Abstract
Introduction
Extracellular volume (ECV) predicts mortality in hemodialysis patients, but it is difficult to assess clinically. Peridialytic blood pressure (BP) measurements can help ECV assessment. Orthostatic BP is routinely measured clinically, but its association with ECV is unknown.
Methods
In a cohort of hypertensive hemodialysis patients, we measured post-hemodialysis ECV/weight with bioimpedance spectroscopy and analyzed its association with post-HD orthostatic BP measurements obtained during routine care. Using linear and logistic regression, the primary outcomes were orthostatic BP change and orthostatic hypotension (OH) defined by systolic BP decrease of at least 20 mmHg or diastolic decrease of at least 10 mmHg. Model 1 controlled for sex, age, diabetes. Model 2 additionally controlled for ultrafiltration rate and antihypertensive medications. We conducted sensitivity analysis using OH definition of systolic BP decrease of at least 30 mmHg.
Findings
Among 57 participants, mean orthostatic systolic BP change was −7.30 (20) mmHg and mean ECV/weight was 0.24 (0.04) L/kg. Post-HD ECV/weight was not associated with orthostatic systolic BP change (β=8.2, p=.6). There were 16 participants with and 41 participants without OH. The ECV/weight did not differ between these groups (0.22 [0.04] vs. 0.24 [0.05] L/Kg, p=.09) and did not predict OH in logistic regression (OR 11, 4.04; 95% CI .2–671, .03–530 in the 2 models.) In a sensitivity analysis, ECV/weight was lower in the OH group (0.22 [0.03] vs. 0.25 [0.04] L/kg, p=0.005), but this was accompanied by differences in sex and diabetes. Using logistic regression, there was no independent association between ECV/weight with OH.
Discussion
Orthostatic systolic BP change after HD completion is not a reliable indicator of post-hemodialysis ECV. When considering other factors associated with orthostatic BP, ECV/weight is not independently associated with OH. Although transient post-dialytic differences in intravascular volume may be associated with OH, post-hemodialysis OH does not necessarily indicate ECV depletion.
INTRODUCTION
Extracellular volume (ECV) overload is an independent risk factor for mortality in hemodialysis (HD) patients(1), but there is no gold standard method to assess or manage ECV overload in clinical practice. Clinicians often infer ECV from HD-unit BP measurements, and there is evidence that individual peridialytic measurements, changes in BP from pre to post-HD, and intradialytic BP slope help with assessment of ECV overload(2–4). A major challenge in fluid management in HD patients is striking balance between fluid removal that is aggressive enough to minimize volume-associated morbidities without inducing intradialytic hypotension, another independent risk factor for mortality(5). This is further complicated by the occurrence of hemodynamic changes related to excessive vasodilation that can occur during HD.
Most literature relating BP with assessment of ECV is based upon seated BP measurements only. Standing BP measurements are often obtained after HD mainly to ensure hemodynamic stability prior to discharge, but such measurements may provide important information regarding the volume status of a patient that could be considered in upcoming dialysis treatments.
Even though HD units frequently measure orthostatic vitals on patients upon completion of dialysis, there is limited evidence on the significance of these measurements with regards to ECV. This study’s purpose was to evaluate the relationship between post-HD orthostatic BP changes with objective measurements of ECV in hypertensive HD patients. From a prospective cohort of hypertensive HD patients, we analyzed patients who had available data on post-HD bioimpedance spectroscopy (BIS) as well as documented post-HD seated and post-HD standing BP measurements. We hypothesized that post-HD orthostatic BP changes would be associated with post-HD BIS-determined ECV/weight.
MATERIALS AND METHODS
Study Design and Subjects
We conducted a retrospective analysis of a cohort of 75 hypertensive HD patients. This cohort consisted of 36 participants from a case-control study comparing BIS measurements in hypertensive HD patients with and without recurrent intradialytic hypertension (systolic BP increase or decrease ≥10 mmHg from pre to post-HD in ≥4/6 screening treatments) and was later expanded with an additional 39 hypertensive HD patients independent of intradialytic BP patterns. Participants were consecutively enrolled based on meeting eligibility criteria from our three university-affiliated DaVita HD units. Inclusion criteria were age>18, end-stage renal disease on maintenance HD >1 month, hypertension defined by a pre-HD systolic BP >140 mmHg or post-HD systolic BP>130 mmHg. Exclusion criteria were major extremity amputation, presence of cardiac defibrillator or pacemaker, coronary artery stents, implanted metallic prostheses, or pregnancy. We obtained written informed consent from all subjects prior to any study procedures. The University of Texas Southwestern Medical Center Institutional Review Board approved the protocol, and all procedures were in accordance with the Declaration of Helsinki. The study was part of a registered clinical trial, NCT01862497(6)
Procedures
All BP and BIS measurements were collected before, during, or after the same mid-week HD treatment.
