Volume 14 · Issue 07 · 2022

Diabetic foot has become one of the important causes of disability and death in diabetic patients, and it is also a common cause of hospitalization, which brings a heavy economic burden to patients, their families and society. The prevention and treatment of diabetic foot often requires multidisciplinary cooperation. It is necessary to not only control various cardiovascular risk factors such as blood sugar, blood pressure, blood lipids and smoking, but also give systematic anti-infective treatment if necessary. It is also necessary to strengthen the local treatment of diabetic foot, such as debridement of foot ulcer, dressing change, improvement of blood supply to lower limbs and foot decompression and braking. This article emphasizes the recent progress of revascularization, wound repair, stem cell transplantation and anti-infection therapy related to the clinical practice of diabetic foot, and emphasizes that the healing of diabetic foot ulcer can be ensured only by comprehensively controlling the systemic and local factors affecting foot ulcer. The author also briefly comments on some new technologies, new methods and related papers in the treatment of diabetic foot, and points out the gap between the clinical prevention and treatment of diabetic foot in China and the advanced level of developed countries.

To investigate the relationship between time in range (TIR) of glucose and diabetic foot.

Diabetic patients hospitalized in the Department of Endocrinology, the First Affiliated Hospital of Xi ′an Medical University from June 2017 to October 2021 were enrolled and divided into non-diabetic foot group (NDF group) and diabetic foot group (DF group) according to whether they had diabetic foot. General and biochemical data were collected before admission. All patients received continuous glucose monitoring for 3 days at admission. TIR, time above range (TAR), time below range (TBR), glucose variability coefficient (CV), standard deviation (SD), daytime blood sugar average absolute deviation (MODD) and so on were calculated. Chi-square test, U test and t test were used to compare the two samples.

A total of 450 diabetic patients were included, of whom 225 were in NDF group and 225 in DF group. Compared with NDF group, patients in DF group had more males, older age, longer course of disease, higher systolic blood pressure, higher white blood cells (WBC), platelets, low density lipoprotein, urea nitrogen, creatinine, prevalence of diabetic nephropathy and coronary heart disease, glycated hemoglobin A_1c (HbA_1c), TAR (>10 mmol/L), CV, SD and MODD (P<0.05), while body mass index (BMI), hemoglobin, indirect bilirubin, albumin, cholesterol, triglyceride, high density lipoprotein, estimated glomerular filtration rate (eGFR) and TIR were lower (P<0.05). In the corrected model, DF risk was reduced by 13.1% for every 10% increase in TIR. After adjusting for confounding factors such as age, course of disease and HbA_1c, TIR was significantly associated with lower grade diabetic foot group [OR (95%CI): 0.985 (0.972-0.999), P=0.035].With the increase of TIR quartile, the prevalence of diabetic foot decreased gradually. Spearman correlation analysis showed that TIR quartile was negatively correlated with the risk of diabetic foot (r=-0.290, P<0.01), TIR level of diabetic foot patients was positively correlated with systolic blood pressure, and negatively correlated with WBC and HbA_1c.

Among diabetic patients, glycemic variability was higher in diabetic foot patients, TIR was significantly associated with diabetic foot, and TIR quartiles were negatively associated with the risk of diabetic foot disease.

To describe the microbiological profile of diabetes foot infection (DFI) and to analyze the correlation of Texas wound classification system with the characteristics of pathogenic bacteria.

Patients hospitalized in the Department of Endocrinology of PLA Strategic Support Force Characteristic Medical Center from January 2014 to May 2019 for DFI with positive bacterial culture results were recruited retrospectively in this study. The patients were assigned into different groups according to Texas grades and Texas stages. General data and bacterial culture results of the study population were collected. Positive rate was used to describe the microbiological profile. The correlations of different Texas grades and stages with pathogen characteristics were analyzed by Chi-square test. The potential confounding factors were corrected by logistic regression.

