Marklund Matti, Pingel Ronnie, Rosqvist Fredrik, Lindroos Anna Karin, Eriksson Jan W, Vessby Bengt, Oscarsson Jan, Lind Lars, Risérus Ulf
Department of Public Health and Caring Sciences, Clinical Nutrition and Metabolism Unit.
National Food Agency, Uppsala, Sweden.
J Nutr. 2017 Nov;147(11):2118-2125. doi: 10.3945/jn.117.254250. Epub 2017 Sep 20.
Fatty acid (FA) proportions in cholesterol esters (CEs) and plasma phospholipids are widely used as dietary biomarkers. Information on how proportions in these fractions correlate could have implications for interpretation and use of FA biomarkers in observational and interventional studies. We investigated correlations between FA proportions in CEs and phospholipids in free-living individuals and assessed how diet-induced alterations of FA proportions correlate between fractions. Spearman's rank correlation coefficients () between FA proportions (percentage of total FAs) in circulating CEs and phospholipids were calculated separately in 8 individual study populations including Swedish females and males ( = 2052; age range: 11-84 y), and pooled by inverse-variance weighted meta-analysis. In addition, study populations were stratified by age, sex, body mass index (BMI; in kg/m), and diabetes status, and strata-specific were pooled by meta-analysis. In 2 randomized trials ( = 79) in which dietary saturated FAs were isocalorically replaced with unsaturated FAs, treatment-wise calculations of were conducted between FA changes in CEs and phospholipids. Overall, FA proportions in CEs and phospholipids correlated well and especially strongly for polyunsaturated FAs (PUFAs), with pooled (95% CIs) ranging from 0.74 (0.72, 0.76) for α-linolenic acid to 0.92 (0.91, 0.93) for eicosapentaenoic acid. Weak correlations (pooled < 0.4) were observed only for palmitic acid and stearic acid, with pooled (95% CIs): 0.29 (0.24, 0.33) and 0.30 (0.25, 0.34), respectively. Overall, correlations were not affected by age, sex, BMI, or diabetes status. Strong correlations ( ≥ 0.6) between diet-induced FA changes in CEs and phospholipids were observed for most PUFAs. Proportions of most FAs in CEs and phospholipids ranked individuals similarly, suggesting that FA proportions in these fractions can be used interchangeably in populations of diverse age, sex, body composition, and diabetes status. Caution is advised, however, when comparing results from studies assessing palmitic acid or stearic acid in different lipid fractions.
胆固醇酯(CEs)和血浆磷脂中的脂肪酸(FA)比例被广泛用作膳食生物标志物。关于这些组分中比例如何相互关联的信息可能对观察性研究和干预性研究中FA生物标志物的解释和应用具有重要意义。我们研究了自由生活个体中CEs和磷脂中FA比例之间的相关性,并评估了饮食诱导的FA比例变化在不同组分之间的相关性。在包括瑞典女性和男性在内的8个个体研究人群(n = 2052;年龄范围:11 - 84岁)中,分别计算了循环CEs和磷脂中FA比例(占总FAs的百分比)之间的Spearman等级相关系数(ρ),并通过逆方差加权荟萃分析进行汇总。此外,研究人群按年龄、性别、体重指数(BMI;单位:kg/m²)和糖尿病状态进行分层,并通过荟萃分析汇总各层特定的ρ。在2项随机试验(n = 79)中,用不饱和脂肪酸等热量替代膳食饱和脂肪酸,对CEs和磷脂中FA变化之间进行了按治疗分组的ρ计算。总体而言,CEs和磷脂中的FA比例相关性良好,尤其是多不饱和脂肪酸(PUFAs),汇总的ρ(95%置信区间)范围从α-亚麻酸的0.74(0.72,0.76)到二十碳五烯酸的0.92(0.91,0.93)。仅棕榈酸和硬脂酸观察到弱相关性(汇总的ρ < 0.4),汇总的ρ(95%置信区间)分别为0.29(0.24,0.33)和0.30(0.25,0.34)。总体而言,相关性不受年龄、性别、BMI或糖尿病状态的影响。对于大多数PUFAs,观察到饮食诱导的CEs和磷脂中FA变化之间存在强相关性(ρ≥0.6)。CEs和磷脂中大多数FA的比例对个体的排序相似,这表明在不同年龄、性别、身体组成和糖尿病状态的人群中,这些组分中的FA比例可以互换使用。然而,在比较评估不同脂质组分中棕榈酸或硬脂酸的研究结果时,建议谨慎。