Hemodialysis Unit BP Measurements
Peridialytic and intradialytic BP measurements were obtained in the HD-unit by dialysis nurses or technicians. Blood pressures were obtained in the brachial artery of the non-access arm using a sphygmomanometer attached to the HD machine and were automatically recorded into the participants’ electronic medical records. Pre-HD measurements were obtained in the seated position just prior to having the dialysis access stuck or being connected to the HD machine. Measurements were obtained per protocol every 30 minutes with more frequent measurements if the patient was symptomatic, required an intervention, or developed intradialytic hypotension or severely elevated BP.
The final HD BP measurement was defined as the last seated measurement during dialysis.
The post-HD seated BP measurement was defined as the measurement obtained after dialysis had been completed and as soon as the blood was returned.
The post-HD standing measurement was defined as the measurement obtained upon initial standing just following the post-HD seated measurement.
We defined the orthostatic BP change as the post-HD standing measurement – the post-HD seated measurement.
We reviewed all nursing notes to ascertain for use of intradialytic antihypertensive medications.
Biompedance Spectroscopy
Approximately 30 minutes following HD, we obtained whole body multi-frequency BIS measurements (Impedimed SFB7, Carlsbad CA, USA) in the supine position. We were able to obtain estimates of extracellular water, as well as intracellular and total body water. We calculated the ratio of BIS-determined ECV/body weight based on the patient’s post-HD weight measurement using the HD unit scale.
Ambulatory Blood Pressure
All subjects underwent ambulatory BP measurements following the mid-week HD treatment using a Spacelabs 90207 monitor with BP measured every 30 minutes during the daytime and every hour at night for the 44-hour interdialytic period.
Study Outcomes
The primary dependent variable was orthostatic systolic BP change, defined as the difference between the documented post-HD standing BP and post-HD seated systolic BP. We also considered the orthostatic diastolic BP change based on these positions.
The secondary dependent variable was orthostatic hypotension (OH), defined as a decrease in orthostatic systolic BP of at least 20 mmHg or a decrease in diastolic BP of at least 10 mmHg(7)
Statistical Analysis
The descriptive statistics are presented as mean and standard deviation for continuous variables and frequencies and percentages for categorical variables. Two-sided Wilcoxon-rank statistics and Mantel-Haenszel chi-square statistics were used to test differences in continuous and categorical variables between different groups, respectively.
General Linear Models (GLM) were used to test associations of post dialysis BIS-determined ECV/body weight (L/kg) with changes in orthostatic systolic and diastolic BP. Post dialysis BIS-determined ECV/body weight (L/kg) was non-normally distributed thus Box-Cox transformation was used to transform non-normal post dialysis BIS-determined ECV/body weight (L/kg) to normally distributed variables.
In Model 1, we controlled for the covariates of age, sex, and presence of diabetes as clinically relevant baseline characteristics. In Model 2, we additionally controlled for the ultrafiltration rate (mL/kg/hr) during the dialysis treatment and the use of the following medications: alpha-adrenergic receptor blockers (α-blocker), calcium channel blockers (CCB), and statins.
The binary outcome of OH was defined as systolic BP decrease more than or equal to 20 mmHg or a diastolic BP decrease more than or equal to 10 mmHg(7). We conducted logistic regression model to test whether post dialysis BIS-determined ECV/body weight (L/kg) explains OH risk post dialysis. We utilized the same covariates as above for Model 1 and Model 2, respectively. We then conducted sensitivity analyses where the definition of OH included systolic BP decrease more than or equal to 30 mmHg or a diastolic BP decrease more than or equal to 10 mmHg as a consideration of a larger magnitude in BP drop for patients with baseline hypertension(7).
RESULTS
Of the 75 participants from the original cohort, there were 62 that had both post-HD ECV/weight and post-HD orthostatic BP measurements available. We excluded 2 participants that did not have complete data available on antihypertensive drug use and 3 that received intradialytic clonidine. Of the remaining 57 subjects, mean age was 48.8 (12) years, and there were a large percentage of participants that were African American or had diabetes mellitus (Table 1). Both the percentage of participants taking CCB and the percentage of participants taking α-blockers were 61%.
Table 1.
