Insulin resistance: looking back, looking forward

Understanding the role of insulin resistance in diabetes has followed a fascinating path, with more than 20 000 matches in in the PubMed (https://www.ncbi.nlm.nih.gov/pubmed/) database having “insulin resistance” as a title term, increasing progressively from the early 1940s (Figure 1). The first such reference, a commentary in the British Medical Journal, stated, “the most characteristic action of insulin is its ability to increase the rate at which the peripheral tissues remove sugar from the blood,”1 an observation as applicable today as 60 years ago. The insulin resistance typically recognized in clinical practice at that time was associated with the impure insulin preparations then available, with administration of serum from patients with insulin resistance shown to prevent the glucose-lowering effect of insulin in mice, a phenomenon taken as evidence of a relationship to the presence of antibodies to insulin,2 as was the typical pattern of glucose-lowering with insulin requiring extremely high doses, with the insulin resistance appearing in some cases to be transient.3 Resistance to administered insulin could also be seen with insulin injection into areas of lipodystrophy.4 Our modern understanding of the nature of insulin resistance began with the development by Yalow and Berson of the technique of radioimmunoassay and its application to the measurement of insulin levels in blood, allowing recognition of the elevated circulating insulin levels seen in many persons with diabetes.5 This tool led to the understanding that obesity itself is associated with elevation in circulating insulin levels, described by Rabinowitz and Zierler as “a link in the chain of evidence incriminating simple obesity as a precursor of maturity onset stable diabetes.”6 Subsequent work by many groups allowed more precise characterization of the phenomenon of insulin resistance in prediabetes7 and in obesity,8 with Reaven integrating what had become a large body of evidence into the recognition that hyperinsulinemia and insulin resistance are associated with a variety of conditions including not only diabetes but also dyslipidemia and hypertension, which he called “syndrome X,”9 leading to the recognition of what is now described as the metabolic syndrome.10 Although we can readily recognize the syndrome, determining useful approaches to quantitation of insulin resistance for clinical practice has been elusive - despite the long history of studies such as those from Reaven's group showing that markers such as the steady state plasma glucose after intravenous insulin and glucose can be estimated either from the serum insulin or from the triglyceride level or triglyceride to high-density lipoprotein (HDL) cholesterol ratio.11 Higher insulin (and triglyceride) levels are predictive of diabetes risk several decades subsequently.12 As a further example of the potential of such an approach, in a meta-analysis of the use of rosiglitazone, risk of myocardial infarction appeared greater with rosiglitazone treatment in groups with triglyceride levels below 150 mg/dL, whereas rosiglitazone appeared to be associated with reduction in myocardial infarction in those with higher triglyceride levels, who presumably had greater degrees of insulin resistance and hence greater benefit from the insulin sensitizing effect of the thiazolidinedione.13 What of current studies of insulin resistance? More than one thousand studies per year have “insulin resistance” as a title term beginning in 2011, so any review of the topic is necessarily limited. Both the Homeostatic Model for Insulin Resistance (HOMA-IR), calculated from the product of serum insulin and glucose levels, and a triglyceride-based measure were found to be associated with the presence of coronary