Associations Between Pancreatic Lipids and β-Cell Function in Black African and White European Men With Type 2 Diabetes

Abstract

Context

Intrapancreatic lipid (IPL) has been linked to β-cell dysfunction. Black populations disproportionately develop type 2 diabetes (T2D) and show distinctions in β-cell function compared with white populations.

Objective

We quantified IPL in white European (WE) and black West African (BWA) men with early T2D and investigated the relationships between IPL and β-cell insulin secretory function (ISF).

Design, Setting, and Participants

We performed a cross-sectional assessment of 18 WE and 19 BWA middle-age men with early T2D as part of the South London Diabetes and Ethnicity Phenotyping study.

Main Outcome Measures

The participants underwent Dixon MRI to determine IPL in the pancreatic head, body, and tail and subcutaneous and visceral adipose tissue volumes. Modeled first- and second-phase ISFs were comprehensively determined using C-peptide measurements during a 3-hour meal tolerance test and a 2-hour hyperglycemic clamp test.

Results

The WE men had greater mean IPL levels compared with BWA men (P = 0.029), mainly owing to greater IPL levels in the pancreatic head (P = 0.009). The mean IPL level was inversely associated with orally stimulated first-phase ISF in WE but not BWA men (WE, r = −0.554, P = 0.026; BWA, r = −0.183, P = 0.468). No association was found with orally stimulated second-phase ISF in either WE or BWA men. No associations were found between the mean IPL level and intravenously stimulated ISF.

Conclusions

The IPL levels were lower in BWA than WE men with early T2D, and the lack of inverse association with first-phase ISF in BWA men indicates that IPL might be a less important determinant of the development of T2D in BWA than in WE men.

Type 2 diabetes (T2D) is more prevalent and develops at a younger age among populations of black compared with white European (WE) ethnicity (1, 2). The pathophysiological processes of T2D are well documented and include insulin resistance, ectopic fat deposition, and pancreatic β-cell dysfunction (3). The role of pancreatic lipid accumulation in the development of T2D has been receiving increasing attention. Through the process of lipotoxicity, intrapancreatic lipid (IPL) is believed to cause β-cell damage (4, 5) through the release of lipid intermediates and free fatty acids, which interfere with cellular signaling and cause β-cell apoptosis (6). In vivo studies have shown consistently that IPL is inversely associated with insulin secretory function (ISF), specifically, the first-phase response (4, 7). Similar to visceral adipose tissue (VAT) and intrahepatic lipid (IHL), IPL has been found to be elevated in individuals with T2D (8, 9). Also, in studies investigating reversal of T2D, mobilization of IPL appears to be a key component for achieving normalization of glycemia (10).

Increasing evidence has shown distinctions in the pathophysiology of T2D in populations of African ancestry. Typically, lower levels of VAT and IHL have been reported compared with those in white populations (11). Additionally, β-cell dysfunction will be more evident. A greater insulin response to glucose stimulation has been consistently reported among healthy and prediabetic black populations compared with those of white ethnicity (12, 13). Also, in the diabetic state, black men might have lower insulin secretion compared with white men in response to both oral and intravenous glucose (14). To date, investigation of the effect of black ethnicity on IPL levels and its relationship to the metabolic abnormalities of T2D has been limited. However, given the lesser deposition of VAT and greater β-cell dysfunction typically observed in black populations, it is reasonable to hypothesize that IPL, and its role in the pathophysiology of T2D, might differ by ethnicity. A small number of studies of healthy adults and adolescents have reported lower IPL levels among black populations compared with other ethnic groups (7, 15, 16). Also, in a comparison of healthy and prediabetic adolescents of black and Hispanic ethnicity, the IPL level was found to be the strongest predictor of prediabetes in the black population but not the Hispanic population (17). Investigations of the effect of IPLs on β-cell function in black populations have shown inconsistent findings. Although no relationship was found between IPL levels and insulin secretory function in adolescents, studies of healthy adults have shown that IPL is more strongly associated with β-cell function in black populations compared with other ethnic groups (7, 15). To date, studies of IPL and β-cell function in black populations have been limited to healthy cohorts with limited development of the metabolic abnormalities of T2D. Furthermore, these studies have assessed only the first-phase insulin secretory response. Also, only indirect assessments of insulin secretion have been performed, which have limited utility in black populations because they do not account for hepatic insulin clearance, for which ethnic differences are well established (18). We recently reported deficits in second-phase insulin secretory function through comprehensive modeling of C-peptide, present in black African but not WE men with early T2D (14). The aim of the present study was to assess IPLs and investigate the relationship with first- and second-phase ISF, assessed comprehensively using C-peptide modeling, to explore the hypothesis that men of black (West) African (BWA) ethnicity with early T2D will have lower IPL levels than those of WE ethnicity.

