United States

BackgroundCystic fibrosis-related diabetes (CFRD) is a unique type of diabetes that is associated with significantly increased morbidity and mortality in both children and adults with cystic fibrosis (CF). The prevalence of CFRD progressively increases with age such that more than half of adults with CF develop CFRD. Early diagnosis and treatment are associated with improvements in body weight and pulmonary function, reduction in the frequency of pulmonary exacerbations, and improved overall survival. Since patients with CFRD may present with no symptoms, screening is recommended starting from the age of 10 years with an annual oral glucose tolerance test (OGTT). Especially during the COVID-19 pandemic, requiring an in-person clinic visit can be challenging, which may lead to a delayed diagnosis of CFRD.ObjectivesThe purpose of this project was to develop a state-of-the-art technique to detect changes in glucose levels of patients with CF by developing a deep learning-based audio classification tool. Preliminary work by our group suggested that voice characteristics could distinguish between patients with CFRD patients and patients with CF but without CFRD. We hypothesize that high blood glucose levels may cause laryngeal soft tissue swelling leading to changes in voice characteristics.MethodsWe performed a prospective cross-sectional study in adult patients with CF recruited from Emory CF Clinic from March to December 2021. We recorded 5-second voice samples of a sustained /a/ vowel via a portable digital microphone. The spectrogram was extracted via the Mel frequency cepstral coefficient. The training to test the dataset ratio was 80: 20. 20% of the training dataset were randomly selected to serve as a validation dataset. We designed a convolutional neural network (CNN) architecture for CFRD patients&#39; voice classification.ResultsThere were a total of 100 subjects consisting of 43 patients with CFRD and 57 patients with CF without diabetes. The male to female ratio was approximately 60: 40 in both groups. Patients with CFRD had similar mean age and mean BMI to patients without CFRD. There was a significantly higher point of care glucose level in CFRD patients. The mean duration of a CFRD diagnosis was 9 years and the mean HbA1c level was 7.26 in the CFRD group. The performance of the VGG model CNN classifier achieved 98.7% and 94.92% accuracy on training and validation datasets, respectively. On the test dataset, the model achieved 73.53% sensitivity 69.77% specificity and 71.43% accuracy.ConclusionsWe found a deep learning-based audio screening tool for CFRD could be potentially used as an alternative tool for screening in the CF community. A convolutional neural network algorithm demonstrated high sensitivity and specificity to adequately differentiate between patients with and without CFRD. Larger prospective studies are required to test this technology in patients with every form of diabetes.

ENDO 2022

Deep learning-based voice screening technique for cystic fibrosis related diabetes

**Program**
ENDO’s comprehensive program provides attendees the opportunity to learn about the latest developments in hormone science and medicine from renowned investigators, expert clinicians, and educators from all over the world. 

This year’s education program features:
more than 100 in-person and virtual sessions with another 70+ in-person only sessions;
six plenary sessions including plenaries on Envisioning Future Therapies for Type 1 Diabetes and Addressing Challenges in Fertility and Reproduction;
over 20 Meet the Professor sessions presented both in-person and live-streamed virtually and another 20+ presented live in Atlanta;
four Basic Science Pathways, including Diabetes and Metabolism; Neuroendocrinology; Nuclear Receptors and Signaling; and Reproductive Endocrinology; and
more than 30 oral sessions and a robust virtual poster hall for accepted scientific abstracts.

**What Programming Is Included in My Virtual Meeting Registration?**
As you build your schedule during the conference, you can select the sessions that are tagged with 'Livestreamed', so you can view virtually. These sessions are being livestreamed, so they are available in real time to our virtual audience. In addition, our in-person and livestreamed sessions will be recorded, so you can access them on demand after 72 hours. We will have over 200 hours available on demand for Standard and Premium pass registrants.

Product theaters and the Communications and Career Center will also include hybrid content delivery.

**Plenaries**
Spanning the spectrum of science, clinical care, and social implications, an outstanding slate of world-renowned speakers will showcase the most cutting-edge advances on research and medicine.

**Symposia**
Our ENDO 2022 lineup of sessions offers a unique opportunity to dive into topics from nearly every major therapeutic area of endocrinology and hormone science. Listen, learn, and be inspired by world-class faculty who are at the forefront of scientific discovery and high-quality health care.

**Basic Science Pathways**

These special “meeting-within-a-meeting” tracks integrate focused discussions, poster sessions, and networking opportunities to provide you with the latest research in:
Diabetes and Metabolism
Neuroendocrinology
Nuclear Receptors and Signaling
Reproductive Endocrinology

**Meet the Professor & Meet the Scientist**
Discover insights from experts on treating a variety of endocrine conditions and emerging science in the field of endocrinology, and learn the latest in patient care and scientific research.

ENDO’s comprehensive program provides attendees the opportunity to learn about the latest developments in hormone science and medicine from renowned investigators, expert clinicians, and educators from all over the world. 

