Arterial Hypertension Determined by Ambulatory Blood Pressure Profiles: Contribution to microalbuminuria risk in a multicenter investigation in 2,105 children and adolescents with type 1 diabetes

A total of 31,278 patients with type 1 diabetes under 18 years of age were consecutively registered at 195 centers for pediatrics and internal medicine using the national quality initiative DPV software (Diabetes Software for Prospective Documentation). The data were generated locally, documented, and transmitted in anonymous form to the University of Ulm for central evaluation and analysis, as described previously (1,12). A total of 5,982 24-h blood pressure recordings were obtained from January 1994 until October 2006.

Pediatric patients between 5 and 18 years of age and the most recent ambulatory blood pressure recording for each patient were included in this cross-sectional investigation. Patients with antihypertensive treatment were excluded (9.2%), and the blood pressure profiles of 2,105 children and adolescents (1,101 boys and 1,004 girls) were analyzed. Altogether, 74% of the recordings were generated in six major centers.

Normalized reference values were obtained from cross-sectional ABPM data in 949 healthy children and adolescents (464 boys and 485 girls). Their age ranged from 5 to 18 years (mean 11.1) and height from 101 to 198 cm (mean 150). Only healthy children with no history of disease affecting blood pressure and without current antihypertensive medication or other blood pressure–affecting drugs were included in this investigation, performed by the German Working Group on Pediatric Hypertension (7). Using the least median squares (LMS) method, sex-specific L M, and S reference values were calculated for 24-h and daytime and nighttime mean values of SBP DBP, and mean arterial pressure (MAP) relative to age and height. These reference values enable the calculation of SDS (SD score) values in individual patients for each of the above-mentioned AMBP recordings (see below).

A1C values were determined in each center and standardized according to the Diabetes Control and Complication Trial reference range of 4.05–6.05% (1). BMIs were compared by SDS values based on a cohort of 34,422 healthy German children (17,147 boys and 17,275 girls) (13).

Ambulatory 24-h blood pressure monitoring was performed by oscillometry with adapted cuff size. Daytime blood pressure was measured every 20 min between 0800 and 2000 h, and nighttime blood pressure was recorded once per hour between 0000 and 0600 h. Each patient recorded his/her daily routine; in case of changes in active and sleeping times, blood pressure readings were individually adjusted. Only ABPM recordings with at least 75% reliable readings were further analyzed. Mean SBP DBP, and MAP were calculated separately for daytime and nighttime, as well as the nocturnal reduction of SBP and DBP (dipping). Age and sex dependency of blood pressure were corrected for using SDS values for SBP DBP, and MAP.

Because the use of pediatric ABPM reference values is comprised by the non-Gaussian distribution of 24-h blood pressure in children, we used the LMS method to calculate appropriate SDS values for ABPM. The LMS method describes the distribution of a measurement Y by its median (M), the coefficient of variation (S), and a measure of skewness (L) required to transform the data to normality. The reference values of L M, and S can be used to calculate individual LMS-transformed SDS (LMS-SDS) by the following equation (7):

where Y is the child's individual blood pressure value, and L(t), M(t), and S(t) represent sex-specific reference values of L M, and S interpolated for the child's height.

The individual LMS-SDS values were compared with those of the published reference population of healthy German children (7,14). Blood pressure profiles (SBP DBP, and MAP) were considered pathological when the LDS-SDS exceeded 1.65, corresponding with the 95th percentile (7).

The nocturnal reduction of blood pressure (dipping) was calculated as (daytime BP − nighttime BP)/daytime BP, where BP is blood pressure.

The generally accepted definition of normal systolic dipping is a nocturnal SBP reduction of >10% (15), whereas pathological diastolic dipping is controversially defined as a nocturnal reduction of <10% (16) or <20% (15). Therefore, prevalence will be given for both definitions.

Persistent microalbuminuria was defined according to the guidelines of the International Society for Pediatric and Adolescent Diabetology as a minimum of two positive out of three consecutive urine specimens at least 4 weeks apart (17) with an albumin excretion rate of 20–200 μg/min in timed overnight urine collections or 30–300 mg/24h in 24-h urine collections and an albumin-to-creatinine ratio of 2.5–25 mg/mmol or 30–300 mg/g in the morning spot urine. Each of the participating centers decided independently which method to use. Blood pressure profiles were compared with the urinary albumin excretion rates taken within ±3 months of the ABPM recording.

Statistical analysis was performed using SAS version 9.1 (SAS Institute, Cary, NC). Group-specific differences were compared using parametric testing (t test) after testing for Gaussian distribution (Kolmogorov-Smirnov) and otherwise by nonparameteric Wilcoxon's test.