Two hundred and thirty-nine patients were included in the study. According to the depth of the wounds, 52 cases were in Texas grade 2 group (limited to skin and subcutaneous tissues), 38 cases in Texas grade 3 group (deep into muscles) and 149 cases in Texas grade 4 group (involving bone and joint). According to the existence of ischemia (as 5 cases missed due to the lack of ischemia-related data, a total of 234 patients were divided into groups by Texas stages), 81 cases were in Texas stage B group (without ischemia) and 153 cases in Texas stage D group (with ischemia). Among 239 cases, the majority of DFI were monomicrobial infections (74.9%) and Gram-positive bacteria (G+, 182 cases, 76.2%) were the major pathogens. Among G+cases, the most common three bacteria were Staphylococcus aureus (71 cases, 39.0%), Staphylococcus epidermidis (35 cases, 19.2%) and Streptococcus agalactiae (19 cases, 10.4%). G+infections kept the major infections of patients in different Texas grade groups while G- infections were not rare. There was no significant difference in the positive rate of G+ infections or G- infections among the three groups (P>0.05). According to Texas stages, G+ infection took the main part in both Texas B group and Texas D group while the infection rate of G+ in Texas B group (85.2%) was statistically higher than that in Texas D group (71.5%, P=0.019). On the other hand, The G- infection rate in Texas D group (43.7%) was higher than that in Texas B group (25.9%). The difference was statistically significant (P=0.007). After adjusting for age, course of diabetes foot, glycosylated hemoglobin level and history of antibiotic use, the infection rate of G- in Texas B group was still significantly lower than that in Texas D group (OR=0.52, 95%CI 0.28-0.97, P=0.041).

G+ was the main pathogen of DFI. Monomicrobial infections took the major part in DFI, and the risk of G- infection was increased significantly in Texas stage D.

To assess the association between social network characteristics and diabetic foot of type 2 diabetes mellitus (T2DM) patients.

This was a cross-sectional study. Patients with T2DM hospitalized in the Department of Endocrinology and Metabolism ofNanfang Hospital from September 2020 to April 2021 were collected and divided into control group and diabetic foot group according to whether they were complicated with diabetic foot. The characteristics of social network were collected in the form of face-to-face interview, including the general information of patients, the structural characteristics of social network and the functional characteristics of social network. At the same time, the characteristics of age, occupation, education level and social isolation of social network members were recorded. A gender-stratified analysis was performed. Binary logistic regression analysis was used to evaluate the association between social network characteristics and diabetic foot.

This study included 57 patients in the diabetic foot group and 62 patients in the control group. Compared with the control group, the score of Lubben social network scale was lower (OR=0.960), the average age of network members was lower (OR=0.918), and the risk of diabetic foot was higher (all P<0.05). In the male patients, there were lower score of Lubben Social Network Scale (OR=0.949), higher proportion of family members (OR=1.024), higher emotional support related to discomfort (OR=1.739), and the higher risk of diabetic foot (all P<0.05). In female patients, there were higher proportion of family members (OR=1.045), higher risk of diabetic foot (P<0.05).

Social isolation, average age of social network members and emotional support related to discomfort were closely related to diabetic foot. The proportion of non-family members in social network members might play a positive role in the prevention and treatment of diabetic foot.

To evaluate the safety and efficacy of autologous whole blood external application in treating diabetes foot chronic wounds, and to explore the mechanism of the therapy.

The patients with chronic wounds of diabetes foot admitted to Hunan People′s Hospital from January to December in 2020 were enrolled. According to the random number table method, the patients were divided into control group and experimental group (autologous whole blood group). In the case of active control of underlying diseases, two groups of patients were advanced to nibble debridement, and the control group was treated with active oxygen and hypochlorite disinfectant (DAVIC) combined with foam dressing. The experimental group was treated with DAVIC, autologous whole blood external application combined with foam dressing. On the day of treatment and the 3rd, 6th, 9th and 15th days of treatment, the local wound conditions were observed, adverse reactions, wound healing rate, capillary count and growth factor expression were recorded. The growth factors included platelet derived growth factor (PDGF), vascular endothelial growth factor (VEGF), epidermal growth factor (EGF) and transforming growth factor-β1 (TGF-β1) and insulin-like growth factor-1 (IGF-1). The general condition, capillary count, growth factor expression and wound healing rate of the two groups were compared by independent sample t-test and χ² test.

