Association of urine phthalate metabolites, bisphenol A levels and serum electrolytes with 24-h blood pressure profile in adolescents

ABPM measurements were made from the patient’s non-dominant arm with the cuff of the appropriate size and the Spacelabs Monitor 90,207 device (Spacelabs Healthcare, Snoqualmie, Washington, USA). Measurements were made every 15–20 min during the waking period and 30 min during the sleep period; ABPM data were evaluated with a measurement time of at least 18 h without interruption; at least 20 measurements during wakefulness and at least 7 measurements during sleep; at least one measurement was made per hour, including the sleep period [36, 37]. SDS for systolic blood pressure, diastolic blood pressure and mean arterial pressure (MAP) were estimated with LMS values according to age, height, and gender [38].

Definitions of ABPM profiles including normal ABPM profile, ABPM-HT, WCHT, masked HT and dipping were given in Fig. 1. ABPM profiles were further evaluated for the presence of daytime (from 08.00 am to 08.00 pm), night-time (from midnight to 06:00 am), and sustained HT (Fig. 1) [38, 39].

Ultrasound examination for cIMT measurement was performed with an Affiniti 70G ultrasound system (Philips Medical Systems, Holland) using 5-18 MHz linear probe according to Mannheim consensus [40]. Five measurements were taken and the average was converted to SDS values using standard LMS values for gender and height [41].

Free BPA (fBPA) and total BPA (tBPA) levels were measured using the MyBioSource Human fBPA ELISA kit (Cat no: MBS047388, Southern California, SanDiego) and the Human tBPA ELISA Kit (Cat no: MBS109499) by the user manual.

Of phthalates, 14 metabolites [mono-n-butyl phthalate (MnBP), mono-benzyl phthalate (MBzP), monocarboxy isononyl phthalate (MCiNP), mono carboxy isooctyl phthalate (MCiOP), mono 3-carboxypropyl phthalate (MCPP), mono 2-carboxymethylhexyl phthalate (MCMHP), mono 2-ethyl-5-carboxypentyl phthalate (MECPP), mono 2-ethylhexyl phthalic acid (MEHP), mono 2-ethyl 5-hydroxyhexyl phthalate (MEHHP), mono 2-ethyl-5-oxy-hexyl phthalate (MEOHP), monoethyl phthalate (MEP), monoisobutyl phthalate (MiBP), monoisononyl phthalate (MiNP), monomethyl phthalate (MMP)] were explored in urine sample using the ‘Waters ACQUITY UPLC-MS/MS System (Waters Acquity, the United States/Milford)’ device using the Acquity UPLC BEH Phenyl Column [42, 43] in Certificated Reference Laboratory (Düzen Norwest Environmental Laboratory, Ankara, Turkey). UPLC-MS/MS system was operated under suitable conditions [the capillary voltage: 3 kV, desolvation temperature: 250 °C, ion source temperature: 150 °C, cone gas: 50 L/h, desolvation gas flow: 600 L/h] according to the manufacturer’s instructions. Details on the analytic methods for the phthalates have been previously [44]. Briefly, urine samples were processed using enzymatic deconjugation and extracted, and analysis was performed. The limit of detection (LOD) was below 0.3 μg/L for all studied metabolites (Suppl Table 1). Quality control (QC) samples were included in every 10 analyzed samples and each analytical run. Recovery rates for phthalate biomarkers ranged from 90.0 to 112.8. Validation of results showed the high accuracy of the method and given in Supplementary Table 1.

The determination of urinary creatinine (Cre) levels was based on Jaffe’s colorimetric method with an automatic biochemical analyzer to correct urine dilution (Beckman Coulter ORS61780 21). The concentrations of urinary phthalate metabolites and BPA were determined as both unadjusted (crude; μg/L) and creatinine corrected values (μg/g-cre).

