基于卷积神经网络-长短期记忆神经网络模型利用光学体积描记术重建动脉血压波信号

目的 直接动脉血压（arterial blood pressure,ABP）连续监测是侵入式的,传统袖带式的间接血压测量法无法实现连续监测。既往利用光学体积描记术（photoplethysmography,PPG）实现了连续无创血压监测,但其为收缩压和舒张压的离散值,而非ABP波的连续值,本研究期望基于卷积神经网络-长短期记忆神经网络（CNN-LSTM）利用PPG信号波重建ABP波信号,实现连续无创血压监测。方法 构建CNN-LSTM混合神经网络模型,利用重症监护医学信息集（medical information mart for intensive care,MIMIC）中的PPG与ABP波同步记录信号数据,将PPG信号波经预处理降噪、归一化、滑窗分割后输入该模型,重建与之同步对应的ABP波信号。结果 使用窗口长度312的CNN-LSTM神经网络时,重建ABP值与实际ABP值间误差最小,平均绝对误差（mean absolute error,MAE）和均方根误差（root mean square error,RMSE）分别为2.79 mmHg和4.24 mmHg,余弦相似度最大,重建ABP值与实际ABP值一致性和相关性情况良好,符合美国医疗器械促进协会（Association for the Advancement of Medical Instrumentation,AAMI）标准。结论 CNN-LSTM混合神经网络可利用PPG信号波重建ABP波信号,实现连续无创血压监测。     

Non-invasive model-based estimation of aortic pulse pressure using suprasystolic brachial pressure waveforms

Elevated central arterial (aortic) blood pressure is related to increased risk of cardiovascular disease. Methods of non-invasively estimating this pressure would therefore be helpful in clinical practice. To achieve this goal, a physics-based model is derived to correlate the arterial pressure under a suprasystolic upper-arm cuff to the aortic pressure. The model assumptions are particularly applicable to the measurement method and result in a time–domain relation with two parameters, namely, the wave propagation transit time and the reflection coefficient at the cuff. Central pressures estimated by the model were derived from completely automatic, non-invasive measurement of brachial blood pressure and suprasystolic waveform and were compared to simultaneous invasive catheter measurements in 16 subjects. Systolic blood pressure agreement, mean (standard deviation) of difference was ?1 (7) mm Hg. Diastolic blood pressure agreement was 4 (4) mm Hg. Correlation between estimated and actual central waveforms was greater than 90%. Individualization of model parameters did not significantly improve systolic and diastolic pressure agreement, but increased waveform correlation. Further research is necessary to confirm that more accurate brachial pressure measurement improves central pressure estimation.     

Oscillometric measurement of systolic and diastolic blood pressures validated in a physiologic mathematical model

ABSTRACT: BACKGROUND: The oscillometric method of measuring blood pressure with an automated cuff yields valid estimates of mean pressure but questionable estimates of systolic and diastolic pressures. Existing algorithms are sensitive to differences in pulse pressure and artery stiffness. Some are closely guarded trade secrets. Accurate extraction of systolic and diastolic pressures from the envelope of cuff pressure oscillations remains an open problem in biomedical engineering. METHODS: A new analysis of relevant anatomy, physiology and physics reveals the mechanisms underlying the production of cuff pressure oscillations as well as a way to extract systolic and diastolic pressures from the envelope of oscillations in any individual subject. Stiffness characteristics of the compressed artery segment can be extracted from the envelope shape to create an individualized mathematical model. The model is tested with a matrix of possible systolic and diastolic pressure values, and the minimum least squares difference between observed and predicted envelope functions indicates the best fit choices of systolic and diastolic pressure within the test matrix. RESULTS: The model reproduces realistic cuff pressure oscillations. The regression procedure extracts systolic and diastolic pressures accurately in the face of varying pulse pressure and arterial stiffness. The root mean squared error in extracted systolic and diastolic pressures over a range of challenging test scenarios is 0.3 mmHg. CONCLUSIONS: A new algorithm based on physics and physiology allows accurate extraction of systolic and diastolic pressures from cuff pressure oscillations in a way that can be validated, criticized, and updated in the public domain.     

