Chemerin and Arterial Stiffness in the Brisighella Heart Study: A Residual-Based Discordance Analysis

Figures

• 

  Figure 1. Flow-chart of participant selection. Selection of the complete-case analytical sample from the Brisighella Heart Study database.

• 

  Figure 2. Association between circulating chemerin and pulse wave velocity. Scatterplot showing the relationship between circulating chemerin and pulse wave velocity in the complete-case analytical sample. Each point represents one individual. The blue solid line represents the unadjusted linear regression fit. Orange markers indicate the mean PWV values plotted at the median chemerin value within each tertile of observed chemerin, and the orange line connects these points to illustrate the distributional pattern. In the multivariable model adjusted for age, sex, body mass index, mean arterial pressure, resting heart rate, current smoking status, and estimated glomerular filtration rate, each one-standard-deviation increase in chemerin was associated with a 0.093 m/s higher PWV (95% confidence interval, −0.026 to 0.212; p = 0.126). Abbreviations: CI, confidence interval; PWV, pulse wave velocity; SD, standard deviation.

• 

  Figure 3. Individual-level chemerin–PWV discordance map. Individual distribution of chemerin and pulse wave velocity residuals in the analytical sample. Each point represents one subject. Chemerin residuals were calculated as observed minus predicted chemerin values from a model including age, sex, body mass index, and estimated glomerular filtration rate. PWV residuals were calculated as observed minus predicted PWV values from a model including age, sex, body mass index, mean arterial pressure, resting heart rate, current smoking status, and eGFR. Dashed lines indicate the tertile cut-off points of each residual distribution and define nine chemerin–PWV phenotype cells. Orange points indicate participants with higher-than-expected chemerin and lower-than-expected PWV; green points indicate participants with higher-than-expected values of both measures; and blue points represent the remaining residual patterns. Abbreviation: PWV, pulse wave velocity.

• 

  Figure 4. Three-by-three classification of chemerin–PWV residual patterns. Three-by-three matrix obtained by independently dividing chemerin and PWV residuals into tertiles. Rows represent tertiles of chemerin residuals, and columns represent tertiles of PWV residuals. Cell values indicate the number of participants assigned to each residual pattern. The bottom-left cell represents participants with higher-than-expected chemerin and lower-than-expected PWV (n = 132), whereas the bottom-right cell represents participants with higher-than-expected values of both chemerin and PWV (n = 151). The middle tertiles were retained as intermediate categories to reduce the classification of participants with residual values close to the distributional cut-off points as extreme phenotypes. Abbreviation: PWV, pulse wave velocity.

• 

  Figure 5. Factors associated with the higher-than-expected chemerin/lower-than-expected PWV phenotype. Forest plot showing odds ratios and 95% confidence intervals for factors potentially associated with membership in the higher-than-expected chemerin/lower-than-expected PWV rather than the higher-than-expected chemerin/higher-than-expected PWV group. Each candidate factor was evaluated in a separate logistic regression model adjusted for age, sex, body mass index, mean arterial pressure, resting heart rate, current smoking status, and estimated glomerular filtration rate. Age and mean arterial pressure were modeled using centered natural cubic splines. Continuous candidate variables were standardized, and odds ratios are expressed per one-standard-deviation increase. The vertical reference line indicates an odds ratio of 1. None of the evaluated factors remained statistically significant after correction for multiple comparisons using the Benjamini–Hochberg false-discovery-rate procedure. Abbreviations: CI, confidence interval; FDR, false discovery rate; OR, odds ratio; PWV, pulse wave velocity.