Characteristics of the cohort, including a comparison of participants with and without orthostatic hypotension
| Data presented as Mean (standard deviation) Or percentage | Total cohort (n=57) | With orthostatic hypotension (n=16) | Without orthostatic hypotension (n=41) | p-value (comparison of with and without orthostatic hypotension) |
|---|---|---|---|---|
| Age (years) | 48.8 (12) | 51.1 (12) | 48.0 (12) | .31 |
| Male n (%) | 37 (65) | 8 (50) | 29 (71) | .14 |
| African American n (%) | 32 (56) | 9 (56) | 23 (56) | .9 |
| Hispanic, n (%) | 18 (32) | 6 (38) | 12 (29) | .55 |
| Diabetes mellitus, n (%) | 33 (58) | 12 (75) | 21 (51) | .11 |
| Total number of antihypertensive medications | 2.96 (1.1) | 3 (1.1) | 2.95 (1.1) | .85 |
| Taking alpha adrenergic receptor blocker, n (%) | 35 (61) | 9 (56) | 26 (63) | .62 |
| Taking calcium channel blocker, n (%) | 35 (61) | 6 (38) | 29 (71) | .02 |
| Taking statin n (%) | 20 (35) | 9 (56) | 11 (27) | .06 |
| Post-hemodialysis extracellular volume/weight (L/kg) | 0.24 (0.04) | 0.22 (0.04) | 0.24 (0.05) | .09 |
| Body Mass Index (kg/m2) | 30.9 (8.8) | 29.7 (6.8) | .62 | |
| Estimated dry weight (kg) | 86.5 (20) | 87.0 (24) | 86.1 (18) | .71 |
| Percentage of interdialytic weight gain | 2.88 (1.6) | 2.21 (1.1) | 3.15 (1.8) | .04 |
| Volume removed (L) | 2.69 (1.2) | 2.34 (1.1) | 2.83 (1.2) | .17 |
| Ultrafiltration rate (mL/kg/hr) | 7.97 (3.6) | 6.7 (2.7) | 8.2 (3.6) | .15 |
| Blood urea nitrogen (mg/dL) | 55.5 (17) | 55.0 (16) | 55.7 (17) | .93 |
| Serum creatinine (mg/dL) | 10.5 (2.6) | 9.50 (2.4) | 10.9 (2.6) | .78 |
| Serum potassium (mmol/L) | 4.74 (0.6) | 4.66 (0.4) | 4.77 (0.6) | .27 |
| Kt/V | 1.48 (0.3) | 1.48 (0.2) | 1.49 (0.3) | .57 |
| Hemoglobin (g/dL) | 10.6 (1.3) | 10.6 (1.0) | 10.6 (1.4) | .36 |
| Parathyroid hormone (ng/L) | 656 (740) | 651 (547) | 657 (804) | .78 |
| Serum calcium (mg/dL) | 9.2 (0.8) | 8.89 (0.6) | 9.30 (0.8) | .62 |
| Serum phosphorus (mg/dL) | 5.81 (1.6) | 5.94 (1.6) | 5.76 (1.6) | .62 |
| Serum albumin (g/dL) | 3.84 (0.3) | 3.73 (0.3) | 3.89 (0.3) | .26 |
| Protein Catabolic Rate | 1.09 (0.4) | 1.04 (0.3) | 1.11 (0.4) | .65 |
| Treatment time (minutes) | 234 (19) | 229 (23) | 235 (16) | .86 |
| Blood flow (mL/minute) | 409 (89) | 434 (150) | 399 (40) | .95 |
| Dialysate flow (mL/minute) | 681 (102) | 638 (110) | 698 (96) | .08 |
Orthostatic BP as a continuous variable
The mean orthostatic systolic BP change was −7.30 (20) mmHg. This change was −12.7 (20) mmHg for participants with diabetes and +0.1 (18) mmHg for those without (p=.06, Figure 1). There were no differences in orthostatic systolic BP change between men and women (−4.7 [17] vs. −12.1 [24], p=.78), those taking vs. not taking α-blockers (−7.4 [16] vs. −7.1 [26] mmHg, p=.76), those taking vs. not taking CCB (−5.1 [19] vs. −10.8 [22] mmHg, p=.65), or those taking vs. not taking statins (−11.3 [20] vs. −5.1 [21] mmHg, p=.28).
Figure 1. Orthostatic Systolic Blood Pressure Change in Patients Without and With Diabetes.
shows a box-and-whisker plot of the orthostatic systolic blood pressure changes in patients without and with diabetes. These changes were +0.1 (18 mmHg) for participants without and −12.7 (20) mmHg for participants with diabetes (p=.06).
The mean orthostatic diastolic BP change was −1.73 (11) mmHg. This change was −3.9 (11) mmHg for participants with diabetes and 0.3 (12) mmHg for those without (p=.30). There were differences in orthostatic diastolic BP change for men vs. women (−3.8 [9.5] vs. +2.4 [14] mmHg, p=0.02), participants taking vs. not taking CCB (0 [9.4] vs. −4.7 [14] mmHg, p=0.05), and participants taking vs. not taking statins (−6.4 [12] vs. 0.8 [10] mmHg, p=0.04).
