Detecting lifestyle risk factors for chronic kidney disease with comorbidities: Association rule mining analysis of web-based survey data

Suyuan Peng, Feichen Shen, Andrew Wen, Liwei Wang, Yadan Fan, Xusheng Liu, Hongfang Liu

Research output: Contribution to journalReview articlepeer-review

3 Scopus citations

Abstract

Background: The rise in the number of patients with chronic kidney disease (CKD) and consequent end-stage renal disease necessitating renal replacement therapy has placed a significant strain on health care. The rate of progression of CKD is influenced by both modifiable and unmodifiable risk factors. Identification of modifiable risk factors, such as lifestyle choices, is vital in informing strategies toward renoprotection. Modification of unhealthy lifestyle choices lessens the risk of CKD progression and associated comorbidities, although the lifestyle risk factors and modification strategies may vary with different comorbidities (eg, diabetes, hypertension). However, there are limited studies on suitable lifestyle interventions for CKD patients with comorbidities. Objective: The objectives of our study are to (1) identify the lifestyle risk factors for CKD with common comorbid chronic conditions using a US nationwide survey in combination with literature mining, and (2) demonstrate the potential effectiveness of association rule mining (ARM) analysis for the aforementioned task, which can be generalized for similar tasks associated with noncommunicable diseases (NCDs). Methods: We applied ARM to identify lifestyle risk factors for CKD progression with comorbidities (cardiovascular disease, chronic pulmonary disease, rheumatoid arthritis, diabetes, and cancer) using questionnaire data for 450,000 participants collected from the Behavioral Risk Factor Surveillance System (BRFSS) 2017. The BRFSS is a Web-based resource, which includes demographic information, chronic health conditions, fruit and vegetable consumption, and sugar- or salt-related behavior. To enrich the BRFSS questionnaire, the Semantic MEDLINE Database was also mined to identify lifestyle risk factors. Results: The results suggest that lifestyle modification for CKD varies among different comorbidities. For example, the lifestyle modification of CKD with cardiovascular disease needs to focus on increasing aerobic capacity by improving muscle strength or functional ability. For CKD patients with chronic pulmonary disease or rheumatoid arthritis, lifestyle modification should be high dietary fiber intake and participation in moderate-intensity exercise. Meanwhile, the management of CKD patients with diabetes focuses on exercise and weight loss predominantly. Conclusions: We have demonstrated the use of ARM to identify lifestyle risk factors for CKD with common comorbid chronic conditions using data from BRFSS 2017. Our methods can be generalized to advance chronic disease management with more focused and optimized lifestyle modification of NCDs.

Original languageEnglish (US)
Article numbere14204
JournalJournal of medical Internet research
Volume21
Issue number12
DOIs
StatePublished - Dec 1 2019

Keywords

  • Association rule mining
  • Behavioral risk factor surveillance system
  • Chronic kidney disease
  • Noncommunicable diseases

ASJC Scopus subject areas

  • Health Informatics

Fingerprint

Dive into the research topics of 'Detecting lifestyle risk factors for chronic kidney disease with comorbidities: Association rule mining analysis of web-based survey data'. Together they form a unique fingerprint.

Cite this