- Email: [email protected]
Cardiovascular Disease Guideline Adherence and Statin Use: Results from the Canadian Study of Longevity in Diabetes
542
047-2015.
Abstracts / Can J Diabetes 39 (2015) 529e547
GLUCOSE AND METABOLISM
Upper Small Intestinal Protein Sensing Improves Glucose Tolerance Through Suppression of Hepatic Glucose Production SOPHIE HAMR*, BRITTANY A. RASMUSSEN, PAIGE V. BAUER, FRANK A. DUCA, TONY K.T. LAM Toronto, ON Intestinal lipid-sensing initiates a neuronal gut-brain-liver negative feedback pathway to lower glucose production (GP) and maintain glucose homeostasis in vivo. However, whether intestinal protein-sensing regulates GP and consequently glucose tolerance via a neuronal network remains unknown, despite the ability of high-protein feeding to improve hyperglycemia and glucose tolerance in healthy and diabetic rodents and humans. Thus, we investigated the effects of upper small intestinal protein-sensing on glucose homeostasis by first infusing intraduodenal 10% w/v casein hydrolysate (CH) during an intravenous glucose tolerance test (IVGTT) in healthy rats. Intestinal infusion of CH for 50 min improved glucose tolerance as compared to saline (AUC saline: 1186.4106.8 vs. CH: 340.29128.96, n¼8, p<0.05), independent of a rise in plasma amino acids. Thus, this protein-induced glucose-lowering effect was preabsorptive. To assess whether intestinal protein-sensing improves glucose tolerance by directly lowering GP or increasing glucose uptake (GU) independent of changes in circulating glucoregulatory hormones, we infused intestinal CH during a pancreatic (basalinsulin) euglycemic clamp and evaluated changes in glucose kinetics using tracer-dilution methodology. Intestinal CH lowered GP (% suppression from basal; saline: -18.24.9% vs. CH: 463.2%, n¼8, p<0.05) compared to saline-treated rats, while GU remained unchanged (basal GU: 13.20.7 mgkg-1min-1 vs. CH GU: 12.60.7 mgkg-1min-1, n¼8, p>0.05). These findings collectively suggest that upper small intestinal protein-sensing improves glucose tolerance by lowering GP in healthy rodents. Future studies will address the underlying preabsorptive signaling mechanisms, the potential involvement of a neuronal gut-brain-liver axis, and the capacity of intestinal proteinsensing in models of obesity and diabetes.
048-2015.
CLINICAL CARE
Health Literacy Among Patients with Type 2 Diabetes Attending an Ambulatory Diabetes Services in Ireland YVONNE FINN*, SITI S. MOHD FARUDZ, MUHAMMAD A. MAD DAN Galway, Ireland Background: The institute of Medicine (2004) defines health literacy as “The degree to which individuals have the capacity to obtain, process and understand basic health information and service needed to make appropriate health decisions”. Sub-optimal health literacy has been described as a risk factor for chronic diseases such as type 2 diabetes. It is also linked with poorer glycemic control and higher complication rates. There is no data on health literacy in patients with type 2 diabetes in Ireland. This study measures health literacy in patients with type 2 diabetes attending the ambulatory diabetes services at Galway University hospital in Ireland. Methods: Patients with type 2 diabetes attending the out-patient diabetes clinic over a 4 week period in June/July 2015 were invited to participate. Functional health literacy was measured by administering the Newest Vital Sign (NVS), a validated tool to measure health literacy. Ethics approval was obtained from Galway University Hospitals’ Ethics committee. Results: Sixty seven participants, ranging from 39 to 84 years, completed the study. Twenty seven (40%) had a high likelihood of limited functional health literacy, 16 (24%) a possible likelihood of limited health literacy and 24 (36%) had adequate health literacy.
Only 2 participants (3%) perceived their level of health literacy level as low. Conclusions: Just over one third of participants in this study had adequate health literacy. Most of the participants with a high likelihood of limited health literacy did not perceive their health literacy skills as being inadequate.
