Family History of Premature Cardiovascular Disease
Family History of Premature Cardiovascular Disease
Baseline characteristics are shown in Table 1 for the overall population and in Supplementary material online, Table S1 for the propensity score-matched group. There were 1408 and 9379 individuals, respectively, with (P-FH) and without parental history (P-FH) of premature CVD, while 1864 individuals (FDR-FH) had at least one first-degree relative with premature CVD and 7571 individuals (FDR-FH) had no first-degree relative (both parents negative) with premature CVD. More than half (51.7%) of the study participants were female, the mean BMI was 27.1 ± 5.2 kg/m, the mean baseline SBP was 169.6 ± 29.6 mmHg, the mean baseline DBP was 100.1 ± 15.3 mmHg, and 23% had at least stage 3 CKD. At first clinic visit, compared with P-FH, P-FH were younger, had lower BP, lower cholesterol, and less prevalent CKD. As the ascertainment period was very long, the baseline demographic profile of patients presenting in different time-periods (epochs) is presented in Figure 1 showing no substantial between-group differences across epochs. The total time at risk was 193 756 person-years with a median survival time of 29.2 years. The incidence rates were 20.5 [95% confidence interval (CI) 19.9–21.1], 13.8 (95% CI 13.3–14.3), and 8.9 (95% CI 8.5–9.4) per 1000 person-years of follow-up for all-cause, CVD, and non-CVD mortality, respectively.
(Enlarge Image)
Figure 1.
Baseline demographic profile of patients at first presentation during different periods (epochs) over the ascertainment period.
Unadjusted and adjusted results of Cox survival analyses are presented in Table 2. In the univariate analysis, P-FH was associated with lower mortality (Model 1), but this was reversed after inclusion of age, without any major attenuation of hazard ratios between models adjusted for age and sex (Model 3) and the fully adjusted model (Model 4). In the fully adjusted model, FDR-FH and P-FH were associated with an 11–12 and 14–20% increased risk of all-cause and CV death, respectively. There was no association with non-CV or stroke mortality. The results were similar in the propensity-matched analysis.
In the GEE analyses, the overall average annual reduction in SBP in the first 5 years was −3.4 (SE 0.06) mmHg and DBP −1.9 (0.03) mmHg after adjustment for age, sex, BMI, smoking, alcohol use, CKD, and prevalent CVD. Family history had no effect on 5-year changes in BP, eGFR, or cholesterol (Figure 2 and Supplementary material online, Table S2).
(Enlarge Image)
Figure 2.
Longitudinal trends in blood pressure, renal function, and serum cholesterol over the first 5 years of follow-up. β-Coefficients with 95% confidence intervals are obtained from generalized estimating equations and show the change over 5 years after adjustment for conventional covariates.
Reclassification analyses (Table 3) showed that the inclusion of family history of CVD did not improve CV mortality risk discrimination over and above conventional CV risk factors, but reclassification improved by 1.9% using risk cut-offs of 0–10%, 10–20%, and >20%.
Dispensed prescription data were available on 1869 patients for at least 5 years between 2004 and 2013 with 5.8 ± 2.9 years interval between their first clinic visit and the start of the prescription review period. FDR-FH and P-FH were younger than those with no family history. All patients were on antihypertensive treatment during this period, the average refill rate was 5.8 ± 1.9, and 51% of patients had a coefficient of variation <1 of the inter-refill interval. There was no significant difference in the number of patients on lipid-lowering (93%) or antiplatelet therapy (47–50%) based on parental or family history of CVD. There was no significant difference in the annual rate of refilled prescriptions or coefficient of variation of inter-refill interval based on parental or family history of CVD (Table 4).
Results
Demographic and Clinical Characteristics of the Study Population
Baseline characteristics are shown in Table 1 for the overall population and in Supplementary material online, Table S1 for the propensity score-matched group. There were 1408 and 9379 individuals, respectively, with (P-FH) and without parental history (P-FH) of premature CVD, while 1864 individuals (FDR-FH) had at least one first-degree relative with premature CVD and 7571 individuals (FDR-FH) had no first-degree relative (both parents negative) with premature CVD. More than half (51.7%) of the study participants were female, the mean BMI was 27.1 ± 5.2 kg/m, the mean baseline SBP was 169.6 ± 29.6 mmHg, the mean baseline DBP was 100.1 ± 15.3 mmHg, and 23% had at least stage 3 CKD. At first clinic visit, compared with P-FH, P-FH were younger, had lower BP, lower cholesterol, and less prevalent CKD. As the ascertainment period was very long, the baseline demographic profile of patients presenting in different time-periods (epochs) is presented in Figure 1 showing no substantial between-group differences across epochs. The total time at risk was 193 756 person-years with a median survival time of 29.2 years. The incidence rates were 20.5 [95% confidence interval (CI) 19.9–21.1], 13.8 (95% CI 13.3–14.3), and 8.9 (95% CI 8.5–9.4) per 1000 person-years of follow-up for all-cause, CVD, and non-CVD mortality, respectively.
(Enlarge Image)
Figure 1.
Baseline demographic profile of patients at first presentation during different periods (epochs) over the ascertainment period.
Association Between Family History of CVD and Mortality
Unadjusted and adjusted results of Cox survival analyses are presented in Table 2. In the univariate analysis, P-FH was associated with lower mortality (Model 1), but this was reversed after inclusion of age, without any major attenuation of hazard ratios between models adjusted for age and sex (Model 3) and the fully adjusted model (Model 4). In the fully adjusted model, FDR-FH and P-FH were associated with an 11–12 and 14–20% increased risk of all-cause and CV death, respectively. There was no association with non-CV or stroke mortality. The results were similar in the propensity-matched analysis.
Association of Family History of Cardiovascular Disease With Longitudinal Blood Pressure and Estimated Glomerular Filtration Rate
In the GEE analyses, the overall average annual reduction in SBP in the first 5 years was −3.4 (SE 0.06) mmHg and DBP −1.9 (0.03) mmHg after adjustment for age, sex, BMI, smoking, alcohol use, CKD, and prevalent CVD. Family history had no effect on 5-year changes in BP, eGFR, or cholesterol (Figure 2 and Supplementary material online, Table S2).
(Enlarge Image)
Figure 2.
Longitudinal trends in blood pressure, renal function, and serum cholesterol over the first 5 years of follow-up. β-Coefficients with 95% confidence intervals are obtained from generalized estimating equations and show the change over 5 years after adjustment for conventional covariates.
Discriminatory Power of Family History of Cardiovascular Disease in Predicting Cardiovascular Mortality
Reclassification analyses (Table 3) showed that the inclusion of family history of CVD did not improve CV mortality risk discrimination over and above conventional CV risk factors, but reclassification improved by 1.9% using risk cut-offs of 0–10%, 10–20%, and >20%.
Drug Adherence
Dispensed prescription data were available on 1869 patients for at least 5 years between 2004 and 2013 with 5.8 ± 2.9 years interval between their first clinic visit and the start of the prescription review period. FDR-FH and P-FH were younger than those with no family history. All patients were on antihypertensive treatment during this period, the average refill rate was 5.8 ± 1.9, and 51% of patients had a coefficient of variation <1 of the inter-refill interval. There was no significant difference in the number of patients on lipid-lowering (93%) or antiplatelet therapy (47–50%) based on parental or family history of CVD. There was no significant difference in the annual rate of refilled prescriptions or coefficient of variation of inter-refill interval based on parental or family history of CVD (Table 4).
Source...