ΣDEHP [the sum of di (2-ethylhexyl) phthalate (DEHP) metabolites] was estimated by adding the molar concentrations of five metabolites: MEHP (MW = 278), MEHHP (MW = 294), MEOHP (MW = 292), MECPP (MW = 308) and MCMHP (MW = 308) and multiplied by the MW of MECPP [44, 45]. The sum of the dibutyl phthalate (DBP) metabolites (ΣDBP) was calculated by adding the molar concentrations of two metabolites: MiBP and MnBP and multiplied by the MW of MnBP [46].

Primary/secondary DEHP metabolites calculated by dividing the molar concentration of MEHP by the sum of MEHHP, MEOHP, MECPP, and MCMHP.

Detectable rates were fBPA 12.8, MBzP 24.4%, MEP 95.3%, MEHP 97.7%, MCINP 77.9%, MCPP 80.2%, MMP 89.5%, MiNP 11.6% (Supplementary Table 2). Parameters having low detection ratios (< 80%) were only compared with the Chi-square test. Frequencies of cases having the median levels or higher for metabolites in groups were also compared.

Logistic regression were used for the differences in rates of urinary phthalate metabolites having ≥ median value or detectable levels among groups of “day or/and night HT” vs normal profile with four models. Model 0 was crude model. Model 1 included adjustment for age, sex, BMI-SDS and parental HT. Model 2 included platelet, eGFR, U-cre in addition to Model 1. Model 3 covered phthalate metabolites having p value < 0.2 with confounding factors in Model 2. Besides, cases were divided into 4 groups according to MEP (≥ or < median value) and MBzP (detectable or not) and association with day-time and/or night-time HT were evaluated with logistic regression.

The association of the rates of the high ΣDEHP metabolites with cIMT SDS higher than 95th percentile were analyzed with logistic regression in crude model, after adjustment for age, sex and BMI-SDS, and after controlling for age, sex, BMI-SDS, eGFR, U-cre.

The Odds ratio (OR) and 95% confidence interval (CI) were calculated for logistic regressions.

Early studies showed that chloride, rather than sodium, may be crucial for HT [13]. In a study by Kurtz and Morris, salt-sensitive HT was induced by a high NaCl diet; but a non-chloride diet with similar Na loading failed to induce HT [47]. Wilcox had previously shown that hyperchloremia induced renal vasoconstriction [13]. In contrast to these findings, a recent study showed a J-shaped association of chloride with mortality and cardiovascular events; the lowest chloride quartile (≤103.9 mEq/L) had significantly higher all-cause mortality in a group of pre-dialysis patients (median chloride was 106.0 mEq/L) [10]. In parallel to their finding, in our study, serum chloride level was lower in WCHT (103.8 ± 2.0) and ABPM-HT group (103.5 ± 2.2), compared to normal blood pressure group (105.0 ± 1.8). Several mechanisms may have a role in the association between low chloride and HT: a) a decrease in NaCl concentration in the macula densa of the kidneys increases renin secretion and results in the activation of the renin-angiotensin system and retention of sodium and water [48]; b) low chloride may cause inflammation, the highest CRP level was reported in these cases [10]. Additionally, similar to the previous studies [49‐52], our results revealed associations between high blood pressure and elevated WBC count and platelet count even after controlling BMI-SDS. Platelets are known to have important inflammatory functions [53]. Considering that HT is an inflammatory process [6], it was not unexpected that both WBC and platelet counts increased in hypertensive patients compared to those with normal blood pressure. In parallel with our finding, a study conducted with children showed that WBC count was higher in hypertensive patients, regardless of dipping status [50]. In two other studies in adults, it was stated that the WBC count was higher in hypertensive patients and could be used as a risk factor for HT [51, 52].