049-2015. CLINICAL CARE Validation of Cooling Detection Threshold as a Marker of Sensorimotor Polyneuropathy in Type 2 Diabetes MOHAMMED A. FAROOQI, LEIF ERIK LOVBLOM, ZOE LYSY, ELISE M. HALPERN, MYLAN NGO, EDUARDO NG, ANDREJ ORSZAG, ARI BREINERy, VERA BRILy, BRUCE A. PERKINS*y Toronto, ON Aim: The measurement of cooling detection thresholds (CDT) has been established in previous cross-sectional studies as a valid test for diabetic sensorimotor polyneuropathy (DSP) in type 1 diabetes. We aimed to validate its diagnostic performance in type 2 diabetes (T2D). Methods: 220 T2D subjects from a larger cohort underwent clinical and electrophysiological examinations including 3 small-fiber function tests: CDT, heart rate variability (HRV) and laser Doppler imaging of axon-mediated neurogenic flare responses to cutaneous heating (LDIFLARE), along with the Toronto Clinical Neuropathy Score (TCNS). Clinical DSP was defined by consensus criteria whereas preclinical DSP was defined by at least one electrophysiological abnormality. Area under the curve (AUC) and optimal thresholds were determined by receiver operating characteristic (ROC) curves. Results: Subjects were aged 6311 years with mean HbA1c of 7.51.6%. The 139(63%) clinical DSP cases had mean CDT value 18.38.9 C; the 52(24%) preclinical DSP had 25.33.5 C; and the 29(13%) controls had 27.13.8 C; (p-value<0.02 for all three comparisons). For identification of clinical DSP AUCCDT was 0.79, which exceeded AUCHRV (0.60, p¼<0.0001), AUCLDI FLARE (0.69, p¼0.0003) and AUCTCNS (0.73, p¼0.03). Optimal threshold for clinical DSP identification was <22.8 C (64% sensitivity and 83% specificity). For Preclinical DSP, AUCCDT was 0.80, and the optimal threshold was 27.5 C (83% sensitivity and 72% specificity). Conclusions: Akin to studies of T1D, CDT has acceptable diagnostic performance for the identification of both clinical and preclinical neuropathy in patients with T2D. Application of CDT as a noninvasive tool for systematic screening of early neuropathy in diabetes clinics should be considered.
050-2015. LIPIDS Cardiovascular Disease Guideline Adherence and Statin Use: Results from the Canadian Study of Longevity in Diabetes JOHNNY-WEI BAI, GENEVIÈVE BOULET, ELISE M. HALPERN, LEIF ERIK LOVBLOM, ALANNA WEISMAN, DEVRIM ELDELEKLI, HILLARY KEENAN, MICHAEL BRENT, NARINDER PAUL, VERA BRILy, DAVID CHERNEYy, BRUCE A. PERKINS*y Toronto, ON; Boston, US Background: Older patients with longstanding type 1 diabetes (T1DM) have high cardiovascular disease (CVD) risk such that statin therapy is recommended independent of prior CVD events. We aimed to determine adherence to CVD prevention guidelines in patients with 50 or more years of diabetes. Research design and methods: 309 Canadians with over 50 years of T1DM completed a medical questionnaire for presence of lifestyle and CVD pharmacological interventions and they were
Abstracts / Can J Diabetes 39 (2015) 529e547
stratified into primary or secondary CVD prevention subgroups based on self-report CVD events. Associations with statin use were analyzed using multivariable logistic regression. Results: The 309 participants were 65.78.5years, with median diabetes duration of 55.0[51.0,59.0]years and A1C of 7.40 [6.80,8.00]%. 159(52.7%) participants reported adhering to diet recommendations, 296(95.8%) smoking avoidance, 217(70.5%) physical activity,, 218(71.5%) renin-angiotensin system blocker use, and 220(72.1%) statin use. Other than physical activity - reported as less common in the secondary prevention subgroup - current statin therapy was significantly lower in the primary prevention subgroup (65.5% vs. 84.8%, p¼0.0004). In multivariable logistic regression, the odds of statin use was 0.38(0.15,0.92) in members of the primary compared to the secondary prevention cohort, adjusting for age, microvascular complications, A1c, blood pressure, cholesterol, and body mass. Conclusion: Though guidelines have for years recommended statin use in longstanding type 1 diabetes regardless of CVD history, its use was substantially lower in those without compared to those with CVD history. Interventions are needed to improve adherence in the high-risk population without CVD history.