Materials and Methods

The present investigation was conducted as a part of the South London Diabetes and Ethnicity Phenotyping study (Soul-Deep), details of which have been previously reported (19). The present study was conducted at King’s College Hospital and Guy’s Hospital, London, and approved by the London Bridge National Research Ethics Committee (approval no. 12/LO/1859). Also, all participants provided written informed consent. Recruitment and data collection were performed from April 2013 to January 2015.

Participants

Potential participants were identified through primary care practices in South London. Men of WE or BWA ethnicity were recruited. Ethnicity was self-declared and confirmed through grandparent birthplace, where the countries included were North West European and West African countries, defined by the United Nations Statistics Division. The participants also provided information on their birthplace. Eligibility was confirmed in a screening visit, and participants were considered eligible if they had met the following criteria: age, 18 to 65 years; body mass index, 20 to 40 kg/m²; a recent diagnosis of T2D (<5 years before starting the study); using only lifestyle alone or metformin to manage T2D. The exclusion criteria included the use of thiazolidinedione, insulin, oral steroids, β-blockers, or other medication that can affect the study outcome; a contraindication for MRI, such as metal implants; kidney or liver damage identified by a serum creatinine of >150 mmol/L or serum alanine transaminase level increased >2.5-fold greater than the upper limit of the reference range.

Procedures

The participants attended 3 assessment visits in random order within a maximum period of 6 months. Each participant underwent MRI for assessment of IPL, a 2-hour hyperglycemic clamp, and a mixed meal tolerance test for assessment of ISF through the measurement and mathematical modeling of C-peptide. For each assessment, the participants presented in a fasted state, having refrained from eating or drinking anything other than water from 10 pm the night before. The participants were instructed to refrain from strenuous exercise and physical activity in the 48 hours preceding the visit, refrain from consuming alcohol in the 24 hours preceding the visit, and to consume a standardized diet the day before (∼50% of calories from carbohydrates, evenly spread throughout the day, with no >30% of daily carbohydrates consumed in the evening meal). The participants taking metformin were instructed to cease taking it for 7 days before each visit.

MRI fat quantification

A Dixon-based MRI sequence was used on a 1.5 Tesla Siemens scanner to acquire images for the assessment of IPL. With the participant lying supine, images were obtained from the neck to the knee (excluding the arms), with coils placed on the abdominal region. During acquisition of the abdominal images, on instruction by the radiographer, the participants undertook three 15-second breath holds to reduce the occurrence of motion artifacts. For each participant, 320 contiguous, 3-mm slice thickness, T1-weighted transverse spin-echo images [repetition time, 6.77 ms; echo time, 4.77 ms (in-phase), 2.39 ms (out-of-phase), flip angle, 10°] were obtained.

Intrapancreatic fat was determined by analysis of MRI scans using HOROS, version 1.1.7, software (available at: www.horosproject.org; accessed December 21, 2017) by locating one or more axial images with the largest area of the head, body, and tail of the pancreas and extracting the corresponding fat and water images. On each of the fat and water images, one circular region of interest of 1 cm² was drawn on each of the head, body, and tail of the pancreas (Fig. 1). A region of 1 cm² was used as recommended by a recent review of MRI methods used to determine IPL, because it ensures the regions are within the border of the pancreas and avoids inclusion of VAT and the splenic vein (9). Using the following formula, %IPL = [F/(F+W)] × 100, where F is the pixel signal intensity of the fat image and W is the pixel signal intensity of the water image, the pancreatic fat fraction was calculated in each region, termed the IPL_HEAD, IPL_BODY, and IPL_TAIL. The IPL_MEAN was calculated as the average of the head, body and tail regions. Owing to the subjective nature of locating the regions of interest on the head, body and tail of the pancreas, IPL quantification was conducted by two independent investigators with a statistically significant correlation reported (r = 0.62; P < 0.001) and an interobserver coefficient of variation of 14%.