BackgroundCystic fibrosis-related diabetes (CFRD) is a unique type of diabetes that is associated with significantly increased morbidity and mortality in both children and adults with cystic fibrosis (CF). The prevalence of CFRD progressively increases with age such that more than half of adults with CF develop CFRD. Early diagnosis and treatment are associated with improvements in body weight and pulmonary function, reduction in the frequency of pulmonary exacerbations, and improved overall survival. Since patients with CFRD may present with no symptoms, screening is recommended starting from the age of 10 years with an annual oral glucose tolerance test (OGTT). Especially during the COVID-19 pandemic, requiring an in-person clinic visit can be challenging, which may lead to a delayed diagnosis of CFRD.ObjectivesThe purpose of this project was to develop a state-of-the-art technique to detect changes in glucose levels of patients with CF by developing a deep learning-based audio classification tool. Preliminary work by our group suggested that voice characteristics could distinguish between patients with CFRD patients and patients with CF but without CFRD. We hypothesize that high blood glucose levels may cause laryngeal soft tissue swelling leading to changes in voice characteristics.MethodsWe performed a prospective cross-sectional study in adult patients with CF recruited from Emory CF Clinic from March to December 2021. We recorded 5-second voice samples of a sustained /a/ vowel via a portable digital microphone. The spectrogram was extracted via the Mel frequency cepstral coefficient. The training to test the dataset ratio was 80: 20. 20% of the training dataset were randomly selected to serve as a validation dataset. We designed a convolutional neural network (CNN) architecture for CFRD patients' voice classification.ResultsThere were a total of 100 subjects consisting of 43 patients with CFRD and 57 patients with CF without diabetes. The male to female ratio was approximately 60: 40 in both groups. Patients with CFRD had similar mean age and mean BMI to patients without CFRD. There was a significantly higher point of care glucose level in CFRD patients. The mean duration of a CFRD diagnosis was 9 years and the mean HbA1c level was 7.26 in the CFRD group. The performance of the VGG model CNN classifier achieved 98.7% and 94.92% accuracy on training and validation datasets, respectively. On the test dataset, the model achieved 73.53% sensitivity 69.77% specificity and 71.43% accuracy.ConclusionsWe found a deep learning-based audio screening tool for CFRD could be potentially used as an alternative tool for screening in the CF community. A convolutional neural network algorithm demonstrated high sensitivity and specificity to adequately differentiate between patients with and without CFRD. Larger prospective studies are required to test this technology in patients with every form of diabetes.

technical paper

Background: The link between birth-size, early childhood growth and later life cardiovascular (CV) risk is well recognised. The contribution to CV risk of growth and cell differentiation during pre-implantation embryo development is unknown.
Aims: Develop a transcriptomic model of developmental trajectory in the pre-implantation embryo and link this model to birthweight, child growth and CV risk in later life.
Methods: Single cell RNA velocity¹ and pseudotime analyses were used to infer developmental trajectory in the preimplantation human embryo². Time series clustering was used to identify genes with similar activity patterns. These data were compared with independent genomic data from four further data sets. I) Birthweight genome wide association study (GWAS)³. II) Genes correlated with gestation⁴. III)  A previously identified set of 55 genes that classify pre-hypertensive young adults using their blood methylome at age 0, 7 and 17 years⁵. IV) A set of genes with differential methylation associated with linear childhood growth⁶. Data sets III) and IV) were derived from the Avon Longitudinal Study of Parents and Children (n=6487). Enrichment was assessed using the hypergeometric test and confirmed via random permutation (1x10⁵-fold) and z-score analysis. Random Forest was used to determine the predictive value presented as the area under the curve (AUC) of the receiver operating characteristic.
Results: 8256 genes contributed to the velocity of mRNA change in the preimplantation embryo and a temporal wave pattern was revealed with peaks of velocity at 3, 4, and 5-days post-fertilisation. Genes associated with the velocity of the developmental trajectory formed 6 clusters, in which metabolic pathways were over-represented (adj. p<0.02).
We found that 16/60 birthweight GWAS (I) and 3096/10521 gestation correlated genes (II) were associated with embryonic developmental trajectory (both 2-fold enrichment, p<5x10-3). Furthermore, 11/55 pre-hypertensive associated genes (III) were linked with embryonic developmental trajectory (>10-fold enrichment, p<1x10-8), these 11 genes predicted adult pre-hypertension in small for gestational age catch-up growth (cuSGA) with an AUC=0.98. Of the 8256 developmental trajectory genes, 307 were differentially methylated in linear childhood growth (IV) (2.3-fold over-enrichment, p<1x10-8).
Conclusion: Waves of gene expression velocity occur in preimplantation human embryos. These waves mark major developmental transitions and form a transcriptomic developmental trajectory which is associated with both birthweight and gestation. The transcriptomic developmental trajectory genes are over-enriched for genes associated with the linear phase of childhood growth and for genes associated with and predictive for the development of adult pre-hypertension in cuSGA. Pre-implantation gene expression can therefore be linked to the origins of adult disease.
1. LaManno, et al. (2018). Nature. 560(7719): 494-8.
2. Petropoulos, et al. (2016). Cell. 165(4): 1012-26.
3. Horikoshi, et al. (2016). Nature 538(7624): 248-252.
4. Rasmussen, et al. (2022). Nature. https://doi.org/10.1038/s41586-021-04249-w
5. Garner, et al (2021). JES. 5: A654–A655.
6. Garner, et al (2019) ESPE Abstracts 92FC12.6