051-2015.
CLINICAL CARE
Neuropathy Prevalence Compared to other Complications: Preliminary Analysis of the Canadian Study of Longevity in Diabetes Cohort ALANNA WEISMAN, RANDY ROVINSKI, LEIF ERIK LOVBLOM, ELISE M. HALPERN, GENEVIÈVE BOULET, DEVRIM ELDELEKLI, HILLARY KEENAN, MICHAEL BRENT, NARINDER PAUL, VERA BRILy, DAVID CHERNEYy, BRUCE A. PERKINS*y Toronto, ON; Boston, US Aim: Type 1 diabetes (T1DM) subjects can survive for extreme duration without retinopathy or nephropathy whereas estimates of neuropathy are uncertain. In a cohort with 50y of T1DM, we aimed to assess the prevalence of microvascular complications and other indicator variables. Methods: 309 participants were recruited (2013-2014) across Canada by public advertisement and mailings to health professionals. Participants answered a comprehensive questionnaire and provided laboratory results. We assessed complications by eye-specialist fundus examinations, Michigan Neuropathy Screening Instrument 3/15, eGFR <60mL/min and albumin:creatinine >2 mg/mmol, and self-report of coronary artery disease (CAD), gastroparesis and peripheral vascular disease (PVD). Results: In preliminary analysis, 309 participants (55% female) were 668.5 years of age, had mean diabetes duration of 556.5y, BMI 25 (23-28.2) kg/m2, and A1c 7.51.1%. Retinopathy was most prevalent at 78% compared to neuropathy (42%), nephropathy (48%), CAD (34%), gastroparesis (15%) and PVD (10%). 37(12%) lacked evidence of any complication. In a Poisson regression generalized estimating equation (GEE) model, greater number of complications was independently associated with higher HbA1c (RR 1.05; 1.01-1.09), older age (RR 1.10; 1.061.13) and greater diabetes-specific emotional distress (RR 1.10; 1.08-1.13). In a logistic regression GEE model, male sex (OR 2.06; 1.16-3.62) and greater diabetes-specific emotional distress (OR 1.03; 1.00-1.06) were associated with greater odds of neuropathy. Conclusions: Despite presence of salutary factors of diabetes management, only a small proportion of patients were resistant to all complications. Retinopathy e and not symptomatic neuropathy e was the dominant complication, indicating a need to better
543
understand mechanisms that could promote resistance to specific complications.
052-2015. CLINICAL CARE Reproducibility of In Vivo Corneal Confocal Microscopy Using an Automated Analysis Program for Detection of Diabetic Sensorimotor Polyneuropathy ILIA OSTROVSKI, LEIF ERIK LOVBLOM, MOHAMMED A. FAROOQI, DANIEL SCARR, GENEVIÈVE BOULET, PAUL HERTZ, TONG WU, ELISE M. HALPERN, MYLAN NGO, EDUARDO NG, ANDREJ ORSZAG, VERA BRILy, BRUCE A. PERKINS*y Toronto, ON Aims: In vivo Corneal Confocal Microscopy (IVCCM) is a validated, non-invasive test for diabetic sensorimotor polyneuropathy (DSP) detection, but its utility is limited by the image analysis time and expertise required. We aimed to determine the intra- and interobserver reproducibility of a novel automated analysis program compared to manual analysis. Methods: In a cross-sectional diagnostic study, 20 non-diabetes controls (mean age 41.417.3y, A1c 5.50.4%) and 26 subjects with type 1 diabetes (42.816.9y, 8.0%1.9%) underwent two separate IVCCM examinations by one observer and a third by an independent observer. Along with nerve density, branch density, and nerve tortuosity, corneal nerve fiber length (CNFL) was obtained by manual analysis (CNFLMANUAL), a protocol in which one image per eye was manually selected for automated analysis (CNFLSEMI-AUTOMATED), and one in which selection and analysis was performed electronically (CNFLFULLYAUTOMATED). Results: Mean CNFLMANUAL was 16.74.0, 13.94.2 mm/mm2 for non-diabetes controls and diabetes subjects, while CNFLSEMI-AUTOMATED was 10.23.3, 8.63.0 mm/mm2 and CNFLFULLY-AUTOMATED was 12.52.8, 10.9 2.9 mm/mm2. Inter-observer intraclass correlation coefficient (ICC) and 95% confidence intervals were 0.73[0.56, 0.84], 0.75[0.59, 0.85], and 0.78[0.63, 0.87], respectively (p¼NS for all comparisons). Intraobserver ICC were 0.72[0.55, 0.83], 0.74[0.57, 0.85], and 0.84 [0.73, 0.91], respectively (p<0.05 for CNFLFULLY-AUTOMATED compared to others). The other IVCCM parameters had lower ICCs than CNFL. Conclusions: Despite an apparent measurement (underestimation) bias in comparison to the manual strategy of image analysis, fully-automated analysis preserves the reproducibility of CNFL measurement. Future work must determine the diagnostic thresholds for fully-automated measures.