The total abdominal VAT mass and subcutaneous adipose tissue (SAT) mass from the neck to the knee region (excluding the arms) were quantified using an automated analysis method (Klarismo Ltd., London, United Kingdom). For VAT mass quantification, each image in the abdominal region was analyzed for the VAT area, and the area was multiplied by the slice thickness of 3 mm to determine the volume of VAT. For SAT mass quantification, all MRI scans acquired were analyzed for the SAT area, which was multiplied by the slice thickness of 3 mm to determine the volume of SAT. The volume of VAT and SAT were converted from cubic milliliters to liters and then converted to kilograms by multiplying by 0.9 kg/L (the density of fat) (20).

Insulin secretory function during hyperglycemic clamp

A 2-hour hyperglycemic clamp was conducted for assessment of ISF (21). After collection of three basal blood samples, a 20% glucose infusion was administered to achieve a hyperglycemic state of 6.9 mmol/L above basal for a period of 2 hours. The blood samples were collected at −20, −10, 0, 2, 4, 6, 8, 10, 15, 20, 30, 40, 50, 60, 75, 90, 105, and 120 minutes to measure the plasma glucose and serum insulin and C-peptide.

Insulin secretory function during a meal tolerance test

A 3-hour mixed meal tolerance test was conducted to assess ISF under physiological conditions. After an overnight fast, participants consumed a liquid milkshake (Ensure Plus; Abbott Nutrition, Berkshire, UK) providing 6 kcal/kg body weight, containing carbohydrates, protein, and fat. Blood samples were taken at −10, 0, 10, 20, 30, 40, 50, 60, 75, 90, 120, 150, and 180 minutes to measure plasma glucose and serum insulin and C-peptide.

Calculations

The incremental area under the curve (iAUC) was calculated, using the trapezoidal rule, for insulin and C-peptide responses to each challenge. To calculate an index of first- and second-phase insulin secretion in the hyperglycemic clamp, we used the iAUC for C-peptide over 0 to 10 minutes for the first phase and 10 to 120 minutes for the second phase, in analogy to DeFronzo et al. (21).

Model-based measurement of β-cell function

The glucose, insulin, and C-peptide curves during the hyperglycemic clamp and meal tolerance test were modeled using methods previously described (22–24) and SAAM-II, version 1.2, software (SAAM Institute, Seattle, WA). The main outputs of the model are glucose sensitivity of first-phase secretion (σ¹), expressed as the amount of insulin secreted in response to a rate of increase in glucose of 1 mmol/L between time 0 and 1 minute of the study: $(pmol·m−2BSA)/(mmol·l−1·min−1)$⁠; glucose sensitivity of second-phase secretion (σ²), expressed as the steady-state insulin secretion rate in response to a step increase in glucose of 1 mmol/L above baseline, in $(pmol·min−1·m−2BSA)/(mmol·l−1)$

Biochemical analyses

We measured plasma glucose using an automated glucose analyzer (2300 STAT Glucose Analyzer; Yellow Spring Instruments, Yellow Springs, OH). Serum insulin was determined by immunoassay using chemiluminescent technology (ADVIA Centaur System; Siemens Health Care, Ltd., Camberly, UK), where the interassay and intra-assay coefficients of variation ≤5.9% and 4.6%, respectively. Serum C-peptide was determined by radioimmunoassay (Millipore Ltd., Hertfordshire, UK).