Preimplantation embryo development is associated with both birth size and post-natal growth and can predict adult cardiovascular risk

Introduction: Early-onset obesity is a cardinal feature of rare genetic obesity disorders. According to the Endocrine Society guideline, genetic screening is indicated in selected cases with age of onset (AoO) of severe obesity (grade ≥2) <5 years. However, this cut-off is not validated.
Aims: To present the detailed BMI characteristics of children and adolescents with rare genetic obesity disorders; to evaluate whether the following growth chart characteristics can aid in assessing which children should be screened for genetic obesity disorders: AoO of obesity, AoO of severe obesity, BMI standard deviation scores (SDS) at yearly age intervals.
Methods: In this prospective observational study, children with non-syndromic and syndromic genetic obesity disorders treated at our tertiary obesity center were included. Growth measurements from birth onwards were collected. Children with obesity from a population-based cohort study with follow-up until age 10 years were included as an unselected reference cohort. Diagnostic performance (sensitivity [sens], specificity [spec], positive likelihood ratio [LR+], area-under-the-curve [AUC]) for AoO of obesity and severe obesity and for BMI-SDS at yearly age intervals was calculated.
Results: We included 64 children with genetic obesity disorders (32 non-syndromic, 32 syndromic) and 298 control children with obesity. At intake, median age of children with genetic obesity was 10.5 years (IQR 7.0–14.7) and mean BMI-SDS +3.1 ± 1.1. Median AoO of obesity was 1.2 years (IQR 0.6–3.7) in non-syndromic genetic obesity, 2.1 years (IQR 0.9–4.2) in syndromic genetic obesity, and 3.8 years (IQR 2.3–6.2) in the control population. For non-syndromic genetic obesity, optimal cut-off value for AoO of obesity was ≤1.5 years: sens 0.60, spec 0.88, LR+ 5.22, AUC 0.79 (p<0.001). For syndromic genetic obesity, optimal cut-off was ≤3.0 years: sens 0.68, spec 0.68, LR+ 2.06, AUC 0.67 (p=0.001). AoO of severe obesity showed worse diagnostic performance than AoO of obesity in both non-syndromic (AUC 0.58, p=0.20) and syndromic genetic obesity (AUC 0.58, p=0.21). Moreover, when the guideline cut-off <5 years was used, AoO of severe obesity showed a negative diagnostic performance (non-syndromic genetic obesity: sens 0.76, spec 0.13, LR+ 0.88; syndromic genetic obesity: sens 0.85, spec 0.14, LR+ 0.98). BMI-SDS showed good diagnostic performance for non-syndromic genetic obesity across the age intervals (AUCs 0.79-0.89, all P<0.001) but not for syndromic genetic obesity (AUCs 0.54-0.71, P-values ranging from <0.001-0.50).
Conclusion: This study shows that growth charts characteristics such as BMI-SDS and AoO of obesity (grade 1), but not AoO of severe obesity (grade ≥2), could be useful to distinguish between children with genetic obesity disorders and children with obesity from a population-based cohort study. However, all investigated growth charts characteristics showed misclassification, especially in syndromic genetic obesity, indicating that additional clinical features should be present to warrant genetic testing in these children.

Age of Onset of Obesity and Childhood BMI Trajectories in Rare Genetic Obesity Disorders