Forty patients with chronic wounds of diabetes foot were included in the study.There were 20 cases in the experimental group and 20 cases in the control group. There was no significant difference in gender distribution, age, wound area, Wagner grade, glycosylated hemoglobin, leukocyte count, hypertension and hyperlipidemia, ankle brachial index between the two groups (P>0.05). There were no adverse reactions in both groups after treatment. There was no significant difference in wound healing rate and capillary count between the two groups on the day and the third day of treatment; on the 6th, 9th and 15th day, there was significant difference between the two groups (P<0.05), and the experimental group was higher than the control group. The growth factors (PDGF, VEGF, EGF, TGF-β1, IGF-1) in the granulation tissue of the experimental group and the control group were expressed in a small amount on the 3rd day, and the expression increased gradually on the 9th and 15th day. There was no significant difference in the expression between the two groups on the 3rd day of treatment, but there was significant difference on the 9th and 15th day of treatment (P<0.05). The expression in the experimental group was higher than that in the control group.

Autologous whole blood can increase the healing rate of diabetes foot chronic wounds. Autologous whole blood may promote wound healing by increasing the number of capillary and growth factor in chronic wound of diabetes foot.

To compare swab sampling method with tissue sampling method to determine whether diabetic foot (DF) were infected pathogenic bacteria and to determine the consistency of pathogenic microorganism by using Meta-analysis method.

Relevant literatures were retrieved from PubMed, Embase, The Cochrane Library, CBM, CNKI, VIP and Wanfang Data until March 25, 2021. Two independent researchers evaluated the included literatures and extracted the data to analysis, RevMan5.3 and STATA13 software were used to calculate the sensitivity (Sen) and specificity (Spe) of the swab sampling methods to determine whether DF were winfected pathogenic bacteria and the consistence of the pathogenic microorganisms, and subgroup analysis was conducted according to the depth of infection. To compare the difference between the two methods, t test was used to analyze the number of pathogens and Chi-square test was used to analyze the classification of Gram stain typing.

A total of 725 samples from 6 studies were included. In the evaluation of swab sampling method to determine whether DF infected pathogenic bacteria, the combined SEN=0.94 [95%CI (0.82-0.98)], the combined SPE=0.59 [95%CI(0.33-0.81)]. While the combined SEN=0.71 [95%CI (0.46-0.87)], the combined SPE=0.28 [95%CI (0.14-0.49)], AUC=0.45 [95%CI (0.41-0.60)], when swab sampling method was used to determine pathogenic microorganisms. In the subgroup analysis, the sensitivity of the shallow infection group was higher than that of the deep infection group. There was no significant difference between the two sampling methods in the number of pathogens obtained by Gram stain typing.

Swab sampling is of high diagnostic value in determining whether pathogenic bacteria are infected, but its accuracy in identifying DF infection of pathogenic microorganisms is poor. Swab sampling method cannot completely replace tissue sampling method.

To investigate the effects of asymptomatic hypoglycemia and blood glucose fluctuation on diabetic cardiovascular autonomic neuropathy (DCAN).

This was a retrospective study. Patients with type 2 diabetes mellitus (T2DM) who were hospitalized in the Department of Endocrinology of Hefei Hospital Affiliated to Anhui Medical University from September 2018 to July 2021 were enrolled in this study. The mean amplitude of glycemic excursions (MAGE), 24-hour mean blood glucose level (24hMBG), time in range (blood glucose 3.9-10.0 mmol/L, TIR), time below range (blood glucose<3.9 mmol/L, TBR) and time above range (blood glucose>10.0 mmol/L, TAR) were collected through the continuous glucose monitoring system (CGMS). The ratio of total hypoglycemia (blood glucose<3.9 mmol/L) and asymptomatic hypoglycemia (the onset of hypoglycemia is not accompanied by palpitation, dizziness, hunger, sweating, hand shaking and other sympathetic nerve excitations) were calculated. According to the results of CGMS and whether hypoglycemia symptoms occurred, the patients were divided into non hypoglycemia group, symptomatic hypoglycemia group and asymptomatic hypoglycemia group. Patients were divided into DCAN and no-diabetic cardiac autonomic neuropathy (N-DCAN) groups based on the results of the Ewing trial. The t test, One way analysis of variance (ANOVA), Mann‐Whitney U test, Kruskal-Wallis H test and Chi-square test were used to compare the risk factors of DCAN. Multivariate logistic regression analysis was used to evaluate the risk factors of DCAN in T2DM patients.

