Journal of Rehabilitation Medicine 51-7 | Page 59

528 P. G. McPhee et al. all 53 individuals for follow-up assessments; however, our sample size calculation confirmed that an n of 28 was reasonable to investigate the longitudinal effects in this study design. GMFCS distribution was as follows: I = 6; II = 7; III = 5; IV = 7; V = 3. The mean time interval between baseline and follow-up assessments was 4.0 ± 1.2 years. Changes in CVD risk factors between baseline and follow-up are shown in Table I. Between baseline and follow-up assessments at the group level, 64% of participants had an increase in waist circumference, absolute and relative FMD decreased in 85% of parti- cipants, 93% of participants experienced an increase in cIMT, carotid distensibility decreased in 59% of parti- cipants, and cfPWV increased in 70% of participants. Three participants at follow-up had cfPWV values that were above the clinical value (i.e. 10 m/s) deemed at risk for future cardiovascular events, whereas only one participant was at risk at baseline. With difference in age in months included as a covariate, repeated measures analysis of variance revealed statistically significant differences in mean absolute FMD (0.31 (SD 0.13) vs 0.22 (SD 0.08) mm, p = 0.045 95% CI 0.040, 0.151), relative FMD (9.9 (SD 4.7) vs 7.5 (SD 2.6) %, p = 0.049, 95% CI 0.464, 4.42), and cIMT (0.52 (SD 0.17) vs 0.67 (SD 0.33) mm, p = 0.041, 95% CI –0.242, –0.074) between baseline and follow-up as- sessments, respectively. As a result of the varying time intervals between the 2 sets of assessments between participants, changes in traditional and non-traditional risk factors were divided by time for each participant and reported as Table II. Multiple linear regression analyses cIMT, n  = 27 GMFCS grouping Age (baseline) Waist circumference, n  = 28 GMFCS grouping Age (baseline) SBP, n  = 28 GMFCS grouping Age (baseline) Distensibility, n  = 27 GMFCS grouping Age (baseline) cfPWV, n  = 23 GMFCS grouping Age (baseline) Absolute FMD, n  = 27 GMFCS grouping Age (baseline) Relative FMD, n  = 27 GMFCS grouping Age (baseline) Coef. p-value 95% CI –0.009 0.001 0.507 0.031 –0.035 to 0.018 3.4E–4 to 0.002 1.01 0.015 0.231 0.659 –0.690 to 2.71 –0.054 to 0.0.84 0.292 0.127 0.825 0.066 –2.41 to 2.99 –0.011 to 0.220 2.5E–4 1.3E–5 0.169 0.086 –1.1E–4 to 6.2E–4 –1.9E–6 to 2.7E–5 –0.053 0.009 0.724 0.119 –0.360 to 0.255 –0.003 to 0.022 –0.011 1.6E–4 0.523 0.784 –0.047 to 0.025 –0.001 to 0.001 –0.461 0.006 0.486 0.796 –1.80 to 0.882 –0.039 to 0.050 95% CI: 95% confidence interval; SBP: systolic blood pressure; FMD: flow mediated dilation; cIMT: carotid artery intima media thickness; GMFCS: Gross Motor Function Classification System; cfPWV: carotid-femoral pulse wave velocity. www.medicaljournals.se/jrm rates of change for the regression analyses. Multiple linear regression analysis for rate of change in cIMT revealed an R-squared of 0.294, p = 0.015, with age at baseline being a significant predictor of change in cIMT. From Cooks distance calculation, participant 19 (age = 58 years; GMFCS III; change in cIMT = 1.07 mm) was identified as an outlier with a residual of 0.14. The regression was performed without this data point, resulting in an R-squared of 0.261, p = 0.031 and again with age at baseline being a significant predictor of change in cIMT (Table II). Age at baseline and/or GMFCS grouping were not significant predictors of rates of change for waist circumference, SBP, carotid artery distensibility, cfPWV, and both absolute and relative FMD (Table II). DISCUSSION The objective of this study was to examine the longi- tudinal changes in traditional and non-traditional risk factors for CVD in a cohort of individuals with CP, in order to better understand the development of CVD in this population. An important finding was that, while risk factors for CVD increased in at least 50% of parti- cipants with CP relative to baseline over approximately a 4-year time period, some non-traditional indices seemed to have higher sensitivity for detecting signi- ficant changes over time. Specifically, after controlling for the varying time intervals between baseline and follow-up assessments among participants, significant changes were apparent only for the non-traditional risk factors absolute FMD, relative FMD, and cIMT. Importantly, this informs us that non-traditional mea- sures may detect changes in CVD risk in individuals with CP when not revealed by tracking traditional risk factors, which is consistent with the pathophysiology of the cardiovascular system in general; functional impairment of the arterial wall may occur at an early stage of the atherosclerotic process, before clinical symptoms of CVD are present (17). Concerning traditional risk factors for CVD, obesity and hypertension have shown a strong association with the non-traditional risk factors arterial stiffness and endothelial dysfunction. In fact, some non-traditional risk factors are considered negative prognostic factors of hypertension (18, 19). The increases in waist circum- ference and significant changes in FMD and cIMT within this cohort of individuals with CP underscore the importance of monitoring and managing CVD risk in this population to prevent the development of hypertension. As this was the first longitudinal cohort study to in- vestigate changes in non-traditional CVD risk variables in individuals with CP, it was important to understand