Background
As a consequence of the global obesity epidemic, prevalence rates of obesity-related co-morbidities such as elevated blood pressure (BP) in children are also increasing [
1]. Hypertension prevalences up to almost 25% are found in overweight children and adolescents [
2‐
4]. Although the cardiovascular sequelae of hypertension are clinically obvious in adulthood, the consequences of high BP in children and adolescents are usually less clear on first sight. Hypertension in children and adolescents is associated with the development of early, often subclinical, hypertensive target-organ damage (TOD) including increased carotid intima-media thickness, left ventricular hypertrophy, insulin resistance, and renal damage [
5‐
8]. In addition, numerous studies have shown that high BP in childhood increases the risk for adult hypertension and metabolic syndrome [
9‐
12].
The diagnosis of elevated BP and hypertension depends on an accurate BP measurement, which can present a challenge to the clinician. Ambulatory 24-hour blood pressure measurement (ABPM) is strongly recommended for the diagnosis and management of hypertension [
13‐
16]. ABPM allows a more representative observation of BP thoughout day and night compared to office blood pressure (OBP) measurements as well as assessment of the circadian and even ultradian BP variability. ABPM is useful to detect white-coat hypertension, masked hypertension, and nocturnal hypertension [
17‐
19]. White-coat hypertension and masked hypertension are known to be more prevalent in obese compared to lean pediatric populations [
3]. Furthermore, ABPM has been shown in children to be more predictive of end-organ damage [
20]. However, healthcare workers may hesitate to perform ABPM, for example because they do not want to burden the patient when they think it is not necessary, especially in the childhood clinic. Furthermore, an adequate APBM could be difficult to obtain.
This study was conducted to evaluate ABPM patterns in a population of overweight and obese children and adolescents referred to our pediatric outpatient clinic, and to compare ABPM patterns with regular OBP measurements, with the aim to show the additional value of ABPM in this population. Our hypothesis is that the prevalence of abnormal ABPM patterns, including white-coat hypertension and masked hypertension, is substantial in childhood obesity. Furthermore, we expect a high prevalence of abnormal circadian variability in this population.
Results
Our 82 participants were aged 4–17 years, and 39% of them were boys (Table
2). Ten participants were classified as overweight, the remaining 72 (88%) were obese. 60% of the participants presented with at least one obesity-related comorbidity. No participant was treated with antihypertensive medication at the time of the ABPM. The average amount of 24-hour ABPM readings per patient was 39 (standard deviation 6).
Table 2
Participant characteristics of the total study population, and for the ABPM categories separately
Male | 32 (39.0) | | 20 (44.4) | 3 (75.0) | 6 (27.3) | 1 (33.3) | 2 (25.0) |
Age | 11.8 (8.8, 14.6) | | 11.5 (8.7, 14.6) | 8.4 (5.9, 11.5) | 12.9 (9.0, 14.9) | 10.8 (6.8, NA) | 9.9 (9.1, 16.3) |
Ethnicity‡ | | | | | | | |
| Dutch / Western immigrant | 66 (81.5) | | 37 (82.2) | 3 (75.0) | 17 (77.2) | 3 (100.0) | 6 (75.0) |
| Non-western immigrant | 16 (19.5) | | 8 (17.8) | 1 (25.0) | 5 (22.7) | 0 (0.0) | 2 (25.0) |
BMI Z-score | 3.3 (2.8, 3.6) | | 3.1 (2.8, 3.5) | 3.1 (2.8, 3.6) | 3.5 (2.9, 3.8) | 2.7 (2.4, NA) | 3.8 (3.1, 4.3)b |
Glucose metabolism | | | | | | | |
| Prediabetes* | 8 (10.4) | | 4 (9.5) | 0 (0.0) | 3 (14.3) | 0 (0.0) | 1 (12.5) |
| Elevated HOMA-IR† | 33 (44.0) | | 16 (39.0) | 0 (0.0) | 9 (45.0) | 0 (0.0) | 8 (100.0)a |
Lipid profile | | | | | | | |
| Elevated total cholesterol* | 3 (3.9) | | 2 (4.8) | 0 (0.0) | 1 (4.8) | 0 (0.0) | 0 (0.0) |
| Elevated LDL-C* | 1 (1.2) | | 1 (2.4) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
| HDL-C below cut-off* | 10 (13.0) | | 4 (9.5) | 1 (25.0) | 3 (14.3) | 0 (0.0) | 2 (25.0) |
| Elevated triglycerides* | 8 (10.4) | | 3 (7.1) | 0 (0.0) | 3 (14.3) | 0 (0.0) | 2 (25.0) |
MetS** | 11 (13.4) | | 2 (5.1) | 1 (25.0) | 6 (30.0)a | 0 (0.0) | 2 (28.6)b |
Based on OBP measurements 54.9% (95% confidence interval [CI] 44.1–65.2; n = 45) of the participants were normotensive, 19.5% had elevated BP, 19.5% classified as stage 1 hypertension, and 6.1% were classified as stage 2 hypertension (Table
3). 22% of the participants had an office systolic BP index ≥ 1.0, 9% an office diastolic BP index ≥ 1.0.
