Retina Department, Vision Research Foundation, Sankara Nethralaya, Chennai, India.
Nightingale-Saunders Clinical Trials and Epidemiology Unit, King's College London, London, UK.
Lancet Glob Health. 2022 Dec;10(12):e1764-e1773. doi: 10.1016/S2214-109X(22)00411-9. Epub 2022 Oct 31.
National and subnational estimates of the prevalence of diabetic retinopathy and vision-threatening diabetic retinopathy (VTDR) are needed to inform the stepwise implementation of systematic retinal screening for people with diabetes in India to decrease the rate of blindness. We aimed to assess these national and subnational estimates and to stratify the prevalence of diabetic retinopathy and VTDR on the basis of people with known versus undiagnosed diabetes, urban versus rural residence, and epidemiological transition level (ETL) and Socio-demographic Index (SDI) categories of states.
We did a multicentre cross-sectional screening study for diabetic retinopathy using a complex cluster sampling design in people aged 40 years or older in ten Indian states and one union territory between Dec 20, 2018, and March 20, 2020. We did non-mydriatic retinal screening and assessed risk factor burden for people with diabetes. We estimated nationally weighted prevalence of diabetic retinopathy and VTDR for individuals with known and undiagnosed diabetes by urban versus rural residence, and by state categorisation by ETL and SDI. We also assessed adjusted risk factors.
From 42 146 participants screened, 7910 (18·8%) were identified to have diabetes. Of these, 6133 (77·5%; 4350 with known diabetes and 1783 with undiagnosed diabetes) had gradable retinal images. 3411 (56%) participants were women and 2722 (44%) were men, and the median age was 56 years (IQR 49-65). The estimated national prevalence was 12·5% (95% CI 11·0-14·2) for diabetic retinopathy and 4·0% (3·4-4·8) for VTDR, with no significant differences between urban and rural residence for diabetic retinopathy. Compared with individuals with undiagnosed diabetes, we observed a higher prevalence of diabetic retinopathy (15·5% [13·4-17·8] vs 8·0% [6·3-10·1]) and VTDR (5·3% [4·5-6·3] vs 2·4% [1·6-3·6]) in individuals with known diabetes. The prevalence was significantly lower in low ETL-SDI states compared with high and middle ETL-SDI states for diabetic retinopathy (by 7·0%, 1·9-12·2, p=0·024) and VTDR (by 4·8%, 3·0-6·6, p<0·0001). Hyperglycaemia was the strongest modifiable risk factor.
We estimate that, in absolute numbers, approximately 3 million people aged 40 years or older have VTDR in India, with a higher prevalence in those with known diabetes residing in high and middle ETI-SDI states.
UKRI Global Challenge Research Fund.
为了在印度逐步实施针对糖尿病患者的系统视网膜筛查，以降低失明率，我们需要对全国和次国家的糖尿病视网膜病变和威胁视力的糖尿病视网膜病变（VTDR）的患病率进行评估。我们旨在评估这些全国和次国家的估计数，并根据已知与未确诊糖尿病、城乡居住、以及州的流行病学过渡水平（ETL）和社会人口指数（SDI）类别，对糖尿病视网膜病变和 VTDR 的患病率进行分层。
我们在 2018 年 12 月 20 日至 2020 年 3 月 20 日期间，在印度的十个州和一个联邦属地，对 40 岁及以上的人群进行了一项基于多中心的糖尿病视网膜病变横断面筛查研究，采用复杂的聚类抽样设计。我们进行了非散瞳视网膜筛查，并对糖尿病患者的危险因素负担进行了评估。我们根据城乡居住、州的 ETL 和 SDI 分类，对已知和未确诊糖尿病患者的糖尿病视网膜病变和 VTDR 的全国加权患病率进行了估计。我们还评估了调整后的危险因素。
在接受筛查的 42146 名参与者中，有 7910 人（18.8%）被诊断为糖尿病。其中，6133 人（77.5%；4350 人有已知糖尿病，1783 人有未确诊糖尿病）有可分级的视网膜图像。3411 名参与者为女性，2722 名参与者为男性，中位年龄为 56 岁（IQR 49-65）。全国糖尿病视网膜病变的估计患病率为 12.5%（95%CI 11.0-14.2），VTDR 的患病率为 4.0%（3.4-4.8），城乡居住对糖尿病视网膜病变的患病率没有显著影响。与未确诊的糖尿病患者相比，我们观察到在已知糖尿病患者中，糖尿病视网膜病变（15.5%[13.4-17.8]vs 8.0%[6.3-10.1]）和 VTDR（5.3%[4.5-6.3]vs 2.4%[1.6-3.6]）的患病率更高。在低 ETL-SDI 州，糖尿病视网膜病变（降低 7.0%，1.9-12.2，p=0.024）和 VTDR（降低 4.8%，3.0-6.6，p<0.0001）的患病率明显低于高和中 ETL-SDI 州。高血糖是最强的可改变危险因素。
我们估计，在绝对数字上，印度约有 300 万 40 岁及以上的人患有 VTDR，在高和中 ETI-SDI 州，有已知糖尿病的居民的患病率更高。
英国研究与创新署全球挑战研究基金。
