.62
PL T-C16:1(n-7)				−0.071	0.11	−0.077	0.06
PL T-C18:1(n-7)				0.131	0.01	0.113	0.02
PL CLA				0.151	0.001	0.169	<0.001
				R^2 =	0.160	R^2 =	0.158

Model 1: Univariate linear regression analysis of the fatty acid with dairy fat intake; Model 2: Multivariate model with all fatty acids from our population in one model; Model 3: Multivariate model with the best combination of fatty acids in our population (backwards regression).

C14:0, C17:0, Trans-C16.1(n-7), Trans-C18.1(n-7) and CLA in plasma TG and Trans-C16.1(n-7), Trans-C18.1(n-7) and CLA in plasma PL were transformed before analyses. Dairy fat intake was transformed and adjusted for energy intake according to the residual method. β’s are standardized beta’s.

TG: Triglycerides; PL: Phospholipids: C14:0: Myristic acid; C15:0: Pentadecanoic acid; C17:0: Heptadecanoic acid; T-C16:1(n-7): Trans-Palmitoleic acid; T-C18:1(n-7): Vaccenic acid; CLA, Cis-9: Trans-11 Conjugated linoleic acid.

Table 3. Multivariate models to predict dairy fat with circulating fatty acids in plasma TG, plasma PL and a combination of the two in the Lifelines Cohort based on 1000 bootstraps.

Variable	Beta ± SE	Std (beta)	p-value	95%CI
Plasma TG
TG C15:0	1.752 ± 0.55	0.162	0.001	0.74; 2.86
TG T-C18:1(n-7)	1.798 ± 0.48	0.181	0.001	0.85; 2.71
Plasma PL
PL C14:0	0.538 ± 0.27	0.074	0.04	0.02; 1.07
PL C15:0	1.102 ± 0.47	0.085	0.02	0.19; 2.02
PL T-C18:1(n-7)	2.032 ± 0.52	0.148	0.001	0.94; 3.05
PL CLA	2.844 ± 0.74	0.151	0.001	1.41; 4.31
Plasma TG + PL
TG C15:0	1.083 ± 0.63	0.100	0.08	−0.20; 2.31
TG T-C18:1(n-7)	1.144 ± 0.59	0.115	0.05	−0.10; 2.56
TG CLA	−1.131 ± 0.67	−0.091	0.09	−2.53; 0.15
PL C15:0	0.912 ± 0.52	0.070	0.08	−0.11; 1.96
PL T-C18:1(n-7)	1.155 ± 0.59	0.084	0.04	0.04; 2.28
PL CLA	3.112 ± 0.78	0.166	0.001	1.56; 4.63

Fatty acids are included in the multivariate model when the circulating fatty acids were present in more than 50% of the models during the bootstrap process based on the backwards stepwise algorithm.

C14:0, C17:0, Trans-C16.1(n-7), Trans-C18.1(n-7) and CLA in plasma TG and Trans-C16.1(n-7), Trans-C18.1(n-7) and CLA in plasma PL were transformed before analyses. Dairy fat intake was transformed and adjusted for energy intake according to the residual method. β’s are standardized beta’s.

TG: Triglycerides; PL: Phospholipids; C14:0: Myristic acid; C15:0: Pentadecanoic acid; C17:0: Heptadecanoic acid; T-C16:1(n-7): T-Palmitoleic acid; T-C18:1(n-7): Vaccenic acid; CLA, Cis-9, Trans-11 Conjugated linoleic acid.

Figure 2. Bland-Altman plots for the agreement between the observed intake (FFQ-derived) and the estimated intake (based on the best circulating fatty acids in plasma TG) for the measurement of dairy fat intake in the general population. Plots of the differences between the observed intake (FFQ-derived) and the estimated intake (based on the best circulating fatty acids in plasma TG) vs the mean of the observed intake (FFQ-derived) and the estimated intake (based on the best circulating fatty acids in plasma TG) method, with (a) representing C15:0 and (b) Trans-C18:1(n-7) as the estimated intake. The red line displays the mean of the difference. The black dotted lines represents the limits of agreement (lower limit: mean - SD*1.96; upper limit: mean + SD*1.96). C15:0: Pentadecanoic acid; FFQ: food frequency questionnaire; SD: standard deviation; TG: triglycerides; Trans-C18:1(n-7): Vaccenic acid.

Figure 3. Bland-Altman plots for the agreement between the observed intake (FFQ-derived) and the estimated intake (based on the best circulating fatty acids in plasma PL) for the measurement of dairy fat intake in the general population. Plots of the differences between the observed intake (FFQ-derived) and the estimated intake (based on the best circulating fatty acids in plasma PL) vs the mean of the observed intake (FFQ-derived) and the estimated intake (based on the best circulating fatty acids in plasma PL), with (a) representing C14:0, (b) C15:0, (c) Trans-C18:1(n-7) and (d) CLA as the estimated intake. The red line displays the mean of the difference. The black dotted lines represents the limits of agreement (lower limit: mean - SD*1.96; upper limit: mean + SD*1.96). C14:0: Myristic acid; C15:0: Pentadecanoic acid; CLA: Conjugated Linoleic; FFQ: food frequency questionnaire; PL: phospholipids; SD: standard deviation; Trans-C18:1(n-7): Vaccenic acid.