Table 3
Summary of blood pressure characteristics according to ABPM classification
OBP | | | | | | |
Normal BP | 45 (54.9) | 35 (77.8) | 0 (0.0) | 9 (40.9) | 1 (33.3) | 0 (0.0) |
Elevated BP | 16 (19.5) | 9 (20.0) | 0 (0.0) | 5 (22.7) | 2 (66.7) | 0 (0.0) |
Stage 1 HTN | 16 (19.5) | 1 (2.2) | 3 (75.0) | 7 (31.8) | 0 (0.0) | 5 (62.5) |
Stage 2 HTN | 5 (6.1) | 0 (0.0) | 1 (25.0) | 1 (4.5) | 0 (0.0) | 3 (37.5) |
24-hour SBP | | | | | | |
Median (IQR) | 109 (104, 116) | 106 (101, 112) | 106 (97, 113) | 111 (107, 119) | 123 (110, NA) | 124 (114, 130) |
BP index, median (IQR) | 0.88 (0.84, 0.93) | 0.86 (0.82, 0.88) | 0.87 (0.83, 0.92) | 0.92 (0.87, 0.95) | 0.96 (0.94, NA) | 1.02 (0.98, 1.05) |
BP index ≥ 1.0, n (%) | 7 (8.5) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (33.3) | 6 (75.0) |
BP load (%), median (IQR) | 8.8 (3.6, 25.4) | 6.1 (0.0, 14.2) | 4.2 (1.0, 13.0) | 16.2 (6.7, 27.9) | 28.6 (17.2, NA) | 58.1 (44.4, 72.9) |
BP load > 25%, n (%) | 20 (24.4) | 3 (6.7) | 0 (0.0) | 8 (36.4) | 2 (66.7) | 7 (87.5) |
Dipping, median (IQR) | 11 (7, 15) | 12.8 (8.5, 17.1) | 9.8 (7.5, 11.1) | 10.4 (4.7, 13.4) | 5.2 (4.9, NA) | 6.8 (0.8, 12.1) |
Dipping < 10%, n (%) | 33 (40.2) | 14 (31.1) | 2 (50.0) | 11 (50.0) | 2 (66.7) | 4 (50.0) |
24-hour DBP | | | | | | |
Median (IQR) | 65 (61, 69) | 63 (60, 66) | 62 (56, 66) | 67 (64, 71) | 73 (64, NA) | 76 (69, 81) |
BP index, median (IQR) | 0.86 (0.81, 0.91) | 0.82 (0.79, 0.87) | 0.83 (0.74, 0.85) | 0.89 (0.84, 0.93) | 0.96 (0.88, NA) | 0.99 (0.91, 1.07) |
BP index ≥ 1.0, n (%) | 4 (4.9) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (33.3) | 3 (37.5) |
BP load (%), median (IQR) | 15.5 (6.3, 25.7) | 11.9 (4.1, 20.5) | 8.8 (3.8, 21.3) | 17.7 (13.1, 30.6) | 37.9 (16.7, NA) | 53.6 (22.3, 73.6) |
BP load > 25%, n (%) | 22 (26.8) | 6 (13.3) | 0 (0.0) | 9 (40.9) | 2 (66.7) | 5 (62.5) |
Dipping, median (IQR) | 19 (14, 22) | 19.4 (14.3, 24.5) | 16.2 (12.3, 23.0) | 17.9 (14.7, 20.7) | 19.4 (0.4, NA) | 14.6 (7.0, 24.0) |
Dipping < 10%, n (%) | 9 (11.0) | 3 (6.7) | 0 (0.0) | 3 (13.6) | 1 (33.3) | 2 (25.0) |
Non-dipper, n (%) | 34 (41.5) | 14 (31.1) | 2 (50.0) | 12 (54.5) | 2 (66.7) | 4 (50.0) |
The ABPM patterns also showed that 54.9% (95% CI 44.1–65.2) of participants had normal BP. However, this group differed from the OBP normotensive group (Table
3). Ten of the 45 participants (22.2%) with normal OBP turned out to have either elevated BP (n = 9) or masked hypertension (n = 1) on ABPM (Table
3). Of the 37 participants with abnormal OBP, ten had normal BP on APBM (27.0%). Eight of the 21 participants (38%) with OBP hypertension were confirmed to have ambulatory hypertension. The others were diagnosed with elevated BP (n = 8) or white-coat hypertension (n = 4). Of the 16 participants with elevated BP in the outpatient clinic, 56% had normal ABPM results, 13% had masked hypertension, and in 31% of the participants elevated BP was confirmed with ABPM. The correlation coefficient for the relationship between indexed office systolic BP (as a proxy of hypertensive status based on OBP) and indexed mean daytime systolic BP (hypertensive status based on ABPM) was 0.39.