Assessment of central haemomodynamics from a brachial cuff in a community setting

ABSTRACT: BACKGROUND: Large artery stiffening and wave reflections are independent predictors of adverse events. To date, their assessment has been limited to specialised techniques and settings. A new, more practical method allowing assessment of central blood pressure from waveforms recorded using a conventional automated oscillometric monitor has recently been validated in laboratory settings. However, the feasibility of this method in a community based setting has not been assessed. METHODS: One-off peripheral and central haemodynamic (systolic and diastolic blood pressure (BP) and pulse pressure) and wave reflection parameters (augmentation pressure (AP) and index, AIx) were obtained from 1,903 volunteers in an Austrian community setting using a transferfunction like method (ARCSolver algorithm) and from waveforms recorded with a regular oscillometric cuff. We assessed these parameters for known differences and associations according to gender and age deciles from <30 years to >80 years in the whole population and a subset with a systolic BP < 140 mmHg. RESULTS: We obtained 1,793 measures of peripheral and central BP, PP and augmentation parameters. Age and gender associations with central haemodynamic and augmentation parameters reflected those previously established from reference standard non-invasive techniques under specialised settings. Findings were the same for patients with a systolic BP below 140 mmHg (i.e. normotensive). Lower values for AIx in the current study are possibly due to differences in sampling rates, detection frequency and/or averaging procedures and to lower numbers of volunteers in younger age groups. CONCLUSION: A novel transfer-function like algorithm, using brachial cuff-based waveform recordings, provides robust and feasible estimates of central systolic pressure and augmentation in community-based settings.     

Association of Processed Meat Intake with Hypertension Risk in Hemodialysis Patients: A Cross-Sectional Study

In this cross-sectional study, we hypothesized that hemodialysis patients consuming greater processed meat is associated with hypertension risk, which can be partly explained by the high sodium content in processed meat. From September 2013 to May 2014, one hundred and four patients requiring chronic hemodialysis treatment were recruited from hemodialysis centers. Data on systolic blood pressure and diastolic blood pressure before receiving dialysis, and 3-day dietary records of the recruited patients were collected. HD patients with systolic and diastolic blood pressures greater than140 mmHg and higher than 90 mmHg, respectively, were considered hypertension risk. Protein foods were divided into 4 categories: red meat, white meat, soybeans, and processed meat (e.g., sausage and ham). In a model adjusted for energy intake and hypertension history, additional servings of processed meats was positively associated to systolic blood pressure >140 mmHg (odds ratio [95% confidence interval]: 2.1 [1.0–4.3]), and diastolic blood pressure > 90 mmHg (odds ratio: 2.5 [1.2–5.5]). After adjustment for dietary sodium contents or body mass index (BMI), most associations were substantially attenuated and were no longer significant. In systolic blood pressure greater than140 mmHg, one serving per day of red meats (β = -1.22, P < .05) and white meats (β = -0. 75, P = .05) was associated with a reduced risk compared with one serving per day of processed meats. Similarly, compared with one serving per day of processed meat, a reduced risk of diastolic blood pressure higher than 90 mmHg was associated with one serving per day of red meat (β = -1. 59, P < .05), white meat (β = -0. 62, P < .05). Thus, in these hemodialysis patients, intake of processed meat is significantly positively associated with higher blood pressure risk, and both sodium contents in processed meat and BMI significantly contributes to this association.     

High altitude and blood pressure in children

We aimed to evaluate the blood pressure of children who had similar demographic characteristics but lived at different altitudes. Blood pressure of the children attending primary schools in Izmir (sea level: n = 425) and Van (altitude: 1725 m, n = 291) were measured by mercurial sphygmomanometer for this study. They were similar with respect to age, sex, weight, height, and BMI. Mean age of the children was 10.51 +/- 0.87 years (range: 9 to 12 years), and 358 (50 percent) of them were female. Mean systolic blood pressure was significantly higher in the children living in Van than in the children living in Izmir (104.72 +/- 11.2 vs. 97.96 +/- 25.5 mmHg, respectively, p < .001). Similarly mean diastolic blood pressure was significantly higher in the children living in Van than in the children living in Izmir (63.98 +/- 9.3 vs. 59.91 +/- 10.0 mmHg, respectively, p < .001). When blood pressure was evaluated with regard to height percentile, the number of children with a blood pressure over 90 percentile were 19 (4.5 percent) and 48 (16.5 percent) for systolic blood pressure, and 25 (5.9 percent) and 37 (12.7 percent) for diastolic blood pressure among the children living in Izmir and Van, respectively (p < .001). Systolic and diastolic blood pressures were found to increase in parallel to the increase in body mass index in children living in Van (r = 0.358, p < .001 and r = 0.235, p < .001, respectively). However, blood pressures were not correlated to body mass index in children living in Izmir. A difference of 1700 m in altitude was associated with higher systolic and diastolic blood pressure levels in children with similar demographic characteristics, and at this altitude, body mass index and blood pressure showed a positive correlation.     