Age (years), diabetes duration (years), sex, A1C (%), BMI-SDS, and insulin dose (insulin units per kilogram body weight per day) were compared with blood pressure LMS-SDS as independent variables by multiple linear regression analysis. Potential factors contributing to microalbuminuria were studied by stepwise multiple logistic regression analysis. We did not correct for the multiple tests. Therefore, all reported P values are nominal. Unless otherwise stated, data are presented as means ± SD. P < 0.05 was considered significant and P < 0.01 as highly significant.

The mean age of the children included in this investigation was 14.05 ± 2.95 years, mean diabetes duration 5.15 ± 4.02 years, mean BMI-SDS 0.49 ± 0.90, and average A1C 8.0 ± 1.8%. The subjects required an average insulin dosage of 0.83 ± 0.28 units · kg body wt⁻¹ · day⁻¹.

The boys were significantly older than the girls (14.2 ± 0.09 vs. 13.9 ± 0.1 years; P = 0.021, Wilcoxon's test) and had a shorter diabetes duration (4.9 ± 0.12 vs. 5.4 ± 0.13 years; P = 0.001, Wilcoxon's test), and the girls had significantly higher BMI-SDS than the boys (0.57 ± 0.03 vs. 0.42 ± 0.03; P < 0.0001, Wilcoxon's test). A1C (P = 0.06) and insulin dosage (P = 0.56) did not differ between the sexes.

Daytime mean LMS-SDS in diabetic patients was higher by 0.06 ± 0.83 for SBP and by 0.11 ± 0.97 for MAP, whereas LMS-SDS for DBP was reduced by 0.12 ± 1.08, compared with the reference population (P < 0.0001 for all three variables) (Fig. 1). More pronounced results were found for nocturnal blood pressure: mean LMS-SDS was increased by 0.51 ± 1.20 for SBP, 0.58 ± 1.10 for DBP, and 0.80 ± 1.20 for MAP in the study population (P < 0.0001 for all three variables) (Fig. 1).

Mean dipping was significantly reduced in the diabetic children compared with the reference population (respectively 10.0 ± 5.7 vs. 13.0 ± 6.0% for systolic dipping and 16.8 ± 8.1 vs. 23.0 ± 9.0% for diastolic dipping; P < 0.0001) (Fig. 2).

The prevalence for pathological blood pressure at daytime was 5.1% for SBP, 4.3% for DBP, and 5.7% for MAP and during the night was 14.6% for SBP, 14.7% for DBP, and 19.0% for MAP (Fig. 1). Pathological dipping was found in 49.1% for SBP and in 17.5% for DBP using a cut off of 10% and in 64.9% for DBP using a cut off of 20%.

Using multiple regression analysis, insulin dosage, female sex, BMI-SDS, A1C, and diabetes duration were significantly associated with increased blood pressure (Table 1). Increased SBP was strongly related to diabetes duration, female sex, and BMI-SDS, both for the diurnal and nocturnal values. Age and A1C were not associated with SBP. DBP was strongly related to diabetes duration, female sex, and A1C. Diurnal DBP was additionally associated with BMI-SDS and nocturnal DBP with age and insulin dosage. Diabetes duration, female sex, BMI-SDS, and insulin dose were significantly correlated with MAP, age, and A1C, however, only with the nocturnal MAP. Nocturnal blood pressure reduction was linked to age and A1C and diastolic dipping additionally to female sex and insulin dosage.

Twenty four–hour blood pressure profiles and data on urinary albumin excretion were available for 1,670 patients. A total of 101 patients (6.1%) had persistent microalbuminuria, which was significantly associated with nocturnal DBP (P < 0.0001) and impaired diastolic dipping (P < 0.01). SBP MAP, A1C, age, diabetes duration, sex, BMI-SDS, and insulin dose were not related to microalbuminuria.

In this bi-national multicenter investigation, diabetic children showed an abnormal blood pressure profile, particularly affecting nocturnal blood pressure. Mean LMS-SDS values of blood pressure in diabetic children differed from those of the control population by +0.51 to +0.80.

The prevalence of arterial hypertension in the diabetic population is 1.5 to 3 times higher than that in nondiabetic age-matched groups (18); ∼75% of adult diabetic patients have blood pressure >140/90 mmHg (19). Age-related changes in blood pressure regulation were observed in both diabetic and nondiabetic subjects, but the changes seem to occur 15–20 years earlier in type 1 diabetic patients compared with control subjects, suggesting accelerated vascular ageing (20). Compared with the nondiabetic population, type 1 diabetes is associated with a deleterious blood pressure pattern in adult patients even in the absence of diabetic kidney disease (20). According to the Strong Heart Study (21), prehypertension is more prevalent in diabetic patients and the risk for cardiovascular disease is increased 2.9-fold in subjects with type 1 diabetes alone and 3.7-fold in diabetic patients with prehypertension. Our data demonstrate that this fatal development already starts in children and adolescents at an early stage of type 1 diabetes.