Serum calcium and ACCa were both higher in the ABPM-HT group, compared to WCHT. Although it did not reach to statistical significance, serum phosphate was lowest in the ABPM-HT group; and lower in WCHT, compared to normal blood pressure group. This trend reflected as higher ACCa to phosphate ratio in the ABPM-HT group and lower ACCa-phosphate product in WCHT, compared to normal blood pressure group. Previous studies have revealed different results for a correlation between serum calcium and HT. In a study by Hazari et al., where the variables were not adjusted by age, BMI, total cholesterol, triglycerides, serum calcium had no effect on HT [54]. But, cross-sectional studies in adult populations from Norway [55] and United States [56], and a recent longitudinal study from Taiwan [8] indicated a positive correlation between serum calcium and HT. The study from Taiwan showed also the association of higher serum calcium levels with metabolic syndrome and diabetes [8]. Calcium may lead to the development of HT by different mechanisms: a) influx of calcium into the smooth muscle of the artery leading to muscle contracture and increase in vascular resistance, b) positive correlation between calcium and cholesterol [55], c) correlation between calcium and PTH may lead to the production of collagen by aortic vascular smooth muscle cells and thickening of vascular wall [57]. Low phosphate, on the other hand, is related to HT, metabolic syndrome, and increased sympathoadrenal activity [11, 12]. Serum phosphate was found to be inversely related to blood pressure in normotensive individuals and to be lowest in hypertensive patients [9, 58]. In a study conducted by Vyssoulis, among 2600 adult patients with WCHT, decreased levels of serum phosphate and calcium-phosphate product were associated with a higher incidence of a non-dipping nocturnal systolic blood pressure and an impaired metabolic profile [59]. In patients with mild essential HT, low phosphate is inversely related to sympathetic adrenal tone and may be caused by increased plasma epinephrine within pathophysiologic arterial concentrations [60]. Epinephrine leads to a shift of phosphate from the extracellular to the intracellular compartment [12]. We cannot conclude, however, a definite cause-and-effect relationship between low phosphate and HT with our study design. Since the relationship between HT and low levels of serum phosphate may be associated with an unbalanced diet (low phosphate and high carbohydrate consumption) [12].

The percentage of detection of MBzP, one of the phthalate metabolites, in the urine was significantly higher in both the WCHT group and the ABPM-HT group compared to the normal blood pressure group. In addition, detectable MBzP remained an independent predictor of either WCHT or ABPM-hypertension. Studies evaluating the relationship between urinary phthalates and blood pressure in children and adolescents are limited and reported controversial results. While an increased risk for blood pressure was found in some studies with urinary MBzP levels in children [61] and adults [62], no correlation was observed in others [63, 64]. In a cross-sectional study of 108 children aged 6–18 years, a positive correlation was reported between urinary MBzP level and systolic blood pressure [61].

In our study, the frequencies of MEP above or equal to median levels were slightly higher in daytime and/or night-time HT group compared to normal blood pressure group. MEP showed an additive interaction with MBzP on daytime and/or night-time HT. However, a study from China in 276 children aged 6–8 years, evaluating 11 phthalate metabolites (MnBP, MEP, MMP, MBzP, MCOP, MCPP, MOP, MEHP, MECPP, MEHHP, and MEOHP) in urine, in boys a 1-ng/ml revealed that an increase in MEP concentration was associated with a 0.016 mmHg decrease in systolic blood pressure [63].