A total of 342 T2DM patients were included. Two hundred and forty-six cases were in the non hypoglycemia group, 44 cases in the symptomatic hypoglycemia group and 52 cases in the asymptomatic hypoglycemia group. There were 214 patients in the DCAN group and 128 in the N-DCAN group, and the incidence of DCAN was 62.57% (214/342). Compared with N-DCAN group, the proportion of asymptomatic hypoglycemia in N-DCAN group was higher (χ²=10.60, P=0.001). Compared with non hypoglycemia group (χ²=10.716, P=0.001) and symptomatic hypoglycemia group (χ²=5.490, P=0.019), the incidence of DCAN was higher in asymptomatic hypoglycemia group. Patients in the DCAN group had higher MAGE, 24hMBG, TAR and lower TIR when compared to the N-DCAN group (all P<0.05). Logistic regression analysis showed that asymptomatic hypoglycemia and MAGE were the risk factors of DCAN in T2DM patients, the odd values (95%CI) were 2.324 (1.093, 4.941) and 1.456 (1.245, 1.723) respectively.

Asymptomatic hypoglycemia and MAGE were the risk factors for DCAN.

To evaluate the renal efficacy of exenatide in diabetic kidney disease (DKD) patients with different baseline renal function.

Between March 2016 and April 2019, a randomized, parallel study conducted in 4 general hospitals was performed. Type 2 diabetes mellitus (T2DM) patients with an estimated glomerular filtration rate (eGFR) ≥30 ml·min^-1·(1.73 m²)^-1 and macroalbuminuria, defined as urinary albumin-creatinine ratio (UAER)>0.3 g/24 h. In this study, a computer-generated random number table was used for block randomization. The selected patients were randomized 1∶1 to receive exenatide plus glargine or lispro insulin plus glargine for 24 weeks. We investigated percentage change in UAER after 24 weeks of intervention comparing to baseline measurement according to eGFR stage at baseline [G1 stage [eGFR≥90 ml·min^-1·(1.73 m²)^-1]; G2 stage [eGFR 60-89 ml·min^-1·(1.73 m²)^-1]; G3a stage [eGFR 45-59 ml·min^-1·(1.73 m²)^-1] and G3b stage [eGFR 30-44 ml·min^-1·(1.73 m²)^-1]. The mixed effect model of repeated measurement was used to analyze the change rate of UAER in the whole analysis set according to the above groups.

Ninety-two patients were randomly assigned to the intervention group (n=46) and the control group (n=46), and the patients had used at least 1 study drug. Eighty-one (88.0%) patients (43 in intervention group, 38 in control group) were included for the full set analysis. Seventeen patients were in G1 stage, 34 in G2 stage, 15 in G3a stage and 15 in G3b stage. After 24 weeks, the mean difference of the percentage change in UAER from baseline between exenatide plus insulin glargine and insulin glargine plus lispro was -55.9% (P=0.015) in patients in G2 stage at baseline. While no difference showed in the other subgroups (P>0.05).

For the DKD patients in G2 stage at baseline, the renal benefit was greater after the administration of exenatide.

To investigate the relationship between lens fluorescence ratio (LFR) and diabetic retinopathy (DR), and to explore its role in the screening of DR in Chinese population.

A cross-sectional study design was used. A total of 590 diabetic patients aged 20 to 70 years in eight provinces of China between May 2020 and January 2021 were enrolled. Questionnaire survey, smoking and past medical history were collected. Systolic blood pressure, diastolic blood pressure, height, and weight were measured, and fasting blood glucose (FPG), 2-hour postprandial blood glucose, glycated hemoglobin A_1c (HbA_1c), blood lipids and other indicators were detected. Estimated glomerular filtration rate (eGFR) and body mass index were calculated. Grading of fundus images was based on the Early Treatment Diabetic Retinopathy Study (ETDRS) score. LFR was measured by the biomicroscopy. All subjects were grouped according to the tertile of LFR levels: the lowest LFR group (group Q₁, LFR<20.20%, 196 cases), the intermediate LFR group (group Q₂, LFR 20.20%-25.88%, 198 cases), the highest LFR group (group Q₃, LFR≥25.89%, 196 cases). The logistic regression analyses were used to examine the association of LFR and DR. Receiver operating characteristic curve (ROC) was used to evaluate the screening efficacy of LFR alone and combined HbA_1c in screening DR.