Objectives: In healthy, normal - weight adults there is a circadian rhythm in blood pressure (BP) and plasma catecholamines (epinephrine and norepinephrine): BP and the levels of norepinephrine and epinephrine decline during sleep, followed by a rapid increase during the early morning hours. Studies in adults with Type 2 diabetes (T2D) show blunted reductions in BP and plasma catecholamine levels during sleep that are associated with increased risks for heart failure, stroke, myocardial infarction, and sudden death. In this study, we explored if the circadian rhythm in catecholamines is disrupted in youth-onset T2D. We hypothesized that increased sympatho-adrenal activity during sleep differentiates youth with T2D from non-diabetic overweight/obese and lean youth of comparable age, pubertal status, and ethnicity. To that end, we measured urine catecholamines in fasting – morning-spot urines and 24-hour urine samples.
Methods: 56 non-diabetic adolescents with overweight/obesity ("obese"), 42 adolescents with T2D ("T2D"), and 43 normal-weight controls ("lean"); aged 12-21 years, were studied. None was diagnosed with hypertension, and none was on antihypertensive treatment. Weight, height, BMI, BMI%, and BP were extracted from medical charts. Body fat percent (BF%) was measured by TANITA. Stress levels were assessed using PHQ-2 and PHQ-9 questionnaires. Fractionated free urine catecholamines (epinephrine, norepinephrine, and dopamine) were analyzed by liquid chromatography/tandem mass spectrometry (LC/MS-MS) in both spot and 24-hour urines, normalized to urinary creatinine. The ratio of fasting morning urine catecholamines to 24-hour urine catecholamines was calculated to assess circadian variation in catecholamines. Group differences were assessed by Kruskal-Wallis or ANOVA.
Results: Groups were comparable for age (obese 14.8 +/- 1.9; T2D 15.7 +/- 2.1 and lean 14.9 +/- 1.9-yr), pubertal status, and ethnicity. Obese youth with and without T2D were predominantly female (T2D, 28 F, 14 M; obese 33 F, 23 M; lean, 17 F, 26 M); those with T2D had highest BF% (obese 37.3 +/- 9.5; T2D 42.9 +/- 9.9; lean 20.1 +/- 6.3 %; p=2.58e-22) and systolic blood pressure (obese 115.36 +/- 11.54; T2D 127.83 +/- 12.08 and lean 111.65 +/- 9.13 mmHg; p=5.49e-10).  Fractionated free urine catecholamines (epinephrine, norepinephrine and dopamine) were comparable across groups in 24-hour urines. However, fasting morning epinephrine levels and the ratio of fasting morning/24-hour epinephrine were higher in T2D (p=0.0035, and p=0.035, respectively) than in either lean or nondiabetic youth with obesity. There were no differences in morning urine catecholamines between lean controls and non-diabetic youth with obesity.
Conclusions: Our results suggest a disrupted circadian rhythm in catecholamines in youth-onset T2D, with a blunted overnight fall in epinephrine levels. Higher levels of epinephrine levels at night in youth with T2D might be associated with, or predispose to, hypertension and long-term cardiovascular complications.

Disrupted circadian rhythm in catecholamines in youth-onset Type 2 Diabetes

Context: Growth hormone (GH) plays a critical role in glucose homeostasis and linear growth during childhood. GH secretion is controlled by multiple neurotransmitters and hormones. Sleep is also a powerful stimulus for GH secretion. In children, GH pulses occur after sleep onset in association with slow-wave sleep (SWS), but it is unclear if undisrupted sleep is required for normal GH secretion. Given the rising incidence of pediatric sleep disorders such as obstructive sleep apnea (OSA) that cause sleep fragmentation, it is critical to determine whether sleep fragmentation during childhood may dysregulate the somatotropic axis.
Objective: To quantify how undisrupted vs fragmented sleep affects GH secretion in children.
Methods: Pubertal subjects (10 M, 9 F; 9.1-14.4 yrs) were randomized to two overnight polysomnographic studies with frequent blood sampling (q10min) to measure GH: fourteen were healthy subjects studied with or without SWS disruption via auditory stimuli (3 sec, 1500 Hz tones, 40-100 dB followed by 18 sec of a 75 dB noise simulating a knock on the door) and five were subjects with OSA studied with or without prescribed continuous positive airway pressure (CPAP); the latter condition was expected to have sleep fragmentation due to untreated OSA. GH pulses were identified using sparse deconvolution. Studies without auditory stimuli or when using CPAP were combined and analyzed as 'undisrupted sleep'. The relationship between GH pulse rate per hour and sleep stage was evaluated using Poisson generalized estimating equations.
Results: In SWS disruption studies, an average of 77.51 ± 13.61 (range 23-190) auditory stimuli were delivered, causing a 40.02 ± 7.78% decrease in SWS in favor of lighter sleep stages. In subjects with OSA, CPAP withdrawal caused a slight increase in the apnea hypopnea index (1.08 ± 0.10 vs. 2.20 ± 1.70 events/hour of sleep, p=0.07) with no change in sleep stage durations. During undisrupted sleep, subjects had an average of 0.91 ± 0.067 GH pulses/hour with a mean pulse amplitude of 10.03 ± 1.00 ng/mL and mean basal secretion rate of 0.26 ± 0.10 ng/ml per min. GH pulses occurred more frequently during SWS (2.3 pulses/hour SWS, amplitude 9.92 ± 2.59 ng/mL) than during any other sleep stage (P < 0.001). Neither form of sleep disruption affected the distribution of GH pulses across sleep stages, GH pulse amplitude and frequency, nor basal GH secretion.
Conclusion: In children, nocturnal GH pulses are temporally associated with SWS. However, neither acute fragmentation of SWS nor nocturnal arousals from OSA altered basal or pulsatile GH secretion, indicating that somatotrophic axis activity is preserved in the face of acute sleep disruption in children.