The mean post-HD BIS-determined ECV/weight was 0.24 (0.04) L/kg. The measurement for women (0.21 [0.03] L/kg) was lower than for men (0.26 [0.03] L/kg, p<0.0001). The measurements for participants with and without diabetes were 0.23 (0.04) and 0.25 (0.05) L/kg (p=.16). Figure 2 shows a plot of orthostatic systolic BP change with BIS-determined ECV/weight. The range of orthostatic systolic BP change narrowed as BIS-determined ECV/weight increased. In a model controlling for age, sex, and the presence of diabetes, there was no significant association between post-HD BIS-determined ECV/weight and orthostatic systolic BP change (β=8.2 [16], p=.62, Table 2). There was an association between diabetes and orthostatic systolic BP change (β= −12.0 [5.3], p=.03). In a model that also controlled for ultrafiltration rate as well as use of α-blockers, CCB or statins, there remained no association between BIS-determined ECV/weight and orthostatic systolic BP change. There was no association between BIS-determined ECV/weight and orthostatic diastolic BP change in either Model 1 or Model 2 (data not shown).
Figure 2. Plot of Post Hemodialysis Extracellular Volume/Weight vs. Orthostatic Systolic Blood Pressure Change.
shows a plot of post hemodialysis extracellular volume/weight (L/kg) measured with bioimpedance spectroscopy compared to the post-hemodialysis orthostatic systolic blood pressure change. Orthostatic hypotension occurred across most of the spectrum of measured ECV/weight, although the overall range of orthostatic systolic BP change became narrower as ECV/weight increased.
Table 2.
Generalized linear models showing associations between predictor variables with the outcome variable orthostatic systolic blood pressure change
| Model 1 | Model 2 | |||
|---|---|---|---|---|
| Estimate (standard error) | p-value | Estimate (standard error) | p-value | |
| Post-hemodialysis extracellular volume/weight (L/kg) | 8.21 (16) | .62 | 16.5 (18) | .41 |
| Sex (male reference) | 7.44 (6.2) | .24 | 8.98 (6.5) | .17 |
| Age (years) | −0.14 (0.2) | .54 | −0.17 (0.3) | .49 |
| Presence of diabetes mellitus | −12.0 (5.3) | .03 | −11.3 (6.0) | .07 |
| Ultrafiltration rate (mL/kg/hr) | 0.71 (0.9) | .41 | ||
| Use of alpha adrenergic receptor blockers | −3.49 (5.7) | .54 | ||
| Use of calcium channel blocker | 5.54 (6.0) | .36 | ||
| Use of a statin | 1.24 (6.2) | .84 | ||
Orthostatic hypotension and ECV/weight
There were 16 participants with and 41 participants without OH. A comparison of these groups is in Table 1. Participants with OH were less likely to be receiving CCB (p=.02), but there was no difference in α-blocker use (p=.62). Fifty percent of the OH participants were women, and 29% of the group without OH were women (p=.14). Diabetes was present in 75% and 51% of the groups with and without OH, respectively (p=.11). The measurements for BIS-determined ECV/weight were 0.22 (0.04) and 0.24 (0.05) L/kg in the groups with and without OH (p=.09, Figure 3). Participants with OH had lower percentage of interdialytic weight gain prior to the treatments than the non-OH group [2.21 (1.1) vs. 3.15 (1.8), p=.04], but the actual volume removed [2.34 (1.1) vs. 2.83 (1.2) L, p=.17] and ultrafiltration rate [6.7 (2.7) vs 8.2 (3.6) mL/kg/hr, p=.15] were not different.
Figure 3: Comparison of Post Hemodialysis Extracellular Volume (L/kg) in Patients With and Without Orthostatic Hypotension.
shows a box-and-whisker plot of the post-HD ECV/weight in participants with and without OH. The measurements for ECV/weight were 0.22 (0.04) and 0.24 (0.05) L/kg in the groups with and without OH (p=.09).
In our logistic regression analyses in either Model 1 or Model 2 (Table 3), there was no significant association between post-HD BIS-determined ECV/weight with OH (OR 11, 4.04; 95% CI .2–671, .03–530 in the 2 models). Being on a CCB was associated with a lower risk of OH. We also conducted sensitivity analyses where we used a definition of OH to be the systolic BP decrease of at least 30 mmHg or a diastolic BP decrease of at least 10 mmHg. The BIS-determined ECV/weight was lower in the OH group (0.22 [0.03], n=14) compared to the non-OH group (0.25 [0.04], n=43; p=0.003). However, there was a larger percentage in the OH group that were women (57 vs. 28%, p=0.04) or had diabetes (86 vs. 49%, p=0.01). In the logistic regression analyses, BIS-determined ECV/weight was not associated with OH in Model 1 (OR=46, 95% CI .41–999) or Model 2 (OR=20.6, 95% CI .06–999).
Table 3.