Blood pressure in 15 inbred mouse strains and its lack of relation with obesity and insulin resistance in the progeny of an NZO/HILtJ × C3H/HeJ intercross

We characterized the systolic and diastolic blood pressures of 10-week-old males from 15 inbred mouse strains and found that blood pressures among strains were continuously distributed and that strain C3H/HeJ had the lowest mean systolic and diastolic pressure (100.5 ± 3.2 and 66.8 ± 3.5 mmHg), and a strain with obesity and diabetes, NZO/HILtJ, had the highest (132.4 ± 3.1 and 86.6 ± 6.9 mmHg). To understand the relationship of blood pressure with insulin resistance and obesity, we produced F₁ and F₂ progeny from reciprocal crosses of NZO, the strain with obesity, diabetes, and high blood pressure, and the strain with the lowest blood pressures, C3H/HeJ. Mean systolic pressures of 10-week-old (NZO × C3H)F₁ and (C3H × NZO)F₁ males were similar to each other (114.9 ± 3.8 and 117.2 ± 5.0 mmHg) and were intermediate to those of the parental strains. Systolic pressure of F₂ males (n = 223) was distributed normally about the mean, suggesting that blood pressure is a polygenic trait. The body mass index (BMI) and plasma insulin levels of F₂ progeny correlated significantly and positively with plasma leptin levels, suggesting that obesity is associated with insulin resistance. In contrast, systolic pressure did not correlate with BMI, plasma leptin levels, and plasma insulin levels, suggesting that genes underlying the development of hypertension in this intercross are not associated with the development of obesity and insulin resistance. Our results demonstrate that the progeny of NZO and C3H intercrosses are a practical and powerful tool for identifying blood pressure genes and for understanding human polygenic hypertension.(Fumihiro Sugiyama) These authors contributed equally to this study.     

Auscultatory indirect measurement of blood pressure in dogs

An indirect method of measuring blood pressure (cuff plus stethoscope) was evaluated in 70 dogs weighing 15 to 30 kg (17.5 +/- 8.8 kg; mean +/- standard deviation). A cuff 12 cm wide was used. The measurements were most audible with the cuff on the upper foreleg of the dog and with the stethoscope placed in the medial epicondylar region just distal to the cuff. The cuff was inflated to greater than systolic pressure and allowed to deflate slowly. In 70 lightly sedated dogs, systolic blood pressures averaged 145 +/- 25 mmHg (mean +/- standard deviation) and diastolic blood pressures averaged 84 +/- 14 mmHg. Indirect measurements were compared to direct measurements (femoral arterial catheter). Systolic pressures obtained by this direct method averaged 138 +/- 29 mmHg (mean +/- standard deviation) and diastolic pressures averaged 84 +/- 17 mmHg. The correlation coefficient for systolic pressure was 0.96 and for diastolic pressure 0.97.     

Direct effects in vivo of angiotensins I and II on the canine sinus node.