Using ABPM, a BP load > 25% was found in 24.4–26.8% of the participants during the whole 24-hour period. During night-time more often a BP load > 25% was detected than during daytime. 24% of all cases (n = 20) showed an isolated night-time BP load > 25% with normal daytime ABPM. Up to 40% of the participants lacked physiologic nocturnal systolic BP dipping.
No significant differences in terms of age and gender were observed between the different ABPM categories (Table
2). Participants with ambulatory hypertension had a significantly higher BMI Z-score (3.8, interquartile range [IQR] 3.1–4.3) compared to the normal BP group (BMI Z-score 3.1, IQR 2.8–3.5). Prediabetes was detected in 8 participants (11.1%); 50% of them had elevated BP or ambulatory hypertension.
The prevalence of metabolic syndrome was significantly higher in participants with elevated BP and ambulatory hypertension than in participants with normal ABPM. Of note, metabolic syndrome was significantly more prevalent in non-dippers when compared to dippers (29.6% versus 6.7%, respectively, p = 0.009). Increasing severity of obesity was not associated with nocturnal non-dipping.
Discussion
Our study confirms the high prevalence of abnormal BP in obese children and adolescents. It also underscores the unreliability of OBP measurement and the need for BP monitoring by APBM.
Using ABPM, 8 of the 82 participants (9.8%) were classified as ambulatory hypertension, of which 75% had severe ambulatory hypertension. Elevated BP was present in 22 of the 82 participants (26.8%). In literature, hypertension prevalence ranges from 3.8 to 24.8% in youth with overweight and obesity [
23]. Prevalences of elevated BP up to around 15% are reported in unselected children, and to 20–30% in childhood obesity [
27‐
31]. Elevated BP, or former ‘prehypertension’, has shown to be associated with cardiovascular TOD in adolescents and young adults and may be a risk factor of progressing to sustained hypertension [
31‐
37].
Three participants (3.7%) in our cohort were diagnosed with masked hypertension and four (4.9%) with white-coat hypertension. In literature, masked hypertension prevalence ranges from 7.6% in unselected children [
8], to 32.3% in obese children with a non-dipping pattern [
38]. White-coat hypertension prevalence ranges from 0.6% in 9–10 year old students [
39], to 59% in a group of children referred for persistently elevated casual BP [
40]. The divergence observed in the prevalence of masked hypertension and white-coat hypertension is likely caused by measurements in different study populations using different diagnostic criteria [
40], and by the choice of the upper limits of normalcy [
40,
41]. In our study 38.1% of the subjects with stage 1 or 2 hypertension based on OBP measurement demonstrated less severe elevation on ABPM and were classified as elevated BP, also suggesting a white coat phenomenon.
