Standardising corneal nerve fibre length for nerve tortuosity increases its association with measures of diabetic neuropathy

AIMS: Recent studies on corneal markers have advocated corneal nerve fibre length as the most important measure of diabetic peripheral neuropathy. The aim of this study was to determine if standardizing corneal nerve fibre length for tortuosity increases its association with other measures of diabetic peripheral neuropathy. METHODS: Two hundred and thirty-one individuals with diabetes with either predominantly mild or absent neuropathic changes and 61 control subjects underwent evaluation of diabetic neuropathy symptom score, neuropathy disability score, testing with 10-g monofilament, quantitative sensory testing (warm, cold, vibration detection) and nerve conduction studies. Corneal nerve fibre length and corneal nerve fibre tortuosity were measured using corneal confocal microscopy. A tortuosity-standardised corneal nerve fibre length variable was generated by dividing corneal nerve fibre length by corneal nerve fibre tortuosity. Differences in corneal nerve morphology between individuals with and without diabetic peripheral neuropathy and control subjects were determined and associations were estimated between corneal morphology and established tests of, and risk factors for, diabetic peripheral neuropathy. RESULTS: The tortuosity-standardised corneal nerve fibre length variable was better than corneal nerve fibre length in demonstrating differences between individuals with diabetes, with and without neuropathy (tortuosity-standardised corneal nerve fibre length variable: 70.5 ± 27.3 vs. 84.9 ± 28.7, P < 0.001, receiver operating characteristic area under the curve = 0.67; corneal nerve fibre length: 15.9 ± 6.9 vs. 18.4 ± 6.2 mm/mm(2) , P = 0.004, receiver operating characteristic area under the curve = 0.64). Furthermore, the tortuosity-standardised corneal nerve fibre length variable demonstrated a significant difference between the control subjects and individuals with diabetes, without neuropathy, while corneal nerve fibre length did not (tortuosity-standardised corneal nerve fibre length variable: 94.3 ± 27.1 vs. 84.9 ± 28.7, P = 0.028; corneal nerve fibre length: 20.1 ± 6.3 vs. 18.4 ± 6.2 mm/mm(2) , P = 0.084). Correlations between corneal nerve fibre length and established measures of neuropathy and risk factors for neuropathy were higher when a correction was made for the nerve tortuosity. CONCLUSIONS: Standardizing corneal nerve fibre length for tortuosity enhances the ability to differentiate individuals with diabetes, with and without neuropathy.

Application of higher-order spectra for automated grading of diabetic maculopathy

Diabetic macular edema (DME) is one of the most common causes of visual loss among diabetes mellitus patients. Early detection and successive treatment may improve the visual acuity. DME is mainly graded into non-clinically significant macular edema (NCSME) and clinically significant macular edema according to the location of hard exudates in the macula region. DME can be identified by manual examination of fundus images. It is laborious and resource intensive. Hence, in this work, automated grading of DME is proposed using higher-order spectra (HOS) of Radon transform projections of the fundus images. We have used third-order cumulants and bispectrum magnitude, in this work, as features, and compared their performance. They can capture subtle changes in the fundus image. Spectral regression discriminant analysis (SRDA) reduces feature dimension, and minimum redundancy maximum relevance method is used to rank the significant SRDA components. Ranked features are fed to various supervised classifiers, viz. Naive Bayes, AdaBoost and support vector machine, to discriminate No DME, NCSME and clinically significant macular edema classes. The performance of our system is evaluated using the publicly available MESSIDOR dataset (300 images) and also verified with a local dataset (300 images). Our results show that HOS cumulants and bispectrum magnitude obtained an average accuracy of 95.56 and 94.39 % for MESSIDOR dataset and 95.93 and 93.33 % for local dataset, respectively.

Diabetes in South Asians

Economic, dietary and other lifestyle transitions have been occurring rapidly in most South Asian countries, making their populations more vulnerable to developing Type 2 diabetes and cardiovascular diseases. Recent data show an increasing prevalence of Type 2 diabetes in urban areas as well as in semi-urban and rural areas, inclusive of people belonging to middle and low socio-economic strata. Prime determinants for Type 2 diabetes in South Asians include physical inactivity, imbalanced diets, abdominal obesity, excess hepatic fat and, possibly, adverse perinatal and early life nutrition and intra-country migration. It is reported that Type 2 diabetes affects South Asians a decade earlier and some complications, for example nephropathy, are more prevalent and progressive than in other races. Further, prevalence of pre-diabetes is high, and so is conversion to diabetes, while more than 50% of those who are affected remain undiagnosed. Attitudes, cultural differences and religious and social beliefs pose barriers in effective prevention and management of Type 2 diabetes in South Asians. Inadequate resources, insufficient healthcare budgets, lack of medical reimbursement and socio-economic factors contribute to the cost of diabetes management. The challenge is to develop new translational strategies, which are pragmatic, cost-effective and scalable and can be adopted by the South Asian countries with limited resources. The key areas that need focus are: generation of awareness, prioritizing health care for vulnerable subgroups (children, women, pregnant women and the underprivileged), screening of high-risk groups, maximum coverage of the population with essential medicines, and strengthening primary care. An effective national diabetes control programme in each South Asian country should be formulated, with these issues in mind.

Glycaemic behaviour during breastfeeding in women with Type 1 diabetes

Aim To describe glycaemia in both breastfeeding women and artificially feeding women with Type 1 diabetes, and the changes in glycaemia induced by suckling. Methods A blinded continuous glucose monitor was applied for up to 6 days in eight breastfeeding and eight artificially feeding women with Type 1 diabetes 2–4 months postpartum. Women recorded glucose levels, insulin dosages, oral intake and breastfeeding episodes. A standardized breakfast was consumed on 2 days. A third group (clinic controls) were identified from a historical database. Results Carbohydrate intake tended to be higher in breastfeeding than artificially feeding women (P = 0.09) despite similar insulin requirements. Compared with breastfeeding women, the high blood glucose index and standard deviation of glucose were higher in artificially feeding women (P = 0.02 and 0.06, respectively) and in the clinical control group (P = 0.02 and 0.05, respectively). The low blood glucose index and hypoglycaemia were similar. After suckling, the low blood glucose index increased compared with before (P < 0.01) and during (P < 0.01) suckling. Hypoglycaemia (blood glucose < 4.0 mmol/l) occurred within 3 h of suckling in 14% of suckling episodes, and was associated with time from last oral intake (P = 0.04) and last rapid-acting insulin (P = 0.03). After a standardized breakfast, the area under the glucose curve was positive. In breastfeeding women the area under the glucose curve was positive if suckling was avoided for 1 h after eating and negative if suckling occurred within 30 min of eating. Conclusions Breastfeeding women with Type 1 diabetes had similar hypoglycaemia but lower glucose variability than artificially feeding women. Suckling reduced maternal glucose levels but did not cause hypoglycaemia in most episodes.