The chronotropic responses to angiotensins I and II (5 micrograms in 1 mL Tyrode's solution) injected into the sinus node artery were assessed before and after the intravenous administration of captopril (2 mg/kg) and saralasin (20 micrograms/kg) in anaesthetized dogs. The effects of angiotensin II given intravenously were also observed. The animals (n = 8) were vagotomized and pretreated with propranolol (1 mg/kg, i.v.) to prevent baroreceptor-mediated responses to increases in blood pressure. Injection of angiotensin I into the sinus node artery induced significant increases in heart rate (114 +/- 6 vs. 133 +/- 6 beats/min) and in systemic systolic (134 +/- 13 vs. 157 +/- 14 mmHg; 1 mmHg = 133.3 Pa) and diastolic (95 +/- 10 vs. 126 +/- 13 mmHg) blood pressures. Similar results were obtained when angiotensin II was injected into the sinus node artery, but intravenous injection induced changes in systolic (138 +/- 8 vs. 180 +/- 25 mmHg) and diastolic (103 +/- 8 vs. 145 +/- 20 mmHg) blood pressures only. Captopril induced a significant decrease in systolic (118 +/- 11 vs. 88 +/- 12 mmHg) and diastolic (84 +/- 9 vs. 59 +/- 9 mmHg) blood pressures without affecting the heart rate (109 +/- 6 vs. 106 +/- 6 beats/min). Saralasin produced a significant increase in systolic (109 +/- 7 vs. 126 +/- 12 mmHg) blood pressure only. Increments in heart rate and systolic and diastolic blood pressures in response to angiotensins I and II were, respectively, abolished by captopril and saralasin. It was concluded that angiotensin II has, in vivo, a direct positive chronotropic effect that can be blocked by saralasin.(ABSTRACT TRUNCATED AT 250 WORDS)     

Biofeedback of R-wave-to-pulse interval normalizes blood pressure

We investigated whether biofeedback of the R-wave-to-pulse interval, a measure related to the pulse wave velocity, enables participants with either high or low arterial blood pressure to modify their blood pressure. Twelve participants with high blood pressure (mean systolic blood pressure = 142.6 ± 13.5 mmHg; mean diastolic blood pressure = 99.9 ± 12.3 mmHg) and 10 participants with low blood pressure (mean systolic blood pressure = 104.8 ± 6.6 mmHg; mean diastolic blood pressure = 73.2 ± 4.2 mmHg) received 3 individual sessions of RPI biofeedback within a 2-week period. Participants with high blood pressure were rewarded for decreasing and participants with low blood pressure for increasing their blood pressure. Standard arm-cuff blood pressure measurements across the sessions served as dependent variables. Participants with high blood pressure achieved significant reductions of systolic (15.3 mmHg) and diastolic (17.8 mmHg) blood pressure levels from the beginning of the first to the end of the last training session. In contrast, participants with low blood pressure achieved significant increases in systolic (12.3 mmHg) and diastolic (8.4 mmHg) blood pressure levels. The degree of blood pressure changes in this study might be of clinical relevance. With prolonged and refined training regimens, even larger effects seem to be likely.     

Evaluation of blood pressure changes and heart rate in young health men based on results of a 24-hour monitoring of pressure]

Systolic and diastolic blood pressures and heart rate were monitored in a group of 20 young healthy men for 24 hours. Period of time between 8 o'clock a.m. and 10 o'clock p.m. was treated as waking state whereas period of time from 12 p.m. to 6 a.m. as sleep phase. Mean value of systolic blood pressure for waking state was 124.6 +/- 7.6 mm Hg, and for sleep phase 110.4 +/- 11.5 mm Hg. (p < .001). Mean diastolic blood pressures were also significantly different (76.5 +/- 5.9 mm Hg and 63.8 +/- 7.7 mm Hg, respectively), the same concerns heart rate (79.6 +/- 6.4 and 63.0-7.2 min-1, respectively). In both cases p < .001. To evaluate dependence of heart rate on systolic blood pressure in waking state the following calculation was made: HR = 0.230 x systolic blood pressure +51.4 (r = 0.24; p < .001) whereas for sleep phase r did not reach a level of statistical significance (HR = 0.074 x systolic blood pressure + 53.9; r = 0.094). Single or even multiple measurements of the arterial blood pressure are not sufficient to evaluate circadian changes.     