The clinical significance of masked hypertension in children lies in the potentially increased risk for TOD and future cardiovascular events [
8,
42,
43]. The impact of white-coat hypertension in children is far less clear [
43]. Although white-coat hypertension in adulthood has been associated with cardiovascular morbidity and mortality and progression to sustained hypertension [
44,
45], the published cardiovascular events incidences and all-cause mortality relative risks are only slightly higher compared to normotensive people and significantly below the risks associated with sustained hypertension [
46,
47].
In our study, more than 20% of the participants with normal OBP turned out to have either elevated BP or masked hypertension on ABPM. These patients would have been missed if classified by OBP. This may convince hesitating healthcare workers to incorporate ABPM in their standard care for overweight and obese children. Discrepancies between OBP and ABPM have been described before in different pediatric populations [
3,
48,
49]. Considering (future) cardiovascular risks in patients with elevated BP or masked hypertension, this underscores the importance of performing ABPM in overweight children, although in some children it could be a challenge to obtain an adequate ABPM.
A high prevalence of abnormal circadian variation was present in our study. Nocturnal hypertension has shown to have significant prognostic implications [
20]. In childhood and adolescence, literature on the association between nocturnal dipping and morbidity is scarce, although some studies show that non-dipping may be associated with insulin resistance [
50,
51]. In adults, a non-dipping status is associated with cardiac structural alterations and a higher risk of CVD events [
52]. Although the suggested scheme for staging of ambulatory BP levels of Flynn et al. incorporates night-time mean BP and BP load, dipping status is not included. As such, dipping status represents an entity that needs separate attention. The high prevalence of abnormal circadian variation in this study, with the associated potential risk for TOD and CVD, confirms the importance of performing ABPM in overweight children in order to detect nocturnal hypertension or a decreased or absent dipping status.
No significant differences in terms of age, gender and ethnicity were observed between the different ABPM categories, perhaps due to small sample size. Previous studies noted that ambulatory BP is affected by sex and ethnicity [
53,
54]. A recently published systematic review showed that when age was dichotomized according to puberty, elevated BP and hypertension were more prevalent in older children. This association was not consistent when using age as a continuous variable [
53].
A higher BMI is an independent risk factor of high BP in children [
53]. In our study participants with ambulatory hypertension had a significantly higher BMI Z-score compared to the normal BP group.
Increased HOMA-IR was present in 39% of the subjects with normal BP in this study. All participants with ambulatory hypertension presented with an elevated HOMA-IR, and almost half of the participants with elevated BP. Moreover, a significantly higher prevalence of metabolic syndrome was detected in children with elevated BP and ambulatory hypertension, as well as in non-dippers, indicating the clustering of other CVD risk factors in overweight subjects with high BP when compared to overweight children with normal BP.
To our knowledge the present study is the first using the BP reference values as presented in the updated Clinical Practice Guideline [
23] in an overweight childhood population, to compare with ABPM results. The main strength of our study is the large number of available ABPMs. Despite this, low patient numbers in the different ABPM classification groups made it difficult to study factors associated with the different ABPM diagnoses. A few other limitations need to be addressed. First, despite the widespread use of the 2014 AHA Scientific Statement values in the interpretation of ABPMs, several limitations has been recognized, i.e. with regard to generalizability [
20]. Robust, universally applicable normative ABPM data in children and adolescents are lacking. Second, by using the current ABPM classification scheme some subjects remain unclassifiable, limiting the comparability between studies due to divergent solutions with regard to the individual classification of these patients. Third, normative data are based on auscultatory measurements, which may provide different values than measurements obtained by using oscillometric devices or ABPM, as obtaining BP by oscillometry could result in an overestimation of BP values [
55].
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.