Associations between clustering of hypoglycemic symptoms, psychological traits, and problem-solving abilities in adults with type 1 diabetes: baseline data analysis of the PR-IAH study.

IF 1.3 Q4 ENDOCRINOLOGY & METABOLISM

Diabetology International Pub Date : 2025-01-16 eCollection Date: 2025-04-01 DOI:10.1007/s13340-024-00788-5

Naoki Sakane, Ken Kato, Sonyun Hata, Erika Nishimura, Rika Araki, Kunichi Kouyama, Masako Hatao, Yuka Matoba, Yuichi Matsushita, Masayuki Domichi, Akiko Suganuma, Seiko Sakane, Takashi Murata, Fei Ling Wu

{"title":"Associations between clustering of hypoglycemic symptoms, psychological traits, and problem-solving abilities in adults with type 1 diabetes: baseline data analysis of the PR-IAH study.","authors":"Naoki Sakane, Ken Kato, Sonyun Hata, Erika Nishimura, Rika Araki, Kunichi Kouyama, Masako Hatao, Yuka Matoba, Yuichi Matsushita, Masayuki Domichi, Akiko Suganuma, Seiko Sakane, Takashi Murata, Fei Ling Wu","doi":"10.1007/s13340-024-00788-5","DOIUrl":null,"url":null,"abstract":"Background: Precision medicine in diabetes care requires a dedicated focus on hypoglycemic symptoms. This study explored the associations between clustering of hypoglycemic symptoms, psychological characteristics, and problem-solving capabilities in adults with type 1 diabetes (T1D).Methods: A total of 251 adults with T1D participated in this survey. Hierarchical clustering was used to analyze 11 hypoglycemic symptoms (Edinburgh scale). The data included diabetic complications, fear of hypoglycemia, depressive symptoms, hypoglycemia problem-solving scale (HPSS), and treatment details. For predicting clusters and identifying feature importance, we utilized a machine learning approach.Results: Three distinct clusters were observed; individuals not sensitive to autonomic or neuroglycopenic symptoms (cluster 1, n = 138), those sensitive to both autonomic and neuroglycopenic symptoms (cluster 2, n = 19), and those sensitive to autonomic but not neuroglycopenic symptoms (cluster 3, n = 94). Compared to cluster 1, individuals from clusters 2 and 3 were of younger age, had higher fear of hypoglycemia, increased depressive symptoms, and greater use of continuous subcutaneous insulin infusion. Cluster 2 displayed enhanced HPSS scores, indicating better detection control and a more proactive approach to seeking preventive strategies than cluster 1. The accuracy for classifying into 3 clusters using machine learning was 88.2%. The feature importance of random forest model indicated that hunger, shaking, palpitation, sweating, and confusion were the top five important factors for predicting clusters.Conclusion: This study identified three distinct clusters of adults with T1D. These findings may provide valuable insights for diabetes professionals seeking to educate these individuals on how to manage hypoglycemia effectively.Trial registration: University Hospital Medical Information Network (UMIN) Center: UMIN000039475); approval date: February 13, 2020.","PeriodicalId":11340,"journal":{"name":"Diabetology International","volume":"16 2","pages":"294-302"},"PeriodicalIF":1.3000,"publicationDate":"2025-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11954777/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Diabetology International","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1007/s13340-024-00788-5","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2025/4/1 0:00:00","PubModel":"eCollection","JCR":"Q4","JCRName":"ENDOCRINOLOGY & METABOLISM","Score":null,"Total":0}

引用次数: 0

Abstract

Background: Precision medicine in diabetes care requires a dedicated focus on hypoglycemic symptoms. This study explored the associations between clustering of hypoglycemic symptoms, psychological characteristics, and problem-solving capabilities in adults with type 1 diabetes (T1D).

Methods: A total of 251 adults with T1D participated in this survey. Hierarchical clustering was used to analyze 11 hypoglycemic symptoms (Edinburgh scale). The data included diabetic complications, fear of hypoglycemia, depressive symptoms, hypoglycemia problem-solving scale (HPSS), and treatment details. For predicting clusters and identifying feature importance, we utilized a machine learning approach.

