Department of Surgical Sciences, Uppsala University, Uppsala, Sweden.
Unit of Cardiovascular and Nutritional Epidemiology, Institute of Environmental Medicine, Karolinska Institutet, Nobelsväg 13, 17177, Stockholm, Sweden.
Diabetologia. 2020 Jan;63(1):116-123. doi: 10.1007/s00125-019-05019-0. Epub 2019 Nov 5.
AIMS/HYPOTHESIS: Epidemiological data on the associations of circulating fatty acid levels with type 2 diabetes are inconsistent. We conducted a two-sample Mendelian randomisation study to explore the causal associations of plasma levels of ten fatty acids with type 2 diabetes and glycaemic traits.
Thirteen SNPs associated with circulating levels of ten individual fatty acids at the genome-wide significance level (p < 5 × 10) were selected as instrumental variables for the exposures. For the outcomes, summary-level data were obtained from the DIAbetes Genetics Replication And Meta-analysis (DIAGRAM) consortium for type 2 diabetes (898,130 individuals) and from the Meta-Analyses of Glucose and Insulin-related traits Consortium (MAGIC) for the glycaemic traits (up to 46,186 non-diabetic individuals). The inverse-variance weighted method was used for analyses.
Genetic predisposition to higher plasma levels of eight of the ten fatty acids were statistically significantly associated with lower or higher odds of type 2 diabetes. The OR per one SD increment of each fatty acid was 0.93 (95% CI 0.90, 0.96; p = 2.21 × 10) for α-linolenic acid, 0.96 (95% CI 0.94, 0.98; p = 1.85 × 10) for linoleic acid, 0.86 (95% CI 0.81, 0.91; p = 6.68 × 10) for palmitoleic acid, 0.87 (95% CI 0.81, 0.93; p = 2.21 × 10) for oleic acid, 1.08 (95% CI 1.03, 1.12; p = 0.002) for eicosapentaenoic acid, 1.04 (95% CI 1.02, 1.07; p = 0.001) for docosapentaenoic acid, 1.03 (95% CI 1.02, 1.05; p = 2.51 × 10) for arachidonic acid and 1.09 (95% CI 1.03, 1.15; p = 0.003) for stearic acid. The same eight fatty acids were also associated with fasting glucose levels and HOMA-B. The associations, except that for palmitoleic acid, were driven by variants in FADS1/2.
CONCLUSIONS/INTERPRETATION: Genetic predisposition to higher circulating levels of eight out of ten fatty acids was associated with type 2 diabetes, fasting glucose and islet beta cell function. However, the associations, except that for palmitoleic acid, were driven by variants in FADS1/2, which encode enzymes with a key role in fatty acid metabolism.
目的/假设：关于循环脂肪酸水平与 2 型糖尿病之间关联的流行病学数据不一致。我们进行了两样本孟德尔随机化研究，以探讨血浆中十种脂肪酸水平与 2 型糖尿病和糖代谢特征之间的因果关系。
选择与十种个体脂肪酸循环水平相关的 13 个单核苷酸多态性（SNP），达到全基因组显著水平（p < 5 × 10），作为暴露的工具变量。对于结局，来自糖尿病遗传学复制和荟萃分析（DIAGRAM）联盟的 2 型糖尿病（898130 人）和葡萄糖和胰岛素相关特征荟萃分析联盟（MAGIC）的糖代谢特征（最多 46186 名非糖尿病个体）获得汇总水平数据。采用逆方差加权法进行分析。
十种脂肪酸中八种的血浆水平遗传易感性与 2 型糖尿病的低或高风险呈统计学显著相关。每种脂肪酸每增加一个标准差的 OR 为 0.93（95%CI 0.90，0.96；p = 2.21 × 10），α-亚麻酸为 0.96（95%CI 0.94，0.98；p = 1.85 × 10），亚油酸为 0.86（95%CI 0.81，0.91；p = 6.68 × 10），棕榈油酸为 0.87（95%CI 0.81，0.93；p = 2.21 × 10），油酸为 1.08（95%CI 1.03，1.12；p = 0.002），二十碳五烯酸为 1.04（95%CI 1.02，1.07；p = 0.001），二十二碳五烯酸为 1.03（95%CI 1.02，1.05；p = 2.51 × 10），花生四烯酸为 1.03（95%CI 1.02，1.05；p = 2.51 × 10），硬脂酸为 1.09（95%CI 1.03，1.15；p = 0.003）。同样的八种脂肪酸也与空腹血糖水平和 HOMA-B 相关。除了棕榈油酸，这些关联都由 FADS1/2 中的变体驱动。
结论/解释：十种脂肪酸中八种的循环水平遗传易感性与 2 型糖尿病、空腹血糖和胰岛β细胞功能有关。然而，除了棕榈油酸，这些关联都由 FADS1/2 中的变体驱动，该基因编码在脂肪酸代谢中起关键作用的酶。