Statistical analysis

The variables that were positively skewed were log-transformed to give a normal distribution. The statistical significance of differences in the variables of interest between the ethnic groups were tested using an independent samples t test for normal data or the Mann-Whitney U test for data that could not be transformed to normal. To investigate ethnic differences in IPL, we initially conducted independent samples t tests. We then used analysis of covariance, adjusting for VAT, to determine whether differences in IPL were independent of the VAT. The associations between IPL and parameters of ISF were explored using Pearson correlation. Partial correlation was used to investigate these associations with adjustment for VAT. To investigate the ethnic differences in the distribution of IPL among the head, body, and tail of the pancreas, a mixed between-within subjects ANOVA was performed. SPSS, version 24.0 (IBM Corp., Armonk, NY) was used for all statistical analyses, and P < 0.05 was considered to indicate statistical significance.

Results

Participant characteristics

We studied data from 19 BWA and 18 WE men (Fig. 2). The BWA men were first-generation West African migrants (born in Nigeria, n = 11; Ghana, n = 5; Sierra Leone, n = 2, Ivory Coast, n = 1). The characteristics of the two ethnic groups are presented in Table 1. The groups were well-matched for age, weight, and body mass index, with no substantial differences in T2D duration, fasting glucose, glycated hemoglobin, blood pressure, and liver function, as represented by alanine aminotransferase and measures of cholesterol (Table 1). Metformin use was not different between the two ethnic groups (P = 0.248), with 56% of the WE and 74% of the BWA men receiving treatment with metformin. The waist circumference was greater, although the difference was not statistically significant, in the WE men. The fasting triglyceride concentrations were significantly greater in the WE men. Analysis of the MRI data from the whole abdominal cavity showed that the weight of the abdominal VAT was significantly greater in the WE men. However, no ethnic difference were found in the SAT measured between the neck and knee (excluding the arms; Table 1).

Insulin secretory function

The C-peptide responses to the intravenous (hyperglycemic clamp) and oral (meal tolerance test) stimulations are presented in Table 2. Basal C-peptide was significantly lower in the BWA than in the WE men, although no ethnic differences were found in basal insulin (Table 2). The C-peptide iAUC during the meal test was significantly lower among the BWA men. This was also the case with the hyperglycemic clamp, specifically in the second phase (Table 2). The C-peptide data were modeled with the glucose curves to provide an estimate of the first- and second-phase glucose sensitivity of the β cells (σ¹ and σ², respectively). The modeled data showed similar findings of lower second-phase insulin secretion, although the results failed to reach statistical significance.

Pancreatic fat analysis

The mean IPL and the IPL of the head, body, and tail regions are shown in Fig. 3. The BWA men exhibited a significantly lower IPL_MEAN than the WE men (WE, 10.08% ± 2.46% vs BWA, 8.22% ± 2.51%; P = 0.029), which was driven by the significantly lower IPL_HEAD in the BWA men (WE, 9.66% ± 3.14% vs BWA, 7.03% ± 2.65%; P = 0.009). After adjustment for VAT, the IPL_MEAN showed no statistically significant ethnic differences (WE, 9.60% ± 0.65% vs BWA, 8.60% ± 0.63%; P = 0.305), and the ethnic difference in IPL_HEAD had decreased in statistical significance (WE, 9.04% ± 0.68% vs BWA, 7.18% ± 0.66%; P = 0.074). We investigated regional IPL depositions in the head, body, and tail regions within and between each ethnic group using mixed between-within subjects ANOVA. We found statistically significant differences in IPL among the pancreatic sections in the BWA men (Wilks lambda, 0.791; P = 0.019), with a substantial main effect for ethnicity (P = 0.029). No statistically significant differences were found in the distribution of IPL between the two ethnic groups (P = 0.474).

Relationships between IPL and ISF

The associations between IPL and measures of insulin secretory function are listed in Table 3. The IPL_MEAN was significantly and inversely associated with the meal test first-phase insulin secretion (σ¹) in the WE men. However, the relationship, although negative, was not statistically significant in the BWA men (WE, r = −0.55, P = 0.026; BWA, r = −0.18, P = 0.468; Fig. 4). The association was specifically with IPL_HEAD in the WE men (P = 0.023; Table 3). No evidence was found for a linear relationship between the IPL_MEAN and the meal test second-phase insulin secretion (σ²) in either ethnic group (WE, r = 0.06, P = 0.813; BWA, r = 0.05, P = 0.856). No statistically significant associations were found between IPL_MEAN and intravenously stimulated insulin secretion (σ¹ and σ²) in either ethnic group (Table 3). However, analysis of region-specific associations with intravenously stimulated insulin secretion (σ¹ and σ²; Table 3) showed inverse associations between IPL_TAIL with both σ¹ (P = 0.092) and σ² (P = 0.074), which had neared statistical significance in the WE men but not the BWA men (σ¹, P = 0.18; σ², P = 0.26). No changes were found in the statistically significance of the relationships between IPL (mean and region specific) and ISF after adjusting for VAT (data not shown).