Logistic regression analysis using orthostatic hypotension as the outcome variable
| Model 1 | Model 2 | |||
|---|---|---|---|---|
| OR | 95% CI | OR | 95% CI | |
| Post-hemodialysis extracellular volume/weight (L/kg) | 11 | 0.2–671 | 4.04 | 0.03–530 |
| Sex (male reference) | 0.66 | 0.2–2.6 | 0.45 | 0.09–2.3 |
| Age (years) | 1.03 | 0.97–1.09 | 1.03 | 0.96–1.1 |
| Presence of diabetes mellitus | 2.43 | 0.64–9.2 | 1.18 | 0.35–4.0 |
| Ultrafiltration Rate (mL/kg/hr) | 0.99 | 0.79–1.3 | ||
| Use of alpha adrenergic receptor blocker | 0.74 | 0.2–3.1 | ||
| Use of calcium channel blocker | 0.22 | 0.05–0.99 | ||
| On statin | 3.9 | 0.84–17 | ||
Orthostatic hypotension and other BP variables
The pre-HD seated systolic BP were similar between the groups (Table 4). The final seated systolic BP (prior to blood return) were not different, but the post-HD seated systolic BP (after blood return and completion of HD) were higher among participants with OH (p=.004). The standing post-HD systolic BP were not different, but the orthostatic systolic and diastolic BP changes were −28.4 (19) and −11.4 (12) mmHg in the OH group and 0.93 (13) and 1.8 (9) mmHg in those without OH (p=0.02 and p=0.03, respectively). A depiction of the pre-HD seated, final HD, post-HD seated, and post-HD standing systolic BP measurements are in Figure 4. Although the delta seated systolic BP from pre to post-HD was different between the two groups (there was an increase in this metric in the OH group), the intradialytic BP slope accounting for all measurements during HD was similar in both groups (Table 4). Ambulatory systolic BP during the following interdialytic period was similar in those with and without OH [142 (12) vs. 141 (14) mmHg, p=.86].
Table 4.
Blood Pressure Characteristics
| Reported as mean (standard deviation) | Total cohort (n=57) | With orthostatic hypotension (n=16) | Without hypotension (n=41) | p-value (comparison of with and without orthostatic hypotension) |
|---|---|---|---|---|
| Pre-hemodialysis seated systolic BP (mmHg) | 153 (20) | 151 (19) | 154 (20) | .57 |
| Intradialytic systolic BP slope (mmHg/min) | −0.07 (0.1) | −0.05 (0.09) | −0.08 (0.1) | .07 |
| Low intradialytic systolic BP (mmHg) | 113 (19) | 115 (21) | 112 (19) | .51 |
| Final seated systolic BP (mmHg) | 136 (25) | 142 (16) | 134 (22) | .44 |
| Post-HD seated systolic BP (mmHg) | 139 (23) | 153 (15) | 133 (23) | 0.004 |
| Post-HD standing Systolic BP (mmHg) | 132 (21) | 125 (17) | 134 (22) | .13 |
| Pre-HD seated diastolic BP (mmHg) | 85.3 (14) | 81.5 (14) | 87 (13) | .30 |
| Last seated diastolic BP (mmHg) | 78.2 (14) | 78.0 (13) | 78.3 (15) | .91 |
| Post-HD seated diastolic BP (mmHg) | 76.6 (14) | 79.9 (12) | 75.4 (14) | .26 |
| Post-HD standing diastolic BP (mmHg) | 75.1 (14) | 69.44 (11) | 77.2 (14) | .07 |
| Orthostatic systolic BP change (mmHg) | −7.30 (20) | −28.4 (19) | 0.93 (13) | 0.02 |
| Orthostatic diastolic BP change (mmHg) | −1.73 (11) | −11.4 (12) | 1.80 (9) | 0.03 |
| Ambulatory systolic BP (mmHg) | 142 (13) | 142 (12) | 141 (14) | .86 |
| Delta seated systolic BP (post seated – pre seated) | −14.5 (31) | 2.19 (27) | −21.2 (30) | .008 |
| HD-Hemodialysis, BP-Blood pressure | ||||
Figure 4. Comparison of Seated Systolic Blood Pressure (Pre-HD, Final HD, Post-HD) and Standing Post-HD Blood Pressure in Patients With and Without Orthostatic Hypotension.
shows the progression of systolic BP from pre-HD seated to last HD BP to post-HD seated BP to post-HD standing BP through box and whisker plots. The OH group is characterized by a decrease in systolic BP from pre-HD to the final measurement (−9.5 [25] mmHg), followed by an increase from final to post-HD seated (+11.4 [14] mmHg), followed by a decrease upon standing (−28.4 [19] mmHg). The non-OH group is characterized a by decrease in systolic BP from pre-HD to the final measurement (−20.5 [30] mmHg). This was followed by a small decrease from final to post-HD seated and then a small increase from post-HD seated to standing (−0.88 [12] and +0.93 [13], respectively).