Except for urinary MBzP and MEP levels, other phthalate metabolites showed no significant positive association with blood pressure profiles in our study. This is the first study evaluating WCHT with these pollutants. A recent cross-sectional study from Isfahan (an industrial city in Iran) included 108 children (6–18 years) observed a positive relationship between systolic blood pressure and some metabolites including urinary MnBP and MEHP; no association with MEHHP, MEOHP, and MMP [61]. A recent cross-sectional study from China including a total of 1044 primary school children (6–8 years old) with an electronic sphygmomanometer were studied reported that MMP, MiBP, MnBP, MCEPP, MCMHP, “the sum of MMP, MEP, MiBP, and MnBP”, and “the sum of MCEPP, MEHHP, MEOHP, MCMHP, MEHP, MMP, MEP, MiBP, and MnBP” in urine samples were associated with elevations in systolic/diastolic blood pressure-SDS, pulse pressure, and MAP. Urine MMP level was also significantly associated with the risk of high blood pressure (blood pressure ≥ the 90th percentile for sex/age/height) [65]. National Health and Nutrition Examination Survey (NHANES 2001–2010; aged 20–80 years) showed relationships between high blood pressure and MECPP, MnBP, MEHHP, MMP, MEOHP, and MBzP in the adult age group after adjusting for urinary creatinine, age, sex, ethnicity, and body mass index [62]. The relationship between urinary phthalates and blood pressure (using an aneroid sphygmomanometer) was examined in cross-sectional analyses using a subsample of US children and adolescents (8–19 years) in 2009 to 2012 NHANES. Di-2-ethylhexylphthalate, di-isononyl phthalate, and di-isodecyl phthalate were associated with higher blood pressure (age-, sex- and height-standardized). No association was detected between LMW phthalates and blood pressure [66]. In the Dutch general population (662 adults) blood pressure was not associated with any of the urinary MEP, MiBP, MnBP, MEHHP, MEOHP, MECPP, MBzP, MEHP, and MMP. On the other hand, they reported nonlinear significant associations for MiBP quartiles with systolic blood pressure compared to the first quartile, lowest exposure [67]. The sum of LMW phthalate metabolites and the sum of HMW phthalate metabolites with the analyses of serially assessed exposure (6 samples per case) were not found to be associated with blood pressure in a cohort of 538 children with chronic kidney disease [68]. A population-based, prospective cohort study among 1064 mother-child pairs revealed sex-specific differences for phthalates on blood pressure; higher third-trimester maternal urine concentrations of HMW phthalates, di-2-ehtylhexylphthalate, and di-n-octylphthalate were associated with lower systolic and diastolic blood pressure among girls [69].

In our study a negative association was detected between ΣDBP metabolites and WCHT. In a mouse model, Xie et al. showed an increase in the levels of angiotensin-converting enzyme (ACE) and angiotensin II (AngII) in the DEHP treatment group without a significant change in estradiol level; on the other hand, there was an increase in the level of estradiol in the DBP treatment group, and the expression of ACE, AngII, AT1R, and eNOS in the DBP treatment groups showed no significant change. They suggested that different effects of DEHP and DBP on blood pressure could be related with the different estradiol levels induced with DEHP and DBP [70].

We showed a negative association between MMP and WCHT. In a study conducted by Yao et al. in children, a positive relationship was reported between MMP and high blood pressure [65]. In addition, an inverted-U-shaped relationship with atherosclerosis, which is one of the cardiovascular risk factors, has been reported in adults [71]. This suggests that different effects can be seen at different doses rather than a linear dose-response relationship. A negative association between insulin resistance and MMP in children was reported by Hashemi et al. [61]. It has been reported that insulin resistance and compensatory hyperinsulinemia may lead to increased blood pressure by causing sympathetic system activation, vascular changes, insufficient vascular dilatation and changes in membrane ion exchange [72‐74]. These may be the underlying reasons for this negative relationship we have found.

In our study, we also showed a negative relation between MCPP and ABPM hypertension. Although higher MCCP concentrations in the first-trimester was found to be associated with pregnancy induced hypertension [75], MCPP was has not been found to be associated with hypertension. Our literature search did not reveal any explanation. Further studies are needed to explain these discrepancies.

The results of the studies are inconsistent, the number and type of metabolites studied are not standardized. Differences might also be due to sample matrix, age groups, exposed dosage, and whether the metabolite level is used directly or corrected by urinary creatinine or urine density. In addition, exposure to multiple contaminants is present at the same time and this might cause interaction between pollutants.

In our study, there is no interaction between carotid IMT-SDS and MBzP. We have also shown that phosphate was significantly lower in adolescents with urinary MBzP equal to or greater than the detectable level. Interestingly, a previous study detected both damage in liver and kidney and abnormalities in the trace element and mineral levels DEHP-administered rats [28]. Therefore, additional studies are needed to evaluate the associations of phthalates with blood pressure in different micronutrient status.

