Association of the Triglyceride–Glucose Index with Ambulatory Blood Pressure Categories and Circadian Blood Pressure Phenotypes in Untreated Adults: A Single-Centre, ABPM-Based Retrospective Cross-Sectional Study
TL;DR
Higher TyG index values were associated with progressively higher ambulatory BP categories and with reverse- or extreme-dipper circadian BP phenotypes after multivariable adjustment, which should be interpreted as hypothesis-generating because of the retrospective cross-sectional design.
Abstract
Background/Objectives: The triglyceride–glucose (TyG) index, a surrogate marker associated with insulin resistance, has been linked to both blood pressure elevation and circadian blood pressure dysregulation; however, its relationship with ambulatory blood pressure (BP) phenotypes across the full spectrum of contemporary ambulatory BP categories remains incompletely characterised. This study investigated the association of the TyG index with ambulatory BP categories and circadian BP phenotypes in untreated adults undergoing ambulatory blood pressure monitoring (ABPM). Methods: This retrospective cross-sectional study included 723 consecutive adults with no current or previous exposure to antihypertensive medication (treatment-naïve status refers exclusively to antihypertensive pharmacotherapy) who underwent 24 h ABPM and same-day laboratory assessment at a tertiary referral centre. Participants were classified according to the 2024 European Society of Cardiology ambulatory BP thresholds into non-elevated BP (n = 199), elevated BP (n = 243), or hypertension (n = 281) groups. Circadian BP phenotypes were categorised as extreme-dippers, dippers, non-dippers, and reverse-dippers. Associations between the TyG index and ambulatory BP categories were evaluated using multivariable proportional-odds ordinal logistic regression. The joint effect of BP category and circadian phenotype was examined by adding a TyG index × circadian phenotype interaction term to the ordinal model, and circadian phenotypes were characterised using multinomial logistic regression. Results: The TyG index increased progressively across ambulatory BP categories (8.8 ± 0.5, 9.1 ± 0.5, and 9.2 ± 0.6; p-trend < 0.001). After adjustment for age, sex, HbA1c, estimated glomerular filtration rate, and body mass index, each one-standard-deviation increase in the TyG index was associated with higher ambulatory BP category (OR 1.57, 95% CI 1.34–1.83; p < 0.001). The Brant test provided no evidence against the proportional-odds assumption (p = 0.379). In an exploratory analysis, the TyG–BP category association differed across circadian phenotypes (TyG × phenotype interaction, likelihood-ratio χ2 = 8.26, df = 3, p = 0.041). In multinomial analysis, a higher TyG index was associated with the reverse-dipper (OR 2.90 per SD, 95% CI 2.06–4.08) and extreme-dipper (OR 2.77, 95% CI 1.87–4.10) phenotypes relative to the results for the dippers (both p < 0.001), but not with non-dipping (OR 1.02, 95% CI 0.84–1.24; p = 0.86) after multivariable adjustment. Conclusions: Among untreated adults undergoing ABPM, higher TyG index values were associated with progressively higher ambulatory BP categories and with reverse- or extreme-dipper circadian BP phenotypes after multivariable adjustment. Because insulin resistance was not measured directly, these results describe an association between the TyG index, a surrogate marker, and ambulatory BP phenotypes; they should be interpreted as hypothesis-generating because of the retrospective cross-sectional design. Prospective studies are warranted to determine their temporal and clinical significance.