DISCUSSION
The principal new finding from this study is that there was no independent association between orthostatic BP change following completion of HD and post-HD BIS-determined ECV in hypertensive HD patients. No association existed when these variables were evaluated as continuous variables or when clinically defined OH was considered as the outcome. The BIS-determined ECV/weight was lower among participants who met a more extreme definition of OH (systolic BP decrease of at least 30 mmHg OR diastolic BP decrease of at least 10 mmHg), but this association appears to have been confounded by other variables including sex and diabetes. The clinical relevance of these findings is that evaluation of post-HD orthostatic BP change may be an unreliable method to screen for ECV depletion in HD patients.
In general, orthostatic BP changes occur due to reduction in venous return and cardiac output related to vascular pooling in the lower extremities upon position change. The appropriate physiologic response to standing would be an increase in sympathetic tone to maintain BP through increased cardiac output and vascular resistance. Orthostatic hypotension is defined as a decrease in systolic BP ≥20 mmHg or in diastolic BP ≥10 mmHg upon standing (7). Factors contributing to exaggerated decreases in BP that are commonly encountered by HD patients include advanced age, intravascular volume depletion, use of vasodilating antihypertensive drugs, and autonomic neuropathy often due to diabetes mellitus(7).
Despite having numerous risk factors for OH, there is not extensive literature on the topic in HD patients. One study found that large post-HD orthostatic BP decreases were associated with increased mortality over two years(8). Another study found that OH occurring during the hospitalization of HD initiation was associated with increased mortality(9). There is also evidence that orthostatic BP changes (on non-HD days) are associated with impaired cognitive function(10). These studies highlight the prognostic significance of OH, but do not address any specific adjustments to therapy that should be considered. One recent analysis of DOPPS data showed that more than half of dialysis units use orthostatic BP measurements to assess dry weight(11). This study did not specify how such information was utilized to make clinical decisions, but the general practice was associated with decreased hospitalizations and cardiovascular events with no mortality benefit. However, no study to our knowledge had used an objective assessment of ECV to evaluate its role in orthostatic BP changes.
We analyzed associations between post-HD BIS-determined ECV/weight with post-HD orthostatic BP changes while considering the latter as both a continuous variable and categorical variable. There was no association with BIS-determined ECV/weight and orthostatic BP (systolic or diastolic) as a continuous variable. We did find that diabetes was associated with a larger decrease in BP upon standing. Using the formal definition of OH, there was only a trend for an association between BIS-determined ECV/weight with OH. Using a more exclusive definition (where systolic BP decreased by at least 30 mmHg), there was a significant difference between groups in BIS-determined ECV/weight. However, there were more participants with OH than without who were either female or had diabetes. Similarly, BIS-determined ECV/weight was lower in women than men with a trend for it to be lower among participants with diabetes. With both OH definitions, there were not any independent associations with BIS-determined ECV/weight when controlling for these other variables. The associations between orthostatic BP change (and OH) with diabetes are consistent with effects of autonomic neuropathy. Consequently, the finding of OH after dialysis should prompt consideration of this volume-independent phenomenon among patients with diabetes.
We did not find any associations between orthostatic systolic BP as a continuous variable and use of vasodilating antihypertensives such as α-blockers and CCB. However, CCB were associated with a lower risk of OH. We suspect these findings might be related to indication bias such that patients previously intolerant to CCB blockers or more prone to orthostatic BP changes in general would be less likely to be prescribed them. We suspect the lack of an association with α-blockers might be in part due to the fact that we considered carvedilol to be an α-blocker. Most of the α-blocker use in our cohort was in fact attributed to carvedilol. Although the degree of beta blockade is far greater in this drug, its alpha blocking properties cannot be ignored.
Because BIS is not considered a gold standard method to measure ECV, we considered alternate metrics that have previously been linked to ECV overload in HD patients. Elevated ambulatory BP, post HD-BP and positive intradialytic BP slopes have been associated with ECV overload based on BP response to dry weight probing over time(12, 13). In our study, ambulatory BP and intradialytic BP slopes were similar in patients with and without OH. The delta systolic BP was positive in the OH group, a finding that has previously been consistent with ECV overload(2, 3). Altogether, these findings make it less likely that a large difference in ECV would have been found even if gold standard methodology such as dilution studies had performed.
An additional finding of interest was that change in BP from the final seated BP to the post-HD seated BP differed between the OH and non-OH groups. Both groups experienced reductions in BP during dialysis shown by both negative intradialytic BP slopes and overall decreases from pre-HD to the final measurement during HD. However, the OH group experienced an increase in BP from the final measurement to the post-HD seated measurements. One interpretation of this is that the OH group had BP that was particularly fluid responsive following blood return. This highlights that caution should be implemented in assessing ECV based on such a BP response. While there might have been differences in intravascular volume at that point in time, this does not necessarily equate to differences in overall extracellular volume. We did not have any measurements of intravascular volume during the treatments to determine this. Technology is available for relative blood volume monitoring (RBV) during dialysis. Flat curves using RBV are associated with increased mortality(14), but prospective studies have shown that clinical use of RBV monitoring is associated with increased hospitalization and mortality(15) possibly due to the occurrence of volume-independent intradialytic BP changes(16, 17).
