Meta-Analysis and Network Analysis Differentially Detect Various Pro-Inflammatory Mediators and Risk Factors for Type 2 Diabetes in the Elderly

Abstract Type 2 diabetes (T2D) has a pathophysiological component that includes inflammation. Inflammation-sensitive marker measurement may be helpful in determining the risk of complications for both older T2D patients and the public. This study aimed to investigate the association between blood pro-inflammatory mediators and the characteristics of elderly patients with T2D using meta and network analyses. The Web of Science, Scopus, PubMed, and Cochrane Library databases were selected as study methodology. The Quality in Prognosis Studies (QUIPS) tool in the meta-analysis assessed the studies’ methodological quality. The selected studies were statistically analyzed using the META-MAR tool based on the standardized mean difference (SMD). The selected studies included nine examinations involving 6399 old people [+>+55 years old, 65.9+±+4.09 (mean+±+SD)]. The meta-analysis showed that pro-inflammatory mediators (SMD 0.82) and patient-related variables [risk factors (SMD 0.71)] were significantly associated with T2D (p+<+0.05). Subgroup analysis revealed that tumor necrosis factor alpha (TNF-α; SMD 1.08), body mass index (SMD 0.64), high-density lipoprotein (HDL; SMD –0.61), body weight (SMD 0.50), and blood pressure (SMD 1.11) were factors significantly associated with T2D (p+<+0.05). Network analysis revealed that among patient characteristics, diastolic blood pressure and, among inflammatory mediators, leptin were the most closely associated factors with T2D in older adults. Moreover, network analysis showed that TNF-α and systolic blood pressure were most closely associated with leptin. Overall, alternate techniques, such as meta-analysis and network analysis, might identify different markers for T2D in older people. A therapeutic decision-making process needs to consider these differences in advance.