In particular, nocturnal arterial hypertension and nondipping seems to be associated with cardiovascular disease in the general population. Systemic arterial vascular tone is increased in patients with essential hypertension during the night compared with normotensive control subjects. This increased vascular tone might contribute to the well-known changes of arterial structure in essential hypertension and eventually lead to cardiovascular disease (22).

Nondippers experience a greater incidence of stroke and myocardial infarction than people with normal dipping (23). Because our diabetic children showed impaired nocturnal blood pressure regulation, they might have an increased risk for macrovascular complications even after short diabetes duration (24). Therefore, early detection of alterations in blood pressure regulation is crucial for adequate diabetes management and sufficient antihypertensive therapy. Performing AMBP in diabetic patients might provide valuable information on early alterations in blood pressure regulation (25).

BMI in particular influenced SBP but had no effect on blood pressure dipping, the most prominent sign of altered blood pressure regulation in these diabetic patients. Performing 24-h AMBP, Wühl et al. (7) demonstrated a significant association between SBP and BMI-SDS, accounting for ∼10.7% of the total variability in SBP in healthy children and adolescents. The contribution of increased BMI on SBP variability in our patients was less pronounced, with a maximum of 5.6%. BMI was significantly associated with the SBP-SDS in children with type 1 diabetes participating in the Oxford Regional Prospective Study (26).

Elevated BMI is associated with peripheral insulin resistance, leading to higher insulin requirements similar to a type 2 diabetes–like metabolic situation. Metabolic syndrome is a frequent finding in type 1 diabetes and increases with inadequate glycemic control (27). Even short-term hyperglycaemia of 48 h may disturb vascular function; therefore, prolonged and repeated episodes of hyperglycemia could lead to permanent vascular dysfunction (28). Both, hyperglycemia and hyperinsulinemia stimulate different pathophysiological pathways, which result in altered blood pressure regulation and endothelial dysfunction.

In our diabetic children, female sex predisposed to early blood pressure alterations, probably due to an increased weight gain and a higher risk for insulin resistance during puberty. Adolescent nondiabetic girls are less insulin sensitive than boys but compensate for decreased sensitivity by increasing their insulin secretion (29). Thus, an elevated BMI is probably a major cause for higher insulin resistance in pubertal girls with type 1 diabetes (30). Increased insulin resistance resulting in hyperinsulinemia and hyperglycemia and elevated BMI substantially contributes to the increased rate of arterial hypertension in diabetic girls.

Alterations in nocturnal blood pressure regulation contribute to microalbuminuria and might enhance the development of diabetic nephropathy even in children with type 1 diabetes. Development of microalbuminuria is linked to insufficient blood pressure control and a progressive increment of glucose values in nondiabetic with mild hypertension (31,32). The Oxford Regional Prospective Study confirmed the link between microalbuminuria and arterial hypertension in children followed from diagnosis of type 1 diabetes. In young adults with type 1 diabetes, an increase in SBP during nighttime preceded the development of microalbuminuria. The risk of microalbuminuria for diabetic subjects with a normal pattern of nocturnal blood pressure was 70% lower than that for patients with an abnormal pattern (33). In accordance, our cross-sectional data suggest that impairment of blood pressure regulation begins at nighttime with a higher prevalence than microalbuminuria. However, office blood pressure measurement did not rise before the onset of microalbuminuria (26,27).

Therefore, metabolic state, body weight, and insulin dosage need to be optimized at the onset of diabetes in order to avoid negative impact on blood pressure regulation and vasculature. Age, female sex, and diabetes duration also affected blood pressure regulation as nonmodifiable risk factors. However, these factors classify patients at a higher risk for arterial hypertension requiring special monitoring.

In conclusion, our data suggest that ABPM might be a valuable tool in monitoring pediatric patients with type 1 diabetes, thus enabling vascular-directed preventive intervention at the earliest possible time.

The DPV Science Initiative is supported by the German Federal Ministry of Health, Novo Nordisk Germany, the Dr. Bürger-Büsing Foundation, the German Diabetes Foundation, the German Research Foundation, and the National Action Forum Against Diabetes Mellitus. The sponsors were not involved in this particular investigation at any time.

The centers within the DPV Science Initiative contributed 24-h blood pressure profiles to this study, a list of the centers collaborating in the initiative can be found at http://dpv.mathematik.uni-ulm.de/.

The authors thank Silvia Rudloff, PhD, for proofreading the manuscript.