Our lack of an association between OH and ECV assessed with BIS does not indicate that orthostatic vitals should be ignored. As mentioned above, the presence of OH might indicate intravascular volume depletion or it may reflect vasodilation related to autonomic neuropathy that, similar to intradialytic hypotension, can create challenges to achieving adequate ultrafiltration over time. The balance between maintaining euvolemia and avoiding hypoperfusion through ultrafiltration is a paramount challenge in the management of HD patients. Safe implementation of dialysis should take the highest priority. Because rapid ultrafiltration rate and short dialysis time are themselves independently associated with mortality(18, 19), longer and slower HD treatments if not additional treatments might be appropriate to achieve euvolemia in patients with OH after HD just as in patients with ECV overload who experience intradialytic hypotension. However, our findings do not support the notion that the OH patients must have their dry weight increased.
Limitations include a relatively small sample size, particularly of subjects with clinically defined OH. Our sensitivity analysis expanded the definition to include a minimum decrease of 30 mmHg, and there was still not an independent association between OH and BIS-determined ECV/weight. There may be a severe threshold of OH beyond this definition where independent differences exist in ECV/weight that would be identified with a larger sample size. There was also not a pre-specified protocol for the timing of standing BP measurements due the retrospective nature of data collection for the purpose of this analysis, but these are typically measured within a few minutes of the post-HD seated BP measurements which are routinely measured right after blood return. While our findings may not be generalizable to HD patients without hypertension, prior large studies show that the mean pre-HD systolic BP is well above our inclusion criteria (140 mmHg) with one study even showed that nearly 75% of HD patients had BP exceeding this this threshold(20–22).
In conclusion, we found no consistent association between post-HD orthostatic BP changes and objective assessments of ECV in hypertensive HD patients shortly after the HD treatment. Even the binary outcome of OH was not associated with BIS-determined ECV/weight when considering other clinically relevant variables. We cannot exclude that intravascular volume depletion from excessively rapid ultrafiltration rates could not have been present in some patients despite persistent ECV excess. However, our findings indicate that post-HD orthostatic systolic BP changes obtained per routine clinical is not a reliable reflection of ECV and that chronic ECV overload can exist even in the context of significant orthostatic BP decreases. The assumption these orthostatic BP decreases immediately after HD reflect the need to increase a patient’s dry weight might result in patients becoming unnecessarily ECV overloaded.
Acknowledgements
Preliminary findings from this research was presented in part by Prince Aryeetey, MD (former UT Southwestern nephrology fellow) as a poster presentation at the American Society of Nephrology Kidney Week Meeting in Washington D.C. in November 2019.
All authors are employees of the Department of Veterans Affairs. The content of this publication do not reflect any official policy or position of the Department of Veterans Affairs.
Funding:
Drs. Van Buren and Jeon-Slaughter receive current funding from Veterans Affairs Merit Award 1 I01 CX002009-01 to support this research. Prior funding for this research came from NIH 1K23DK096007-01A1 and the University of Texas Southwestern O’Brien Kidney Research Core (National Institutes of Health [NIH] grant P30DK079328).