Noninvasive measurement of blood pressure in African green monkeys (Cercopithecus aethiops)

Indirect measurements of arterial blood pressure were made in African green monkeys (Cercopithecus aethiops) employing a Doppler ultrasound stethoscope and standard cuff and an Infrasonde automatic blood pressure recorder. Measurements were obtained from anesthetized (10 mg/kg ketamine (HCI) and unanesthetized (1.5 mg/kg ketamine HCI) animals. Ketamine had no significant effect on blood pressure. Indirect measurements from the brachial artery were compared with direct femoral artery measurements and with each other. Systolic blood pressures measured by the Doppler (r = .948) and Infrasonde (r = .920) methods correlated closely with direct measurements but were significantly lower than systolic blood pressures measured by the direct method. Diastolic blood pressures measured by the Infrasonde method agreed closely with direct measurements (r = .947). Systolic blood pressures measured by the indirect methods correlated closely in both anesthetized (r = .973) and unanesthetized (r = .834) animals and were not significantly different. Mean blood pressures calculated from direct and Infrasonde measurements also correlated closely (r = .963), with direct measurements being 4 mmHg higher on the average. Mean blood pressures are less influenced by methodology and are more reproducible than other pressures. These noninvasive methods can be used to obtain simple and accurate measurements of blood pressure from anesthetized and unanesthetized monkeys and are of value in long-term studies in monkeys.     

Sedentary Behavior and Light Physical Activity Are Associated with Brachial and Central Blood Pressure in Hypertensive Patients

Background

Physical activity is recommended as a part of a comprehensive lifestyle approach in the treatment of hypertension, but there is a lack of data about the relationship between different intensities of physical activity and cardiovascular parameters in hypertensive patients. The purpose of this study was to investigate the association between the time spent in physical activities of different intensities and blood pressure levels, arterial stiffness and autonomic modulation in hypertensive patients.

Methods

In this cross-sectional study, 87 hypertensive patients (57.5 ± 9.9 years of age) had their physical activity assessed over a 7 day period using an accelerometer and the time spent in sedentary activities, light physical activities, moderate physical activities and moderate-to-vigorous physical activities was obtained. The primary outcomes were brachial and central blood pressure. Arterial stiffness parameters (augmentation index and pulse wave velocity) and cardiac autonomic modulation (sympathetic and parasympathetic modulation in the heart) were also obtained as secondary outcomes.

Results

Sedentary activities and light physical activities were positively and inversely associated, respectively, with brachial systolic (r = 0.56; P < 0.01), central systolic (r = 0.51; P < 0.05), brachial diastolic (r = 0.45; P < 0.01) and central diastolic (r = 0.42; P < 0.05) blood pressures, after adjustment for sex, age, trunk fat, number of antihypertensive drugs, accelerometer wear time and moderate-to-vigorous physical activities. Arterial stiffness parameters and cardiac autonomic modulation were not associated with the time spent in sedentary activities and in light physical activities (P > 0.05).

Conclusion

Lower time spent in sedentary activities and higher time spent in light physical activities are associated with lower blood pressure, without affecting arterial stiffness and cardiac autonomic modulation in hypertensive patients.     

Wave reflection augments central systolic and pulse pressures during facial cooling

Cardiovascular events are more common in the winter months, possibly because of hemodynamic alterations in response to cold exposure. The purpose of this study was to determine the effect of acute facial cooling on central aortic pressure, arterial stiffness, and wave reflection. Twelve healthy subjects (age 23 +/- 3 yr; 6 men, 6 women) underwent supine measurements of carotid-femoral pulse wave velocity (PWV), brachial artery blood pressure, and central aortic pressure (via the synthesis of a central aortic pressure waveform by radial artery applanation tonometry and generalized transfer function) during a control trial (supine rest) and a facial cooling trial (0 degrees C gel pack). Aortic augmentation index (AI), an index of wave reflection, was calculated from the aortic pressure waveform. Measurements were made at baseline, 2 min, and 7 min during each trial. Facial cooling increased (P < 0.05) peripheral and central diastolic and systolic pressures. Central systolic pressure increased more than peripheral systolic pressure (22 +/- 3 vs. 15 +/- 2 mmHg; P < 0.05), resulting in decreased pulse pressure amplification ratio. Facial cooling resulted in a robust increase in AI and a modest increase in PWV (AI: -1.4 +/- 3.8 vs. 21.2 +/- 3.0 and 19.9 +/- 3.6%; PWV: 5.6 +/- 0.2 vs. 6.5 +/- 0.3 and 6.2 +/- 0.2 m/s; P < 0.05). Change in mean arterial pressure but not PWV predicted the change in AI, suggesting that facial cooling may increase AI independent of aortic PWV. Facial cooling and the resulting peripheral vasoconstriction are associated with an increase in wave reflection and augmentation of central systolic pressure, potentially explaining ischemia and cardiovascular events in the cold.     