053-2015. CLINICAL CARE Agreement Between Automated and Manual Quantification of Corneal Nerve Fiber Length: Implications for Diabetic Neuropathy Research DANIEL SCARR, ILIA OSTROVSKI, LEIF ERIK LOVBLOM, TONG WU, ELISE M. HALPERN, MYLAN NGO, EDUARDO NG, ANDREJ ORSZAG, VERA BRILy, BRUCE A. PERKINS*y Toronto, ON Aim: Quantification of corneal nerve fiber length (CNFL) by in vivo corneal confocal microscopy represents a promising diabetic neuropathy biomarker, but applicability is limited by resource-intensive image analysis. We evaluated agreement between the established manual analysis protocol and a novel automated protocol. Methods: Sixty-eight controls, 139 participants with type 1 diabetes, and 249 participants with type 2 diabetes underwent CNFL measurement (N¼456). Neuropathy severity was determined by
047-2015.
Abstracts / Can J Diabetes 39 (2015) 529e547
GLUCOSE AND METABOLISM
Upper Small Intestinal Protein Sensing Improves Glucose Tolerance Through Suppression of Hepatic Glucose Production SOPHIE HAMR*, BRITTANY A. RASMUSSEN, PAIGE V. BAUER, FRANK A. DUCA, TONY K.T. LAM Toronto, ON Intestinal lipid-sensing initiates a neuronal gut-brain-liver negative feedback pathway to lower glucose production (GP) and maintain glucose homeostasis in vivo. However, whether intestinal protein-sensing regulates GP and consequently glucose tolerance via a neuronal network remains unknown, despite the ability of high-protein feeding to improve hyperglycemia and glucose tolerance in healthy and diabetic rodents and humans. Thus, we investigated the effects of upper small intestinal protein-sensing on glucose homeostasis by first infusing intraduodenal 10% w/v casein hydrolysate (CH) during an intravenous glucose tolerance test (IVGTT) in healthy rats. Intestinal infusion of CH for 50 min improved glucose tolerance as compared to saline (AUC saline: 1186.4106.8 vs. CH: 340.29128.96, n¼8, p<0.05), independent of a rise in plasma amino acids. Thus, this protein-induced glucose-lowering effect was preabsorptive. To assess whether intestinal protein-sensing improves glucose tolerance by directly lowering GP or increasing glucose uptake (GU) independent of changes in circulating glucoregulatory hormones, we infused intestinal CH during a pancreatic (basalinsulin) euglycemic clamp and evaluated changes in glucose kinetics using tracer-dilution methodology. Intestinal CH lowered GP (% suppression from basal; saline: -18.24.9% vs. CH: 463.2%, n¼8, p<0.05) compared to saline-treated rats, while GU remained unchanged (basal GU: 13.20.7 mgkg-1min-1 vs. CH GU: 12.60.7 mgkg-1min-1, n¼8, p>0.05). These findings collectively suggest that upper small intestinal protein-sensing improves glucose tolerance by lowering GP in healthy rodents. Future studies will address the underlying preabsorptive signaling mechanisms, the potential involvement of a neuronal gut-brain-liver axis, and the capacity of intestinal proteinsensing in models of obesity and diabetes.