Discussion

In our comparison of BWA and WE men with early T2D, we found ethnic differences in the deposition of pancreatic fat and its association with β-cell function. The men of BWA ethnicity exhibited lower IPL compared with the WE men, predominantly owing to the lower IPL deposition in the head of the pancreas. Furthermore, we recognized ethnic distinctions in the relationship between IPL and ISF, such that IPL is inversely associated with ISF only in WE men, leading us to speculate that IPL might be a less important determinant of the development of T2D in BWA than in WE men.

The accumulation of IPL and other depots of ectopic fat has been thought to occur owing to reduced expandability of SAT during energy surplus and prolonged release of free fatty acids from SAT, due to adipocyte insulin resistance, which subsequently deposits as ectopic fat (8, 25). The interrelated nature of ectopic fat depots has been demonstrated by Lê et al. (15), who showed a correlation among VAT, IHL, and IPL in a multiethnic cohort. Because it has been well established that VAT and IHL are lower in black populations (26, 27), we hypothesized that the IPL would also be lower. In line with previous studies, our BWA men exhibited lower VAT compared with the WE men. We found that IPL no longer differed by ethnicity after adjusting for VAT. This suggests that the lower IPL in the BWA men is driven by lower VAT, indicating a central role of VAT in determining IPL deposition. Our findings extend the results from previous T2D studies of healthy individuals, which reported lower IPL among black groups compared with white and Hispanic ethnic groups (7, 15). Furthermore, we have provided data on differences in regional deposition of IPL according to ethnicity. We found the greater IPL deposition in WE men to be specific to the head of the pancreas. This is important because the development of T2D has been shown to be specifically associated with loss of β-cell mass from the pancreatic head (28). Also, several studies have shown links between lipotoxicity and β-cell apoptosis (6, 29). During the progression from normal glucose tolerance to T2D, the loss of first-phase ISF is understood to be the most critical dysfunction of β cells (30). Additionally, the detrimental effects of IPL have consistently been shown to relate to first-phase ISF (4, 31). Our protocol enabled us to differentiate first- and second-phase insulin secretory function and to investigate, for the first time, to the best of our knowledge, ethnic differences in the relationship between these and IPL. Thus, our findings suggest that β-cell lipotoxicity might be a less important determinant of ISF in BWA than in WE men. This is consistent with findings from a recent investigation of prepubertal youth of black and white ethnicity (32) that reported greater declines in β-cell function in white youth in response to a lipid infusion, suggesting greater susceptibility of the β cells to acute lipotoxicity in white youth compared with black youth. However, it should be noted that IPL was not measured in that study. To date, studies of IPL in black populations have been limited to first-phase ISF, expressed as the “acute insulin response” measured using the intravenous glucose tolerance test (7, 15, 16). In contrast, our findings indicated the potential importance of also assessing second-phase ISF. Our results showing that BWA men have lower second-phase ISF, which we have previously explored in more detail (14), indicate that a decrease in second-phase ISF might have a more prominent etiological role in β-cell dysfunction in black populations but is not related to IPL.

In contrast to our findings, Szczepaniak et al. (7) reported that IPL was associated with the intravenous glucose tolerance test “acute insulin response” in both normal glucose-tolerant white and black ethnic groups. The acute insulin response is considered comparable to the first-phase response of the hyperglycemic clamp but in this case, only insulin was measured. In our study, we have quantified ISF through the measurement of C-peptide, which provides a more accurate estimation of β-cell function than measurement of insulin alone. This is especially important when studying ethnic comparisons of β-cell function, as it has been extensively reported that black populations exhibit different insulin responses to glucose compared with other ethnic groups and that this response results from a combination of altered insulin secretion and hepatic insulin clearance. To date, many ethnic comparison studies have been limited to the measurement of insulin.