Daily Liquorice Consumption for Two Weeks Increases Augmentation Index and Central Systolic and Diastolic Blood Pressure

Background

Liquorice ingestion often elevates blood pressure, but the detailed haemodynamic alterations are unknown. We studied haemodynamic changes induced by liquorice consumption in 20 subjects versus 30 controls with average blood pressures of 120/68 and 116/64 mmHg, respectively.

Methods

Haemodynamic variables were measured in supine position before and after two weeks of liquorice consumption (daily glycyrrhizin dose 290–370 mg) with tonometric recording of radial blood pressure, pulse wave analysis, and whole-body impedance cardiography. Thirty age-matched healthy subjects maintaining their normal diet were studied as controls.

Results

Two weeks of liquorice ingestion elevated peripheral and central systolic and diastolic blood pressure (by 7/4 and 8/4 mmHg, 95% confidence intervals [CI] 2-11/1-8 and 3-13/1-8, respectively, P<0.05), and increased extracellular volume by 0.5 litres (P<0.05 versus controls). Also augmentation index adjusted to heart rate 75/min (from 7% to 11%, 95% CI for change 0.3-7.5, P<0.05) and aortic pulse pressure (by 4 mmHg, 95% CI 1-7, P<0.05) were elevated indicating increased wave reflection from the periphery. In contrast, peripheral (−3/−0.3 mmHg) and central blood pressure (−2/−0.5 mmHg), aortic pulse pressure (−1 mmHg), and augmentation index adjusted to heart rate 75/min (from 9% to 7%) decreased numerically but not statistically significantly without changes in extracellular volume in the control group. Heart rate, systemic vascular resistance, cardiac output, and pulse wave velocity did not differ between the groups.

Conclusions

Two weeks of daily liquorice consumption increased extracellular volume, amplified pressure wave reflection from the periphery, and elevated central systolic and diastolic blood pressure.

Trial Registration

EU Clinical Trials Register EudraCT 2006-002065-39</url>ClinicalTrials.gov {"type":"clinical-trial","attrs":{"text":"NCT01742702","term_id":"NCT01742702"}}NCT01742702     

Numerical study to evaluate the effect of a surface-based sensor on arterial tonometry

Abstract

Arterial tonometry is a widely used non-invasive blood pressure measurement method. In contrast to the cuff-based method, it is possible to obtain a continuous pressure profile with respect to systolic and diastolic pressures using this method. However, due to a requirement of arterial tonometry—that a sensor needs to be placed directly above a blood vessel—placement error is inevitable if the measurement device is only capable of measuring local regions. This study assumed that the plate sensor is flexible, thus reducing the placement error. We investigated the pressure distribution along the wrist surface rather than the local region through the contact simulation between the flexible plate sensor and the wrist. As a result, we concluded that there is a unique pressure distribution for any specific wrist, regardless of the length and position of the plate, and that it is possible to measure the blood pressure using the response at the wrist surface to the pressure inside the radial artery.     

Device-Guided Paced Respiration as an Adjunctive Therapy for Hypertension in Obstructive Sleep Apnea: A Pilot Feasibility Study

Data suggest that device-guided paced respiration (<10 breaths/min) may reduce blood pressure in hypertensive patients. We hypothesized that daily device-guided slow breathing may lower blood pressure in patients with hypertension and obstructive sleep apnea (OSA). In this one-arm pilot study, we enrolled 25 subjects with hypertension and OSA. Subjects were asked to perform device-guided paced respiration 30 min a day for 8 weeks. Our primary outcome was change in office systolic and diastolic blood pressures from baseline to 8 weeks. Twenty-four subjects completed the study. Mean baseline blood pressure was 140.0 ± 10.2 mmHg systolic and 82.7 ± 8.9 mmHg diastolic. Complete device data were available for 17 subjects. Mean device adherence was 81 ± 24% and 51% achieved a mean breath rate ≤10 breaths/min over 8 weeks. Three subjects had changes in their anti-hypertensive medications during the study. Among the remaining 21 subjects, mean difference in office blood pressure from baseline to 8 weeks was −9.6 ± 11.8 mmHg systolic (p ≤ 0.01) and −2.52 ± 8.9 mmHg diastolic (p = 0.21). Device-guided paced respiration may lower systolic blood pressure in patients with hypertension and OSA; however, our findings need to be confirmed with larger randomized controlled trials.     