The prevalence of DR [5.10% (10/196) in group Q₁, 11.62% (23/198) in group Q₂, and 30.10% (59/196) in Q₃, respectively; P value for trend<0.001] increased with increasing LFR levels. Based on ETDRS score ≥31, the odds of DR was 2.34 for one SD increase (95%CI 1.68-3.26) after adjustment for confounders including age, gender, body mass index, systolic blood pressure, diastolic blood pressure, HbA_1c, FPG, 2-hour postprandial blood glucose, high-density lipoprotein cholesterol, triglycerides, eGFR, smoking history. Based on ETDRS score≥31, the area under the curve (AUC) of LFR was 0.783, the sensitivity and specificity were 64.15% and 84.36%, respectively. The AUC increased to 0.865 and the sensitivity increased to 94.34% after combining LFR with HbA_1c.

LFR was positively associated with the risk of DR. It also showed that high AUC and specificity for DR screening and the screening efficiency were improved after combining with HbA_1c.

To investigate of the mechanism of liraglutide regulation of high-fat-induced hepatic lipid metabolism in mice.

Eighteen healthy male C57BL/6 mice were randomly divided into three groups: normal control group (NC group), obesity control group (OC group) and liraglutide group (six mice in each group). Mice in NC group were fed with low-fat diet, and mice in OC group and liraglutide group were fed with high-fat diet for 12 weeks to establish the high-fat induced obesity mouse model. Then, liraglutide group was intraperitoneally injected with 400 μg·kg^-1·d^-1 for 7 days, equal volumes of saline were injected intraperitoneally into the NC and OC groups. Adipose tissue and liver weight were measured. Body weight, fasting blood glucose, glucose tolerance and insulin resistance, serum and liver triglyceride (TG) and total cholesterol (TC) levels were measured. Serum insulin, interleukin 6 (IL-6) and tumor necrosis factor α (TNF-α) levels were measured by enzyme-linked immunosorbent assay. The mRNA levels of silent information regulator 1 (SIRT-1), peroxisome proliferator-activated receptor γ coactivator 1α (PGC-1α) and phosphoenolpyruvate carboxykinase (PEPCK) in liver were detected by polymerase chain reaction. The protein expression levels of mouse liver SIRT-1, PGC-1α and PEPCK, were detected by Western blotting. One-way analysis of variance (ANOVA) was used for comparison between groups.

Compared with NC group, in OC group, body weight, fat weight, fasting blood glucose and fasting insulin levels were increased (P<0.05), glucose and insulin tolerance levels were decreased (P<0.05), serum TG, TC, IL-6 and TNF-α levels and liver TG, liver TC and liver weight were increased (P<0.05), however, liver SIRT-1, PGC-1α, PEPCK mRNA and protein levels were decreased (P<0.05). Compared with the OC group, in the liraglutide group, body weight, fat weight, fasting blood glucose and fasting insulin levels were decreased (P<0.05), glucose and insulin tolerance levels were increased (P<0.05), serum insulin, IL-6, TNF-α and triglycerides were significantly decreased (P<0.05), liver lipids were reduced (P<0.05), however, liver SIRT-1, PGC-1α, PEPCK mRNA and protein levels were significantly increased (P<0.05).

Liraglutide improved glucolipid metabolism, increased hepatic fatty acid oxidation and reduced hepatic fat accumulation in high-fat diet-induced obese mice, and the mechanism might be related to activation of hepatic SIRT-1/PGC-1α/PEPCK pathway.

Diabetic foot is a serious complication of diabetes. Patients are often accompanied by lower limb ischemia, neuropathy, infection, and various basic diseases. Especially for elderly diabetic foot patients with poor general condition, many complications and poor economic conditions, the available treatment measures are limited. Two elderly diabetic foot patients with coronary heart disease and severe arterial occlusive disease of lower extremities were reported. When conventional treatment was not available, antibiotic bone cement was used to treat them with good results. Antibiotic bone cement provides a new way of thinking and treatment for diabetic foot, especially for elderly patients with complicated disease.

This paper reports a 50-year-old female with iron overload syndrome, which mainly showed diabetes, amenorrhea, skin pigmentation, osteoporosis and other endocrine organ damage. Iron deposition in liver, pancreas, pituitary gland, ventricle and other tissues, liver CT and MRI showed "white liver" and "black liver" respectively. Cranial MR magnetosensitivity imaging revealed diffuse paramagnetic material deposition in the intraventricular choroid plexus region. Patients with diabetes complicated with pigmentation should be alert to iron overload syndrome and improve iron metabolism indexes to avoid missed diagnosis.