We found ethnic differences in the regional distribution of IPL within the pancreas such that BWA men had greater IPL in the tail compared with the head of the pancreas. In contrast, in the WE men, we found no apparent regional variation in IPL deposition. However, when we studied the region-specific relationships between IPL and ISF, we found, in the WE men, an inverse relationship, which neared statistical significance, between IPL in the tail with both intravenously stimulated first-phase and second-phase ISF, which was not seen in the BWA men. Studies conducted in humans have shown regional variation in the distribution of β cells within the pancreas, concluding that the tail of the pancreas has a more than twofold greater density of β cells compared with the head and body (28). The inverse relationship we observed, albeit of borderline statistical significance, between IPL in the tail and ISF in WE men might indicate greater β-cell lipotoxicity in WE men and, in turn, reduced β-cell function in the tail of the pancreas, an association not seen in the BWA men. We propose further work should be conducted to investigate ethnic differences in the role of regional IPL on region-specific β cells within the pancreas.

To the best of our knowledge, our study design, using both the hyperglycemic clamp and mixed meal tolerance test, enabled us to compare, for the first time, in a single study, distinctions in the associations between IPL and ISF in response to orally vs intravenously stimulated glycemia. Our findings of substantial associations only between IPL and orally stimulated ISF help to explain previous contradictory results between studies that used oral vs intravenous methods (31, 33). These findings suggest an interaction among incretin hormones, IPL, and ISF. Recent studies have shown a link between β-cell lipotoxicity and a reduced incretin effect, in which increasing concentrations of free fatty acids were associated with a downregulation of the GLP-1 receptor in a mouse model (34). Our results might suggest that IPL negatively affects incretin signaling in β cells and further hinder an insulin secretory response to a meal in the WE men but not in the BWA men. This ethnic difference can be explained by differences in incretin levels, which have previously differed between black and white populations, with some studies reporting greater incretin levels in blacks (14, 35, 36) and others have reported lower (37). Further investigations are needed to understand the ethnic differences in the relationships between incretins and IPL.

Our study had several strengths, including the measurement of C-peptide for the assessment of insulin secretion and the use of two methods to comprehensively measure insulin secretory function, distinguish first- and second-phase secretion, and determine the role of incretin hormones. Another strength was our investigation of regional variation of IPL in the head, body, and tail of the pancreas and how this differs between and within each ethnic group. Our study also benefited from MRI analysis of pancreatic fat, which has been suggested to be superior to magnetic resonance spectroscopy for IPL analysis owing to the irregular size and morphology of the pancreas, especially in diabetic populations (9, 38, 39). However, our study also had limitations. Small regions of interest were used in the MRI analysis of IPL to reduce contamination with VAT, as recommended by investigators of the methods used to quantify IPL using Dixon-MRI (9, 40), although we could not guarantee that VAT contamination did not occur owing to poor participant compliance with the breath holds. Also, we could not determine whether the mean IPL represents the total IPL, because we did not measure the volume of the pancreas, which has been reported to be 33% less in individuals with early T2D compared with healthy controls (41). Our study was deliberately limited to studying men, because consistent evidence has shown sex differences in the pathophysiology of T2D in African populations (42, 43), limiting the generalizability of our findings. However, despite the greater prevalence of T2D in black women compared with black men (1), our data are valuable owing to the lack of studies of men in this field. Furthermore, our study focused on BWA ethnicity and all our participants were first-generation migrants (born in countries of West Africa). Thus, when comparing our study to previous works, differences might be present between black populations residing in the United Kingdom and those residing in the United States or other regions in terms of lifestyle behaviors, socioeconomic factors, and access to health care that could influence the development of T2D. We could not determine a causal relationship between β-cell function and IPL accumulation owing to the cross-sectional nature of the present study. Our study was also conducted on a small sample size although, despite this, the size was comparable to that in other studies of IPL and ISF. An investigation of normal glucose-tolerant and impaired glucose-tolerant groups of both WE and BWA ethnicity would enable us to gain an understanding of the effect of IPL accumulation on β-cell function during the progression of T2D.