International gestational age-specific centiles for blood pressure in pregnancy from the INTERGROWTH-21st Project in 8 countries: A longitudinal cohort study

BackgroundGestational hypertensive and acute hypotensive disorders are associated with maternal morbidity and mortality worldwide. However, physiological blood pressure changes in pregnancy are insufficiently defined. We describe blood pressure changes across healthy pregnancies from the International Fetal and Newborn Growth Consortium for the 21st Century (INTERGROWTH-21st) Fetal Growth Longitudinal Study (FGLS) to produce international, gestational age-specific, smoothed centiles (third, 10th, 50th, 90th, and 97th) for blood pressure.Methods and findingsSecondary analysis of a prospective, longitudinal, observational cohort study (2009 to 2016) was conducted across 8 diverse urban areas in Brazil, China, India, Italy, Kenya, Oman, the United Kingdom, and the United States of America. We enrolled healthy women at low risk of pregnancy complications. We measured blood pressure using standardised methodology and validated equipment at enrolment at <14 weeks, then every 5 ± 1 weeks until delivery.We enrolled 4,607 (35%) women of 13,108 screened. The mean maternal age was 28·4 (standard deviation [SD] 3.9) years; 97% (4,204/4,321) of women were married or living with a partner, and 68% (2,955/4,321) were nulliparous. Their mean body mass index (BMI) was 23.3 (SD 3.0) kg/m². Systolic blood pressure was lowest at 12 weeks: Median was 111.5 (95% CI 111.3 to 111.8) mmHg, rising to a median maximum of 119.6 (95% CI 118.9 to 120.3) mmHg at 40 weeks’ gestation, a difference of 8.1 (95% CI 7.4 to 8.8) mmHg. Median diastolic blood pressure decreased from 12 weeks: 69.1 (95% CI 68.9 to 69.3) mmHg to a minimum of 68.5 (95% CI 68.3 to 68.7) mmHg at 19⁺⁵ weeks’ gestation, a change of −0·6 (95% CI −0.8 to −0.4) mmHg. Diastolic blood pressure subsequently increased to a maximum of 76.3 (95% CI 75.9 to 76.8) mmHg at 40 weeks’ gestation. Systolic blood pressure fell by >14 mmHg or diastolic blood pressure by >11 mmHg in fewer than 10% of women at any gestational age. Fewer than 10% of women increased their systolic blood pressure by >24 mmHg or diastolic blood pressure by >18 mmHg at any gestational age. The study’s main limitations were the unavailability of prepregnancy blood pressure values and inability to explore circadian effects because time of day was not recorded for the blood pressure measurements.ConclusionsOur findings provide international, gestational age-specific centiles and limits of acceptable change to facilitate earlier recognition of deteriorating health in pregnant women. These centiles challenge the idea of a clinically significant midpregnancy drop in blood pressure.

Lauren Green and colleagues study blood pressure in pregnant women across a range of countries.     

Blood pressure in young adults as influenced by copper and zinc intake

The relationship of copper and zinc status with blood pressure in young adults was studied (N=59). Copper and zinc status was assessed through analysis of serum, urine, hair, and diet records. Males (N=27) had greater systolic and pulse pressures than females (N=32). Multiple regression analysis suggested that males and higher caloric and dietary copper intakes were the predominant factors associated with higher systolic and diastolic pressures, whereas higher dietary zinc intakes were associated with lower pressures. Higher urine zinc excretion rates and serum zinc concentrations were associated with higher diastolic pressures. The sex variable was the strongest factor affecting pulse pressures, with males having higher pulse pressures. Serum zinc concentrations had an inverse relationship to pulse pressures. The data suggest that the weight/height index and age had a low association with blood pressure in this normotensive sample.