Acute sleep disruption does not diminish pulsatile growth hormone secretion in pubertal children

Objectives: We have previously demonstrated that insulin resistance (IR) in youth is associated with elevated levels of branched-chain amino acids (BCAAs).  BCAAs compete with aromatic amino acids (AAA) including tryptophan, the precursor of serotonin, for uptake into β-cells and other tissues via the large neutral amino acid transporter. Serotonin has been reported to increase β-cell mass and glucose-dependent insulin secretion. In this study we have explored how BCAA, tryptophan (one of the AAA), and a subset of their metabolites are modulated in youth-onset T2D. Based on our prior studies in neuronal BCAA metabolism, we hypothesized that elevated BCAA could induce diversion of tryptophan metabolism towards production of kynurenine rather than serotonin in youth with T2D. To test this, we analyzed 24-hour urine samples and compared levels of byproducts of BCAAs and tryptophan metabolism in obese youth with T2D with those in non-diabetic obese and lean youth of comparable age, pubertal status and ethnicity.
Methods: 56 non-diabetic adolescents with overweight/obesity ("obese"), 42 adolescents with T2D ("T2D"), and 43 normal weight controls ("lean"), ages 12-21 years-old were studied. Weight, height, BMI, BMI% were extracted from medical charts. Body fat percent (BF%) was measured by TANITA. We also measured metabolites derived from BCAA catabolism, including the branched-chain ketoacids (BCKAs), and tryptophan metabolism, including intermediates of the serotonergic and kynurenine pathways, in spot and 24-hour urine samples by liquid chromatography/tandem-mass spectrometry (LC/MS-MS). Levels were normalized to urine creatinine. Group differences were assessed by Kruskal Wallis or ANOVA.
Results: Groups were comparable for age (obese 14.8 +/- 1.9; T2D 15.7 +/- 2.1 and lean 14.9 +/- 1.9-yr), pubertal status, and ethnicity. Youth with T2D were predominantly female (T2D, 28 F, 14 M; obese 33 F, 23 M and lean, 17 F, 26 M), and had highest BF% (obese 37.3 +/- 9.5; T2D 42.9 +/- 9.9; lean 20.1 +/- 6.3 %; p=2.58e-22). In 24-hour urine samples, BCKAs, tryptophan, and kynurenine levels were higher in T2D (p=0.0002, p= 0.0045 and p= 0.00009 respectively) than in either lean controls or nondiabetic youth with obesity; in contrast, there were no differences between lean controls and non-diabetic youth with obesity. The levels of 5-HIAA, the principal metabolite of serotonin, were comparable across groups; however, the ratio of kynurenine/tryptophan was higher (p= 0.0112) in youth with T2D and the ratios of 5-HIAA/ tryptophan (p=0.027) and 5-HIAA/Kynurenine (p=0.0067) were lower compared to the other two groups. Those ratios were comparable between lean controls and non-diabetic youth with obesity.
Conclusions: Increased BCKAs are accompanied by diversion of tryptophan metabolism from the serotonin pathway to the kynurenine pathway, suggesting perturbations in both BCAA and AAA metabolism in youth-onset T2D. These alterations could contribute to development of beta-cell dysfunction and progression to T2D in youth.

Branched-chain Amino Acid and Tryptophan Metabolism and the Pathogenesis of Youth-Onset Type 2 Diabetes Mellitus (T2D)

Type 1 diabetes (T1D) is a complex disease characterized by the loss of pancreatic beta-cells and lifelong dependence on insulin replacement therapy. As the worldwide incidence of T1D has increased and the prevalence of high-risk HLA haplotypes has declined among new diagnoses, a greater focus has been placed on environmental factors contributing to T1D pathogenesis.  Lactiplantibacillus plantarum 299v (Lp299v) supplement is reported to increase plasma and stool levels of anti-inflammatory short-chain fatty acids (SCFAs) and promote IL-10 signaling in colonic derived macrophages and T-cells.  Therefore, we investigated the effect of Lp299v supplement on T1D pathogenesis in diabetes-prone BioBreeding DRlyp/lyp rats.  Rats were weaned at 21 days of age (D21) onto a normal cereal diet (ND) or a gluten-free hydrolyzed casein diet (HCD), with or without daily Lp299v supplementation. All rats provided ND developed diabetes by D83 and supplementing ND with Lp299v did not significantly delay progression.  Rats provided HCD showed a significant delay in diabetes onset over ND, while progression was even more delayed in the HCD+Lp299v group, with 25% of animals remaining diabetes-free to D130 (study end).  Relative to the other groups, HCD+Lp299v rats exhibited significant increases in the plasma levels of the SCFAs propionate and butyrate (beneficial metabolites produced by intestinal bacteria). In addition, stool microbiota showed increased abundances of SCFA-producing taxa at D40.  Gene expression analysis of pancreatic islets was conducted at D40, prior to insulitis. Lp299v supplementation favorably modulated islet expression levels of pathways related to endoplasmic reticulum (ER) stress, and the major arms of the unfolded protein response (UPR), especially in combination with the HCD.  Notably, ER stress has been implicated in the formation of islet neoantigens that may underlie the initial loss of immune tolerance in T1D.  Lp299v supplement differentially modulated the UPR to favor the expression of the transcription factor Atf6, increase expression of transcripts related to cell survival/proliferation, and the ER-associated protein degradation (ERAD) pathway of the UPR.  This contrasted with HCD alone, which promoted a transcriptional program related to autophagy.  To colocalize and confirm activities detected in the transcriptional analyses to β-cells, pancreata of D40 rats were subjected to dual immunofluorescence staining for insulin and UPR-related proteins.  Consistent with elevated ERAD activity, significantly greater EIF4G1 and OS9 expression were observed in HCD+Lp299v islets. Phospho-PERK staining was significantly higher in islets of the HCD-alone group, consistent with the elevated expression of autophagy transcripts. Analysis of plasma proinsulin: C-peptide ratios revealed that Lp299v alone could improve islet function, as significant decreases were measured in all groups relative to ND.  While ongoing studies aim to define the bacterial metabolites underlying the favorable outcomes, these studies suggest that Lp299v supplementation improves proteostasis capacity in β-cells and slows T1D progression in DRlyp/lyp rats.