Results: Three distinct clusters were observed; individuals not sensitive to autonomic or neuroglycopenic symptoms (cluster 1, n = 138), those sensitive to both autonomic and neuroglycopenic symptoms (cluster 2, n = 19), and those sensitive to autonomic but not neuroglycopenic symptoms (cluster 3, n = 94). Compared to cluster 1, individuals from clusters 2 and 3 were of younger age, had higher fear of hypoglycemia, increased depressive symptoms, and greater use of continuous subcutaneous insulin infusion. Cluster 2 displayed enhanced HPSS scores, indicating better detection control and a more proactive approach to seeking preventive strategies than cluster 1. The accuracy for classifying into 3 clusters using machine learning was 88.2%. The feature importance of random forest model indicated that hunger, shaking, palpitation, sweating, and confusion were the top five important factors for predicting clusters.

Conclusion: This study identified three distinct clusters of adults with T1D. These findings may provide valuable insights for diabetes professionals seeking to educate these individuals on how to manage hypoglycemia effectively.

Trial registration: University Hospital Medical Information Network (UMIN) Center: UMIN000039475); approval date: February 13, 2020.

查看原文本刊更多论文

成人1型糖尿病患者低血糖症状、心理特征和问题解决能力聚类之间的关联：PR-IAH研究的基线数据分析

背景：糖尿病护理中的精准医学需要专注于低血糖症状。本研究探讨了成人1型糖尿病（T1D）患者低血糖症状聚类、心理特征和问题解决能力之间的关系。方法：对251名成年T1D患者进行调查。采用分层聚类法对11例低血糖症状（爱丁堡量表）进行分析。数据包括糖尿病并发症、对低血糖的恐惧、抑郁症状、低血糖问题解决量表（HPSS）和治疗细节。为了预测聚类和识别特征的重要性，我们使用了机器学习方法。结果：观察到三个明显的簇；对自主神经或神经低糖症状不敏感的个体（第1组，n = 138），对自主神经和神经低糖症状均敏感的个体（第2组，n = 19），以及对自主神经但不神经低糖症状敏感的个体（第3组，n = 94）。与聚类1相比，聚类2和聚类3的个体年龄更小，对低血糖有更高的恐惧，抑郁症状增加，并且更多地使用持续皮下胰岛素输注。与集群1相比，集群2显示出更高的HPSS评分，表明更好的检测控制和更积极主动地寻求预防策略。使用机器学习将其分类为3类的准确率为88.2%。随机森林模型的特征重要性表明，饥饿、颤抖、心悸、出汗和困惑是预测聚类的前五大重要因素。结论：本研究确定了三种不同类型的成人T1D。这些发现可能为糖尿病专业人员提供有价值的见解，以教育这些人如何有效地管理低血糖。试验注册：大学医院医疗信息网络中心（UMIN）： UMIN000039475；批准日期：2020年2月13日。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Diabetology International ENDOCRINOLOGY & METABOLISM-

CiteScore

3.90

自引率

4.50%

发文量

期刊介绍： Diabetology International, the official journal of the Japan Diabetes Society, publishes original research articles about experimental research and clinical studies in diabetes and related areas. The journal also presents editorials, reviews, commentaries, reports of expert committees, and case reports on any aspect of diabetes. Diabetology International welcomes submissions from researchers, clinicians, and health professionals throughout the world who are interested in research, treatment, and care of patients with diabetes. All manuscripts are peer-reviewed to assure that high-quality information in the field of diabetes is made available to readers. Manuscripts are reviewed with due respect for the author''s confidentiality. At the same time, reviewers also have rights to confidentiality, which are respected by the editors. The journal follows a single-blind review procedure, where the reviewers are aware of the names and affiliations of the authors, but the reviewer reports provided to authors are anonymous. Single-blind peer review is the traditional model of peer review that many reviewers are comfortable with, and it facilitates a dispassionate critique of a manuscript.