Carotid IMT-SDS was found to be associated with ΣDEHP metabolites in our study. Similarly, matrix metalloproteinases-2 and -9 expression which are inducers of atherosclerosis was reported to be increased in rats exposed to DEHP compared with control rats [76]. In addition, MEHP, ΣDEHP, and MnBP exposures in a human study are strongly found to be associated with thicker CIMT in adolescents and young adults in Taiwan [77].

Although it is not known exactly how phthalates cause changes in blood pressure, it is thought that oxidative damage may be responsible. Interestingly, a positive association between MBzP and increased oxidative stress and impaired vascular function was reported in the pediatric age group [78]. There are some proposed mechanisms in experimental models. In a mice study, exposure to DINP, another HMW phthalate, was shown to increase systolic blood pressure, diastolic blood pressure, and MAP, decrease endothelial nitric oxide synthase expression, and nitric oxide production [79]. Exposure to DEHP was found to cause an increase in mouse blood pressure through the renin-angiotensin-aldosterone system depending on different estradiol levels [70].

In our study, we did not find a significant relationship between the tBPA levels and blood pressure profiles. Similarly, a study conducted with 471 Dutch children aged 6–10 years reported that there was no significant relationship between blood pressure and BPA after multiple testing corrections [27]. However, in a study conducted with 132 children aged 6–18 years in Iran, the urinary BPA concentrations of the participants were found to be 282.53 ± 166.02 μg/g-cre and a linear increase in blood pressure was reported among the tertiles determined according to the BPA concentration [80]. In a study examining 39 obese and overweight children aged 3–8 years, a positive correlation was reported between urinary BPA levels and diastolic blood pressure in girls. No relationship was found between BPA and diastolic blood pressure in boys and between BPA and systolic blood pressure in both genders [81]. A multi-center prospective cohort study of children aged 6 months to 16 years with mild-to-moderate chronic kidney disease showed no interaction between blood pressure and urinary BPA levels [68]. No association was found between blood pressure and urinary BPA levels in 662 native Dutch adult subjects [67]. Analytic method, exposed dosage, obesity, and age groups of children might influence the results. We detected high BPA levels in our study. However, the frequency of fBPA above the detection limit was more in WCHT and APBM-HT groups compared to the normal blood pressure profile groups (p = 0.081). Although there was no significant association, further studies are needed with a larger sample size. fBPA is considered to be more toxicologically active than the conjugated BPA [82]. There are no studies evaluating the interaction between fBPA and cardiovascular events.

We identified a higher percentage in parental HT in adolescents with HT. Similarly, familial aggregation with an increased liability of childhood-onset essential HT with parental essential HT is known [83]. Besides genetic predisposition, the same environmental exposures might have a role in theses aggregation.

In our study, we focused on asymptomatic adolescents detected in school screening, without antihypertensive medication. Patients with heart disease and having any drug therapy were excluded to avoid bias. In patients, different treatment interventions may affect the exposure level. The inclusion of asymptomatic adolescents without hypertensive medication might influenced the results and only statistically significant association could be detected with the phthalate MBzP. The small sample size might also affect the detection of associations. The cross-sectional design, relatively small study group, lack of detailed data for exposure, a single measurement of urinary metabolites were the limitations of the study. To some extent, these features might limit the generalizability of the current findings. Fetal origins of HT could not be evaluated in our study due to the single-exposure design. However, prenatal pollutant exposure might have an additive role in childhood blood pressure [64]. To the best of our knowledge, for the first time, 24-h ABPM was performed to examine the relationship between blood pressure and BPA and phthalate metabolites. Demographic characteristics (age, gender, ethnicity, etc.), prenatal and environmental exposure, study design, measurement methods of urinary metabolites and blood pressure, parameters included in the multivariate analysis may all have an impact on the findings and interpretation.