048-2015.
CLINICAL CARE
Health Literacy Among Patients with Type 2 Diabetes Attending an Ambulatory Diabetes Services in Ireland YVONNE FINN*, SITI S. MOHD FARUDZ, MUHAMMAD A. MAD DAN Galway, Ireland Background: The institute of Medicine (2004) defines health literacy as “The degree to which individuals have the capacity to obtain, process and understand basic health information and service needed to make appropriate health decisions”. Sub-optimal health literacy has been described as a risk factor for chronic diseases such as type 2 diabetes. It is also linked with poorer glycemic control and higher complication rates. There is no data on health literacy in patients with type 2 diabetes in Ireland. This study measures health literacy in patients with type 2 diabetes attending the ambulatory diabetes services at Galway University hospital in Ireland. Methods: Patients with type 2 diabetes attending the out-patient diabetes clinic over a 4 week period in June/July 2015 were invited to participate. Functional health literacy was measured by administering the Newest Vital Sign (NVS), a validated tool to measure health literacy. Ethics approval was obtained from Galway University Hospitals’ Ethics committee. Results: Sixty seven participants, ranging from 39 to 84 years, completed the study. Twenty seven (40%) had a high likelihood of limited functional health literacy, 16 (24%) a possible likelihood of limited health literacy and 24 (36%) had adequate health literacy.
Only 2 participants (3%) perceived their level of health literacy level as low. Conclusions: Just over one third of participants in this study had adequate health literacy. Most of the participants with a high likelihood of limited health literacy did not perceive their health literacy skills as being inadequate.
049-2015. CLINICAL CARE Validation of Cooling Detection Threshold as a Marker of Sensorimotor Polyneuropathy in Type 2 Diabetes MOHAMMED A. FAROOQI, LEIF ERIK LOVBLOM, ZOE LYSY, ELISE M. HALPERN, MYLAN NGO, EDUARDO NG, ANDREJ ORSZAG, ARI BREINERy, VERA BRILy, BRUCE A. PERKINS*y Toronto, ON Aim: The measurement of cooling detection thresholds (CDT) has been established in previous cross-sectional studies as a valid test for diabetic sensorimotor polyneuropathy (DSP) in type 1 diabetes. We aimed to validate its diagnostic performance in type 2 diabetes (T2D). Methods: 220 T2D subjects from a larger cohort underwent clinical and electrophysiological examinations including 3 small-fiber function tests: CDT, heart rate variability (HRV) and laser Doppler imaging of axon-mediated neurogenic flare responses to cutaneous heating (LDIFLARE), along with the Toronto Clinical Neuropathy Score (TCNS). Clinical DSP was defined by consensus criteria whereas preclinical DSP was defined by at least one electrophysiological abnormality. Area under the curve (AUC) and optimal thresholds were determined by receiver operating characteristic (ROC) curves. Results: Subjects were aged 6311 years with mean HbA1c of 7.51.6%. The 139(63%) clinical DSP cases had mean CDT value 18.38.9 C; the 52(24%) preclinical DSP had 25.33.5 C; and the 29(13%) controls had 27.13.8 C; (p-value<0.02 for all three comparisons). For identification of clinical DSP AUCCDT was 0.79, which exceeded AUCHRV (0.60, p¼<0.0001), AUCLDI FLARE (0.69, p¼0.0003) and AUCTCNS (0.73, p¼0.03). Optimal threshold for clinical DSP identification was <22.8 C (64% sensitivity and 83% specificity). For Preclinical DSP, AUCCDT was 0.80, and the optimal threshold was 27.5 C (83% sensitivity and 72% specificity). Conclusions: Akin to studies of T1D, CDT has acceptable diagnostic performance for the identification of both clinical and preclinical neuropathy in patients with T2D. Application of CDT as a noninvasive tool for systematic screening of early neuropathy in diabetes clinics should be considered.