Probiotic Supplementation with Lactiplantibacillus Plantarum 299v Modulates β-Cell Endoplasmic Reticulum Stress and Prevents Type 1 Diabetes in Biobreeding Rats

Background: Peripubertal hyperandrogenemia (HA) often represents a forerunner to adolescent/adult polycystic ovary syndrome (PCOS), but risk factors for peripubertal HA are unclear. Studies have associated obesity with HA throughout puberty (e.g., McCartney, JCEM 2007;92: 430-436, Knudsen, Obesity 2010;18: 2118-2124). However, these studies were performed in academic centers and could have been influenced by recruitment bias. We therefore investigated if there was a difference in calculated free testosterone (T) levels between girls with healthy weight and those with obesity in a nationally-representative U.S. sample.
Methods: We analyzed data for 1,196 girls aged 6–18 years from the National Health and Nutrition Examination Survey (NHANES) 2013-2016 databases, utilizing a cross-sectional study design. We calculated free T from total T and sex hormone-binding globulin (Vermuelen equation). We divided girls into 3 groups by age, based on presumed pubertal stages: "Pre-/Early" (6–9 years), "Middle" (10–14 years), and "Late" (15–18 years). Each individual was also categorized as healthy-weight (BMI-for-age percentile 5-85) or obese (BMI-for-age percentile ≥95). We analyzed mean free T concentrations by weight status (healthy-weight vs. obese) for each age group by calculating population estimates. We also performed survey-weighted regression analysis to assess the relationships between weight status and age category (independent variables) and free T (outcome).
Results: Among Pre-/Early girls, free T was 0.33 pg/ml (0.28-0.38) [mean (95% CI)] for healthy-weight vs. 0.86 pg/ml (0.67-1.05) for obese. Among Middle girls, free T was 2.53 pg/ml (2.27-2.78) for healthy-weight vs. 4.35 pg/ml (3.75-4.95) for obese. Among Late girls, free T was 3.33 pg/ml (2.96-3.70) for healthy-weight vs. 5.64 pg/ml (4.93-6.36) for obese. (Notably, in each age group, the 95% CIs for free T did not overlap between healthy-weight vs. obese subgroups). Mean free T increased with advancing age in both healthy-weight and obese groups, especially between Pre-/Early and Middle age ranges (Pre-/Early-to-Middle differences were 2.20 pg/ml for healthy-weight and 3.49 pg/ml for obese). Our regression model accounted for 45% of the variance in free T (R²=0.45). Both weight status (obese vs. healthy-weight) and age group were independent predictors of free T concentrations (p<0.0001 for both).  When adding an age-in-months-by-BMI-percentile interaction term, the model R² value was 0.47, with age category, weight category, and the interaction term all being independent predictors of free T (p<0.001 for all).  Conclusion: We conclude that obesity is associated with higher free T levels in U.S. girls aged 6–18 years. Age also predicts free T, likely reflecting increases related to pubertal maturation, and free T differences according to weight status are more prominent in older girls. This study in a nationally-representative sample of U.S. girls supports previous analyses of girls studied at academic centers, further suggesting an important relationship between obesity and peripubertal androgen excess.