Insulin resistance syndrome type A (TAIRS) is a syndrome characterized by hyperinsulinemia and hyperandrogenemia, which can be accompanied by polycystic ovary syndrome (PCOS) in women. Its onset is hidden, and there are no clear criteria for diagnosis and treatment. This article reports three patients in a TAIRS family: the proband and his sister and mother. The proband and his sister presented with acanthosis nigricans, hirsutism, polycystic ovaries and hyperinsulinemia, but the proband's mother had no obvious skin symptoms and no hyperinsulinemia. Genetic testing showed that all three patients had a c.3769C>T (p.Gln1257Ter; Het) mutation, which was located in exon 21 of the insulin receptor gene. After improving insulin sensitivity and anti-hyperandrogen therapy, the symptoms of the proband and his sister were relieved. Epigenetic changes may be the reason why they have the same genetic changes but different clinical symptoms.

Diabetic foot infection (DFI) is one of the important causes of worsening, amputation, and death in diabetic patients, as well as a common cause of increased hospitalization and medical costs. In terms of treatment, systemic application of antibiotics, thorough debridement, regular dressing change, local use of special dressings, negative pressure closed drainage, decompression therapy and other methods are mainly used. Those with large wounds need timely skin grafting or flap transplantation. Antibiotic-loaded bone cement is widely used in the field of orthopedics. It uses induction membrane technology to repair large segments of long bone defects, exerts a lasting local anti-infection effect, and forms bio-induction membrane around bone cement. In recent years, it has been gradually applied to the treatment of DFI wounds, and has achieved satisfactory results. This article reviews the clinical application progress of antibiotic-loaded bone cement in the treatment of DFI wounds, aiming to improve the understanding of its product development, mechanism of action and clinical efficacy, and provide reference for clinical treatment choice.

Diabetes mellitus is a chronic metabolic disease whose incidence is on the rise worldwide. Diabetic foot ulcer (DFU) is a common serious complication of diabetes. About 25% of diabetic patients will develop foot ulcer, and more than 70% of them need lower limb amputation, which has a serious impact on the quality of life of patients. With the continuous development and application of 3D printing technology, the role of 3D printed biological scaffolds in DFU treatment has become more and more obvious. This paper summarizes the causes and pathogenesis of DFU, and introduces the role of 3D printed scaffold in wound healing of DFU in detail, and the future development direction, which provides reference for the treatment of DFU and the development of 3D printed scaffold.

Type 1 diabetes mellitus (T1DM) is an autoimmune disease characterized by absolute insulin deficiency due to islet beta cell failure. On the basis of genetic susceptibility, environmental factors such as diet, viruses and bacteria play an important role in the initiation of the body's autoimmune response. In recent years, it has been found that gluten can participate in the occurrence and development of T1DM by inducing immune response, mediating the changes of intestinal flora, and changing the intestinal barrier function. This paper reviewed the basic and clinical research progress of gluten in T1DM, and prospected the application of gluten in T1DM intervention.

The intestinal flora is closely related to diabetes, and oral hypoglycemic drugs can act through the intestinal flora. As a new generation of oral hypoglycemic drugs, sodium-glucose cotransporter 2 inhibitor (SGLT2i) can inhibit the reabsorption of glucose by the proximal tubules of the kidney, promote urinary glucose excretion and achieve the goal of hypoglycemic. Studies have shown that SGLT2i may change the intestinal flora. This article reviews the research progress of the influence of SGLT2i on the intestinal flora of diabetic patients.

Diabetic gastrointestinal dysfunction (DGD) is the immune, metabolic and nervous system damage of the gastrointestinal tract caused by long-term glucose metabolism imbalance, which seriously affects the whole gastrointestinal tract function. Enteric nervous system (ENS) is the main component of the digestive tract, which contains intestinal neurons, intestinal glial cells and intestinal interstitial cells, and is involved in regulating the local function of gastrointestinal tract. The changes in ENS function are closely related to DGD. Targeting ENS can reduce gastrointestinal lesions and slow down the occurrence and development of DGD. In this paper, the role and regulatory mechanism of ENS involved in DGD are reviewed.