050-2015. LIPIDS Cardiovascular Disease Guideline Adherence and Statin Use: Results from the Canadian Study of Longevity in Diabetes JOHNNY-WEI BAI, GENEVIÈVE BOULET, ELISE M. HALPERN, LEIF ERIK LOVBLOM, ALANNA WEISMAN, DEVRIM ELDELEKLI, HILLARY KEENAN, MICHAEL BRENT, NARINDER PAUL, VERA BRILy, DAVID CHERNEYy, BRUCE A. PERKINS*y Toronto, ON; Boston, US Background: Older patients with longstanding type 1 diabetes (T1DM) have high cardiovascular disease (CVD) risk such that statin therapy is recommended independent of prior CVD events. We aimed to determine adherence to CVD prevention guidelines in patients with 50 or more years of diabetes. Research design and methods: 309 Canadians with over 50 years of T1DM completed a medical questionnaire for presence of lifestyle and CVD pharmacological interventions and they were
Abstracts / Can J Diabetes 39 (2015) 529e547
stratified into primary or secondary CVD prevention subgroups based on self-report CVD events. Associations with statin use were analyzed using multivariable logistic regression. Results: The 309 participants were 65.78.5years, with median diabetes duration of 55.0[51.0,59.0]years and A1C of 7.40 [6.80,8.00]%. 159(52.7%) participants reported adhering to diet recommendations, 296(95.8%) smoking avoidance, 217(70.5%) physical activity,, 218(71.5%) renin-angiotensin system blocker use, and 220(72.1%) statin use. Other than physical activity - reported as less common in the secondary prevention subgroup - current statin therapy was significantly lower in the primary prevention subgroup (65.5% vs. 84.8%, p¼0.0004). In multivariable logistic regression, the odds of statin use was 0.38(0.15,0.92) in members of the primary compared to the secondary prevention cohort, adjusting for age, microvascular complications, A1c, blood pressure, cholesterol, and body mass. Conclusion: Though guidelines have for years recommended statin use in longstanding type 1 diabetes regardless of CVD history, its use was substantially lower in those without compared to those with CVD history. Interventions are needed to improve adherence in the high-risk population without CVD history.
051-2015.
CLINICAL CARE
Neuropathy Prevalence Compared to other Complications: Preliminary Analysis of the Canadian Study of Longevity in Diabetes Cohort ALANNA WEISMAN, RANDY ROVINSKI, LEIF ERIK LOVBLOM, ELISE M. HALPERN, GENEVIÈVE BOULET, DEVRIM ELDELEKLI, HILLARY KEENAN, MICHAEL BRENT, NARINDER PAUL, VERA BRILy, DAVID CHERNEYy, BRUCE A. PERKINS*y Toronto, ON; Boston, US Aim: Type 1 diabetes (T1DM) subjects can survive for extreme duration without retinopathy or nephropathy whereas estimates of neuropathy are uncertain. In a cohort with 50y of T1DM, we aimed to assess the prevalence of microvascular complications and other indicator variables. Methods: 309 participants were recruited (2013-2014) across Canada by public advertisement and mailings to health professionals. Participants answered a comprehensive questionnaire and provided laboratory results. We assessed complications by eye-specialist fundus examinations, Michigan Neuropathy Screening Instrument 3/15, eGFR <60mL/min and albumin:creatinine >2 mg/mmol, and self-report of coronary artery disease (CAD), gastroparesis and peripheral vascular disease (PVD). Results: In preliminary analysis, 309 participants (55% female) were 668.5 years of age, had mean diabetes duration of 556.5y, BMI 25 (23-28.2) kg/m2, and A1c 7.51.1%. Retinopathy was most prevalent at 78% compared to neuropathy (42%), nephropathy (48%), CAD (34%), gastroparesis (15%) and PVD (10%). 37(12%) lacked evidence of any complication. In a Poisson regression generalized estimating equation (GEE) model, greater number of complications was independently associated with higher HbA1c (RR 1.05; 1.01-1.09), older age (RR 1.10; 1.061.13) and greater diabetes-specific emotional distress (RR 1.10; 1.08-1.13). In a logistic regression GEE model, male sex (OR 2.06; 1.16-3.62) and greater diabetes-specific emotional distress (OR 1.03; 1.00-1.06) were associated with greater odds of neuropathy. Conclusions: Despite presence of salutary factors of diabetes management, only a small proportion of patients were resistant to all complications. Retinopathy e and not symptomatic neuropathy e was the dominant complication, indicating a need to better
543
understand mechanisms that could promote resistance to specific complications.