Obesity is Associated with Adolescent Hyperandrogenemia in Girls Aged 6-18, National Health and Nutrition Examination Survey (NHANES) 2013-2016

PCOS confers an increased risk for obesity-related comorbidities. It has been challenging to identify precise risk genes associated with the pathogenesis of PCOS. A novel variant located within MYO10 intron 2 (rs9312937) was identified in a PCOS GWAS in the genetically isolated Finnish and related Estonian populations. We identified PCOS risk variants at the MYO10 locus in additional PCOS populations. We hypothesized that PCOS risk variants at the locus were associated with metabolic traits.
Women from Iceland (n=1347) and Boston (n=591) were diagnosed with PCOS using the Rotterdam criteria and NIH criteria, respectively, with a non-overlapping replication cohort (n=586) identified using an algorithm incorporating ICD codes and natural language processing. Controls included population controls from Iceland (effective n=1200), well-phenotyped controls from Boston (n=441) and a cohort with no PCOS diagnoses or features in the electronic medical record (EMR; n=749). We performed an association study to identify variants in the MYO10 locus associated with PCOS. We also examined the association of variants in MYO10 and 15 independent phenotypic traits in women of European ethnicity from the Boston PCOS cohort. Data were log-normalized and analyzed using linear regression, then corrected for the false discovery rate using the Benjamini-Hochberg procedure.
Two novel risk variants for PCOS were identified within intron 2 of MYO10, rs31506-T (p = 1.6X10-05, OR 0.83 [95% CI 0.77-0.91]) and rs253336-A (p = 2.4X10-06, 0.82 [0.75-0.89]). The rs31506 variant is in moderate linkage disequilibrium with the novel variant rs9312937 (r² ≥ 0.4) and rs253336 (r² ≥ 0.4). The variant rs31506 maps to enhancer sites in pancreas, osteoblast and fetal adrenal cells, whereas rs253336 maps to methylation sites in adipose stem cells and ovary (pubs.broadinstitute.org/mammals/haploreg). Neither variant had eQTLs, and expression was highest in lung, thyroid, neurons, breast and ovary (gtexportal.org).
Among cases and controls, the variant rs253336 was associated with BMI (β=0.015±0.0057; p=0.01), waist circumference (β=0.011±0.0047; p=0.02), hip circumference β=0.010±0.0035; p=0.003) and triglycerides (β=0.028±0.013; p=0.03). The variant rs31506 was also associated with BMI (β=0.0065±0.0027; p=0.02), waist circumference (β=0.0051±0.0022; p=0.02), hip circumference (β=0.0039±0.0017; p=0.02), HDL (β=-0.0083±0.0037; p=0.02) and triglycerides (β=0.018±0.0063; p=0.003). The association between rs253336 and hip circumference, and rs31506 and triglyceride levels remained significant after correction. There was no association between the variants and BMI in the subjects identified through the EMR.
We identified two novel risk variants for PCOS in intron 2 of MYO10 (rs31506 and rs253336), a gene recently implicated within a risk locus for PCOS. We also demonstrated that the variants are associated with hip circumference and triglyceride levels. Previous data demonstrate increased expression of MYO10 in omental adipose of obese type 2 diabetics and an association with lean body mass. The data suggest a potential role for MYO10 in the metabolic etiology of PCOS.

Risk Variants at the MYO10 Locus are Associated with Metabolic Traits

Introduction and Purpose: Polycystic Ovary Syndrome (PCOS) is the most common endocrine disorder in women of reproductive age. Notably, PCOS women with hyperandrogenism have a pronounced increased risk for cardio-metabolic comorbidities. The gut microbiome is responsible for fermentation of indigestible fiber to produce short-chain fatty acids (SCFA), which help maintain metabolic function and reduce gut inflammation. Alteration in the gut microbiome, or dysbiosis, affects overall metabolic homeostasis and contributes to pathogenesis of metabolic diseases. While dysbiosis has been observed in women with PCOS, the molecular mechanisms of hyperandrogenemia-induced dysbiosis and its contribution to SCFA excretion and gut inflammation have not been studied in PCOS.
Methods: Four-week old C57BL/6N female mice were implanted subcutaneously with dihydrotestosterone (DHT, 8.0 mg) or vehicle silastic tubes (n=8/grp). Animals were treated day 90 and the gastrointestinal tract was collected. Expression levels of mRNA were assessed by RT-qPCR. Fecal microbiota was assessed by 16S rRNA gene sequencing and profiled using MicrobiomeAnalyst; SCFA concentrations were assessed by LC-MS.
Results: DHT-treated female mice had significantly increased lean mass (21.35 ± 0.84 vs 19.25 ± 0.55 g, p<0.05) and increased fat mass (5.18 ± 1.45 vs 2.73 ± 0.64 g, p<0.05) compared to the vehicle counterparts. DHT-treated female mice showed increased HbA1C (5.18 ± 1.45 vs 2.73 ± 0.64 %, p<0.05). Hyperandrogenemic females showed altered gastrointestinal morphology, having decreased small intestine (33.03±2.05 vs 28.69±2.604 cm, p<0.05) and colon length (7.88 ± 0.64 vs 6.60 ± 0.55 cm, p<0.05) as well as increased immune cell infiltration assessed by histology. DHT-treated female mice have an increase in the relative percent abundance of Proteobacteria (2.04 ± 0.62 vs 0.35 ± 0.23%, p<0.05) and in the ratio of Firmicutes/Bacteriodetes (1.92 ± 0.47 vs 1.11 ± 0.17, p<0.05). Fecal concentrations of short-chain fatty acids were higher in DHT-females for butyric acid (92.33 vs 404.8 ng/mg feces, p<0.05), acetic acid (571.8 vs 1592 ng/mg feces, p<0.05), and propionic acid (80.59 vs 167.3 ng/mg feces, p<0.05). However, expression of the SCFA receptor Ffar2 in the colon was upregulated by DHT (1.31 ± 0.15-fold, p<0.05), indicating the possibility of a compensation to ameliorate increased SCFA excretion. The mRNA expression of lipopolysaccharide induced inflammation markers were significantly upregulated by DHT in the proximal colon (Litaf, 1.69 ± 0.29-fold; Tlr4 1.27 ± 0.15-fold, p<0.05) compared with vehicles.
Conclusions: Androgen-mediated gut dysbiosis may contribute to altered glucose handling, adiposity, and metabolic homeostasis in PCOS. Increased elimination of SCFA in DHT-treated female mice may indicate less efficient absorption, which can contribute to colonic inflammation, altered gastrointestinal morphology, and increased gut permeability. Together, our results highlight the potential role of androgen-mediated gut dysbiosis to influence metabolic dysfunction.