In conclusion, our results have demonstrated ethnic differences in the deposition of pancreatic fat and its association with β-cell function. Our findings of lower IPL among BWA men with T2D suggest that the lipotoxicity in the pancreas might be less dominant in the pathogenesis of T2D in BWA than in WE men. Furthermore, ethnic distinctions in the relationship between IPL and ISF such that it relates to insulin secretory function in WE men but not in BWA men suggest that IPL might be a lesser determinant of β-cell dysfunction in BWA men with early T2D.

Abbreviations:

Abbreviations:
 
• BWA

  black West African

 
• iAUC

  incremental area under the curve

 
• IHL

  intrahepatic lipid

 
• IPL

  intrapancreatic lipid

 
• IPL_HEAD

  intrapancreatic lipid of pancreatic head

 
• IPL_BODY

  intrapancreatic lipid of pancreatic body

 
• IPL_MEAN

  mean intrapancreatic lipid

 
• IPL_TAIL

  intrapancreatic lipid of pancreatic tail

 
• ISF

  insulin secretory function

 
• SAT

  subcutaneous adipose tissue

 
• VAT

  visceral adipose tissue

 
• WE

  white European

 
• σ¹

  first-phase insulin secretory function

 
• σ²

  second-phase insulin secretory function

Acknowledgments

The authors thank Andrew Pernet, Bula Wilson and Ines De Abreu (research nurses, Diabetes Research Group, King’s College Hospital) for assisting with the metabolic assessments, Toyosi Bello (King’s College London), Anne-Catherine Perz (King’s College London), Daniel Curtis (University of Surrey), and Tracy Dew (ViaPath, London, United Kingdom) for assistance with sample processing and laboratory analysis, Elka Giemsa (CRF Manager, King’s College Hospital) for accommodating the participant visits, Maddalena Trombetta (University of Verona, Verona, Italy) for assisting with the minimal modeling analysis, and Brandon Whitcher and Haris Shuaib (Klarismo Ltd., London, United Kingdom) for conducting the automated MRI analysis. The authors also thank the staff of the Clinical Research Facility at King’s College Hospital for help in performing the studies and the study participants for their time and commitment.

Financial Support: J.L.P. is supported by the National Institute for Health Research (NIHR) Biomedical Research Centre based at Guy’s and St. Thomas’ NHS Foundation Trust and King’s College London. J.L.P. is an NIHR senior investigator. O.H. was supported by the NIHR Collaboration for Leadership in Applied Health Research and Care South London at King's College Hospital NHS Foundation Trust. The views expressed are those of the authors and not necessarily those of the National Health Service, NIHR, or Department of Health. L.B. is supported in part by funds from the Italian Ministry of Education, University and Research (grant PRIN 2015 2015373Z39_004) and University of Parma research funds (both to R.C.B.). The present study was funded by the Diabetes UK project (grant 12/0004473) and in part by funds from the Italian Ministry of Education, University and Research (grant PRIN 2015 2015373Z39_004) and University of Parma research funds (both to R.C.B.).

Author Contributions: L.M.G. formulated the research question and designed the study, supervised data collection and interpretation, and performed the minimal modeling analysis. S.A.A. formulated the research question and designed the study and supervised data collection and interpretation. J.L.P. formulated the research question, designed the study, and provided statistical advice. A.M.U. formulated the research question and designed the study. K.G.M.M.A. supervised data collection and interpretation. C.M. coordinated the study and data acquisition and performed the metabolic assessments. Z.B. and A.S. undertook data analysis. G.C.-E. coordinated MRI data acquisition. R.C.B. and L.B. performed the minimal modeling analysis. O.H. undertook the data analysis and statistical analysis and drafted the manuscript. All the authors contributed to the intellectual content and reviewed the final version of the submitted manuscript. L.M.G. is the guarantor of the present study, had full access to all the data, and takes full responsibility for the integrity of the data and the accuracy of the data analysis.

Disclosure Summary: The authors have nothing to disclose.

References