052-2015. CLINICAL CARE Reproducibility of In Vivo Corneal Confocal Microscopy Using an Automated Analysis Program for Detection of Diabetic Sensorimotor Polyneuropathy ILIA OSTROVSKI, LEIF ERIK LOVBLOM, MOHAMMED A. FAROOQI, DANIEL SCARR, GENEVIÈVE BOULET, PAUL HERTZ, TONG WU, ELISE M. HALPERN, MYLAN NGO, EDUARDO NG, ANDREJ ORSZAG, VERA BRILy, BRUCE A. PERKINS*y Toronto, ON Aims: In vivo Corneal Confocal Microscopy (IVCCM) is a validated, non-invasive test for diabetic sensorimotor polyneuropathy (DSP) detection, but its utility is limited by the image analysis time and expertise required. We aimed to determine the intra- and interobserver reproducibility of a novel automated analysis program compared to manual analysis. Methods: In a cross-sectional diagnostic study, 20 non-diabetes controls (mean age 41.417.3y, A1c 5.50.4%) and 26 subjects with type 1 diabetes (42.816.9y, 8.0%1.9%) underwent two separate IVCCM examinations by one observer and a third by an independent observer. Along with nerve density, branch density, and nerve tortuosity, corneal nerve fiber length (CNFL) was obtained by manual analysis (CNFLMANUAL), a protocol in which one image per eye was manually selected for automated analysis (CNFLSEMI-AUTOMATED), and one in which selection and analysis was performed electronically (CNFLFULLYAUTOMATED). Results: Mean CNFLMANUAL was 16.74.0, 13.94.2 mm/mm2 for non-diabetes controls and diabetes subjects, while CNFLSEMI-AUTOMATED was 10.23.3, 8.63.0 mm/mm2 and CNFLFULLY-AUTOMATED was 12.52.8, 10.9 2.9 mm/mm2. Inter-observer intraclass correlation coefficient (ICC) and 95% confidence intervals were 0.73[0.56, 0.84], 0.75[0.59, 0.85], and 0.78[0.63, 0.87], respectively (p¼NS for all comparisons). Intraobserver ICC were 0.72[0.55, 0.83], 0.74[0.57, 0.85], and 0.84 [0.73, 0.91], respectively (p<0.05 for CNFLFULLY-AUTOMATED compared to others). The other IVCCM parameters had lower ICCs than CNFL. Conclusions: Despite an apparent measurement (underestimation) bias in comparison to the manual strategy of image analysis, fully-automated analysis preserves the reproducibility of CNFL measurement. Future work must determine the diagnostic thresholds for fully-automated measures.
053-2015. CLINICAL CARE Agreement Between Automated and Manual Quantification of Corneal Nerve Fiber Length: Implications for Diabetic Neuropathy Research DANIEL SCARR, ILIA OSTROVSKI, LEIF ERIK LOVBLOM, TONG WU, ELISE M. HALPERN, MYLAN NGO, EDUARDO NG, ANDREJ ORSZAG, VERA BRILy, BRUCE A. PERKINS*y Toronto, ON Aim: Quantification of corneal nerve fiber length (CNFL) by in vivo corneal confocal microscopy represents a promising diabetic neuropathy biomarker, but applicability is limited by resource-intensive image analysis. We evaluated agreement between the established manual analysis protocol and a novel automated protocol. Methods: Sixty-eight controls, 139 participants with type 1 diabetes, and 249 participants with type 2 diabetes underwent CNFL measurement (N¼456). Neuropathy severity was determined by