Contribution of Hyperandrogenemia to Dysbiosis, Gut Inflammation, and Dysregulation of Short-Chain Fatty Acid Homeostasis in Mouse Model of Polycystic Ovary Syndrome

Objective: Adolescent polycystic ovary syndrome (PCOS) is a heterogenous clinical syndrome characterized by hyperandrogenism and irregular menses. Heterogeneous features of the syndrome include the presence/absence of obesity, degree of insulin resistance, and clinical and/or biochemical hyperandrogenism. In adult women with PCOS, discrete metabolic and reproductive subgroups can be identified. A recent study described distinct phenotypes using unsupervised hierarchical cluster analysis: a metabolic subgroup characterized by higher BMI, glucose, and insulin levels, and a reproductive subgroup demonstrating higher LH and SHBG levels. We hypothesized that non-obese and obese adolescent girls with PCOS would exhibit similar distinct phenotypes at the time of diagnosis.

Methods: Data were extracted from a multi-site PCOS registry including seven academic centers in the US with the following inclusion criteria: PCOS confirmed per 2018 international guidelines, diagnosis before age 18 years, and at least one follow-up visit. Data collected included demographics, medical and family history, physical characteristics including hirsutism, acanthosis nigricans, and laboratory measures such as HbA1c, ALT, lipid profiles and reproductive hormones (androgens, LH, FSH, SHBG). Hormone levels were normalized to the upper limit provided by the measuring laboratory. Patients were divided by obesity status, defined as obese (OB) with BMI ≥ 95th percentile for age and sex and non-obese (NO) < 85th percentile for age and sex. Differences in continuous variables were assessed with unpaired t-tests using GraphPad Prism.

Results: Complete diagnostic data from 390 patients from 7 different sites were included for analysis, with 284 OB (Age 15.3±1.5 years, BMI 36.1±5.6 kg/m²) and 68 NO (Age 15.8±1.3 years, BMI 22.6±3.4 kg/m²). OB presented with more severe acanthosis (p<0.0001) and higher HbA1c (OB 5.5±0.5, NO 5.2±0.3; %, p<0.0001), ALT (OB 35±30, NO 22±17; IU/L, p=0.0006), and triglycerides (OB 145±90, NO 110,±88; mg/dL, p=0.002). OB also presented with higher free testosterone (OB 219±188% vs NO 158±125% above upper limit of normal; p=0.003), while NO presented with higher total testosterone (OB 121±68 %, NO 152±88 % above upper limit of normal; p=0.003), androstenedione (OB 87±37 %, NO 125±62 above upper limit of normal; p=0.002), and LH (OB 9.7±4.9, NO 15.2±10.4; mIU/mL, p=0.0001). No differences in Ferriman Gallwey score or number of menses in the 6 months prior to diagnosis were detected.

Conclusions: Metabolic and reproductive phenotypes differed in OB and NO adolescent girls at the time of PCOS diagnosis. The reproductive phenotype in NO was characterized by increased LH, total testosterone and androstenedione levels. In contrast, OB girls had a phenotype characterized by insulin resistance-driven hyperandrogenemia established by lower SHBG and higher free testosterone levels. Our data are consistent with findings reported for adult women with two divergent phenotypes: an insulin resistance-driven metabolic subtype and a primarily reproductive subtype, potentially driven by neuroendocrine mechanisms.

Deep learning-based voice screening technique for cystic fibrosis related diabetes

Downloads

Next from ENDO 2022

Preimplantation embryo development is associated with both birth size and post-natal growth and can predict adult cardiovascular risk

Similar lecture

Ethnic Differences in Albuminuria among Adults with Diabetes and Normal Range Estimated Glomerular Filtration Rate (eGFR).

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES