| Sumari: | <p><strong>Background:</strong> Frailty and cardiovascular disease (CVD) are an increasing burden in rapidly ageing societies like China. While frailty and CVD commonly coexist, there is substantial uncertainty about their relationships particularly about the relevance of modifiable CVD risk factors (RFs) for frailty and frailty for CVD. Moreover, frailty research is constrained by a lack of consensus on the optimum definition for frailty with limited evidence on their relative performance. This thesis first compared the measurement properties of the most widely used frailty models, the frailty index (FI) and frailty phenotype (FP), to guide selection of a frailty instrument, and second investigated the associations of CVD RFs with frailty and of frailty with CVD in Chinese adults.</p>
<p><strong>Methods:</strong> The validity and reliability of the FI and FP were compared using a systematic review of the predictive accuracy for all-cause mortality and a prospective analysis of repeat frailty assessments from the Irish Longitudinal Study on Ageing (TILDA), respectively. The China Kadoorie Biobank, involving 0.5 million adults aged 30-79 years recruited between 2004-08 from 10 diverse regions, was used to assess the associations of CVD RFs (systolic blood pressure [SBP], random plasma glucose [RPG], alcohol consumption, smoking, fat mass, and lean body mass) with incident frailty and frailty with incident CVD (stroke and ischaemic heart disease [IHD]) among participants without prevalent CVD. Logistic and Cox regression yielded adjusted odds (ORs) and hazard ratios (HRs), respectively. Frailty and CVD associations were corrected for regression dilution bias. Modifications of the FI (excluding relevant cardiovascular items) assessed the presence of circularity in relationships.</p>
<p><strong>Results:</strong> The comparative analyses of the measurement properties of the main frailty instruments found that continuous instruments were more accurate predictors of all-cause mortality and FI was more reliable than FP (kappa statistic [95%CI)] 0.60 [0.58-0.62] vs. 0.37 [0.35-0.40]), supporting the selection of FI for subsequent analyses in CKB. Among ~20,000 CKB participants without prevalent frailty or CVD, positive associations were found between SBP, RPG, fat mass, and incident frailty after adjustment for confounders, but associations were substantially attenuated after accounting for circularity. During 13-years’ follow-up, 54,193 incident stroke events (46,043 ischaemic [IS] and 10,535 haemorrhagic [HS] stroke) and 49,937 IHD events (7,749 myocardial infarction [MI] and 10,193 major coronary event [MCE]) were recorded. Among ~0.5M CKB participants without prevalent CVD, there were positive log-linear associations between frailty and incident stroke and IHD (and their subtypes) after adjustment for confounders, which became stronger with correction for regression dilution bias. Each 1 SD higher usual FI was associated with HRs (95%CI) of 1.31 (1.29-1.33) and 1.37 (1.35-1.40) for incident stroke and IHD, respectively, which attenuated to 1.13 (1.11-1.15) and 1.34 (1.31-1.36), respectively, after accounting for circularity. Associations were stronger for fatal than overall events, amounting to nearly 2-fold differences, and there were some differences in the strength of associations by subtypes of stroke (IS: 1.28 [1.25-1.30] and HS: 1.46 [1.41-1.51]) and IHD (MI: 1.44 [1.38-1.50] and MCE: 1.51 [1.46-1.57]).</p>
<p><strong>Conclusions:</strong> Among Chinese adults without prior CVD, there was no clear association between the modifiable CVD RFs and risk of frailty after accounting for circularity in the relationships. However, frailty was associated with significantly increased risk of major CVDs, which was stronger for IHD than for stroke and fatal events, and largely similar within each type.</p>
|
|---|