artery disease (CAD) on computerized tomographic angiography in persons not having diabetes, and HOMA-IR was also associated with CAD among persons with diabetes.14 Among more than 20 000 persons evaluated in the National Health and Nutrition Examination Survey (NHANES) population-based survey, HOMA-IR, adjusted for a variety of measures including the degree of obesity, was associated with general health condition, with the triglyceride and triglyceride-HDL ratio, with hepatic enzymes, with all the parameters of the complete blood count, and (inversely) with levels of vitamins B6, C, D, and folic acid;15 interestingly, among nondiabetic but not diabetic persons in NHANES, HOMA-IR was associated with diabetic retinopathy.16 Insulin resistance has been shown to be associated with abnormal macrophage function, leading to decreased bactericidal response in a peritonitis model.17 Insulin resistance has also been demonstrated in animal models in cutaneous tissue, in association with impaired barrier function, with increased transepidermal water loss and altered keratin and cell cycle regulatory molecule expression, as well as with increased subcutaneous adipose tissue monocyte chemoattractant protein-1 and with macrophage infiltration of the subcutaneous tissue.18 Strikingly, insulin resistance is associated with worsening of cognitive function. In a study of persons with serum insulin measurement at age 55, elevated HOMA-IR was associated with decreased executive function and cognitive processing speed at age 70.19 This and many similar studies have led to the concept that insulin resistance may be an important explanatory factor in the development of Alzheimer's disease,20 with much evidence pointing to the hippocampus as an insulin responsive tissue playing a major role in memory and involved in age- and diabetes-related cognitive issues.21 In this regard, it is fascinating that insulin resistance may have epigenetic effects on expression of brain-derived neurotrophic factor, with the potential for multigenerational impairment of memory mechanisms being caused by the insulin resistant state.22 The relationship of insulin resistance to obesity has been well demonstrated, and in this regard the progressive increase in obesity prevalence is of great concern.23 New approaches to insulin sensitization involving leptin,24perhaps involving non-thiazolidinedione agents directed at targets such as the PPARγ inhibitor transcriptional co-activator with PDZ-binding (TAZ),25 or involving metabolomic abnormalities associated with insulin resistance such as branched chain amino acids26 and bioactive lipids,27 may ultimately allow us to address this complex and far-reaching set of metabolic disturbances. 我们对糖尿病中胰岛素抵抗的理解经历了一条极为有趣的路径, 在PubMed (https://www.ncbi.nlm.nih.gov/pubmed/) 数据库中题目包含“胰岛素抵抗”的文章超过20000篇, 并且从1940年代早期开始迅速增加。 第一篇关于胰岛素的研究源自60年前发表在British Medical Journal上的文章, 陈述“胰岛素最典型的作用是加速外周组织从血液中去除糖的功能” , 这个观点也一直沿用至今。胰岛素抵抗最早在临床中被意识到是与当时不纯的胰岛素制剂有关, 将胰岛素抵抗患者的血清用于小鼠实验中,发现胰岛素的降糖作用在小鼠体内被抑制, 这种现象被视为胰岛素抗体存在的证据, 因此需要使用剂量极高的胰岛素才能达到降血糖的效果。 当代对胰岛素抵抗的理解始于Yalow和Berson发明的放射免疫分析技术, 并用此发现了糖尿病患者外周血中升高的胰岛素水平。这一技术也让学者们发现了肥胖与胰岛素水平升高存在关联, 并且将肥胖称之为糖尿病的始动因素。随后的研究能够更精确地描述胰岛素抵抗在糖尿病前期和肥胖中的现象。Reaven提出了著名的“X综合征”概念, 即胰岛素抵抗不仅存在于糖尿病, 还出现在高脂血症、高血压等疾病中, 而“X综合征”就是我们如今的“代谢综合征”。 目前关于胰岛素抵抗的研究都有哪些呢?从2011年起每年有超过一千项含有“胰岛素抵抗”标题的研究, 因此任何关于此话题的回顾都必然是不全面的。研究发现即使没有糖尿病, 胰岛素抵抗指数(HOMA-IR)和甘油三酯水平升高也会加大冠心病的发病风险, 而已患有糖尿病的患者HOMA-IR升高更是会加大冠心病的发病风险。基于超过2万人的国家健康和营养调查(NHANES)显示, 校正包括肥胖程度在内的各种参数后, HOMA-IR与健康状况、甘油三酯、甘油三酯/高密度脂蛋白的比值、肝酶水平和全血细胞计数成正相关, 而与维生素B6、C、D和叶酸的水平呈负相关; 在非糖尿病患者中, HOMA-IR与糖尿病视网膜病变有关。研究发现胰岛素抵抗与巨噬细胞功能异常相关, 从而降低腹膜炎患者的杀菌能力。此外, 胰岛素抵抗与认知功能障碍恶化也有关系, 研究发现HOMA-IR在年龄55岁时升高会导致老年人在70岁时执行能力和认知能力下降, 因此可能与阿尔茨海默氏病的发生发展相关。 目前, 胰岛素抵抗与肥胖的关系已得到充分证明, 就这一点而言, 肥胖患病率的快速增长应该引起广泛关注。新型的胰岛素增敏剂包括瘦素、非噻唑烷二酮或支链氨基酸和生物活性脂质等与胰岛素抵抗有关的代谢异常产物, 这或许最终可以帮助我们解决这一复杂而影响深远的难题。