Footnotes
The authors have no conflicts of interest
REFERENCES
- 1.Zoccali C, Moissl U, Chazot C, Mallamaci F, Tripepi G, Arkossy O, et al. : Chronic Fluid Overload and Mortality in ESRD. J Am Soc Nephrol, 28, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Nongnuch A, Campbell N, Stern E, El-Kateb S, Fuentes L, Davenport A: Increased postdialysis systolic blood pressure is associated with extracellular overhydration in hemodialysis outpatients. Kidney Int, 87: 452–457, 2015. 10.1038/ki.2014.276 [DOI] [PubMed] [Google Scholar]
- 3.Van Buren P, Zhou Y, Neyra J, Xiao G, Vongpatanasin W, Inrig J, et al. : Extracellular Volume Overload and Increased Vasoconstriction in Patients With Recurrent Intradialytic Hypertension. Kidney Blood Press Res, 41: 802–814, 2016. 10.1159/000450565 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Liu H, Lu R, Shastri S, Sonderman M, Van Buren P: Assessing Extracellular Volume in Hemodiaylsis Patients Using Intradialytic Blood Pressure Slopes. Nephron Clinical Practice, 2018. 10.1159/000487093 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Flythe J, Xue H, Lynch K, Curhan G, Brunelli S: Association of mortality risk with various definintions of intradialytic hypotension. Journal of the American Society of Nephrology, 2015: 724–734, 2015. 10.1681/ASN.2014020222 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.With Intradialytic Hypertension. In: ClinicalTrials.gov [Internet]. Bethesda (MD): National Library of Medicine (US). 2000-[cited January 16, 2016] Available from https://www.clinicaltrials.gov/ct2/show/NCT01862497?term=van+buren&rank=2; NLM identifier NCT01862497 NIoDaDaKDNTUoTSMCaDMaToIABPiP,, [Google Scholar]
- 7.Freeman R, Wieling W, Axelrod F, Benditt D, Benarroch E, Biaggioni I, et al. : Consensus statement on the definition of orthostatic hypotension, neurally mediated syncope and the postural tachycardia syndrome. Autonomic Neuroscience: Basic and Clinical, 161: 46–48, 2011 [DOI] [PubMed] [Google Scholar]
- 8.Shoji T, Tsubakihara Y, Fujii M, Imai E: Hemodialysis-associated hypotension as an independent risk factor for two-year mortality in hemodialysis patients. Kidney International, 66: 1212–1220, 2004 [DOI] [PubMed] [Google Scholar]
- 9.Sasaki O, Nakahama H, Nakamura S, Yoshihara F, Inenaga T, Yoshii M, et al. : Orthostatic hypotension at the introductory phase of haemodialysis predicts all-cause mortality. Nephrol Dial Transplant, 20: 377–381, 2005 [DOI] [PubMed] [Google Scholar]
- 10.Liu W, Wang L, Huang X, Yuan C, Li H, Yang J: Orthostatic blood pressure reduction as a possible explanation for memory deficits in dialysis patients. Hypertension Research 42: 1049–1056, 2019 [DOI] [PubMed] [Google Scholar]
- 11.Dasgupta I, Thomas G, Clarke J, Sitch A, Martin J, Bieber B, et al. : Associations between Hemodialysis Facility Practices to Manage Fluid Volume and Intradialytic Hypotension and Patient Outcomes. Clin J Am Soc Nephrol, 14: 385–393, 2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Agarwal R, Alborzi P, Satyan S, Light R: Dry-Weight Reduction in Hypertensive Hemodialysis Patients (DRIP): A Randomized, Controlled Trial. Hypertension, 53: 500–507, 2009. 10.1161/HYPERTENSIONAHA.108.125674 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Agarwal R, Light R: Intradialytic hypertension is a marker of volume excess. Nephrol Dial Transplant, 25: 3355–3361, 2010. 10.1093/ndt/gfq210 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Preciado P, Zhang H, Thijssen S, Kooman J, van der Sande F, Kotanko P: All-cause mortality in relation to changes in relative blood volume during hemodialysis. Nephrol Dial Transplant (in press), 2018. 10.1093/ndt/gfy286 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Reddan D, Szczech L, Hasselbad V, Lowrie E, Lindsay R, Himmelfarb J, et al. : Intradialytic blood volume monitoring in ambulatory hemodialysis patients: a randomized trial. Journal of the American Society of Nephrology, 16: 2162–2169, 2005. 10.1681/ASN.2004121053 [DOI] [PubMed] [Google Scholar]
- 16.Levin N, de Abreu M, Borges L, Tavares Filho H, Sarwar R, Gupta S, et al. : Hemodynamic response to fluid removal during hemodialysis: categorization of causes of intradialytic hypotension. Nephrol Dial Transplant (in press), 2018 [DOI] [PubMed] [Google Scholar]
- 17.Hussein W, Arramreddy R, Sun S, Doss-McQuitty S, SChiller B: Blood Volume Monitoring to Assist Fluid Management in Hemodialysis Patients. American Journal Of Kidney Diseases, 67: 166–168, 2016 [DOI] [PubMed] [Google Scholar]
- 18.Flythe J, Curhan G, Brunelli S: Disentangling the ultrafiltration rate-mortality association: the respective roles of session length and weight gain. Clin J Am Soc Nephrol, 8: 1151–1161, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Flythe J, Kimmel S, Brunelli S: Rapid fluid removal during dialysis is associated with cardiovascular morbidity and mortality. Kidney International, 79: 250–257, 2011. 10.1038/ki.2010.383 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Agarwal R, Nissenson A, Batlle D, Coyne D, Trout J, Warnock D: Prevalence, Treatment, and Control of Hypertension in Chronic Hemodialysis Patients in the United States. American Journal of Medicine, 115: 291–297, 2003 [DOI] [PubMed] [Google Scholar]
- 21.Zhang H, Preciado P, Wang Y, Meyring-Wosten A, Raimann J, Kooman J, et al. : Association of all-cause mortality with pre-dialysis systolic blood pressure and its peridialytic change in chronic hemodialysis patients. Nephrol Dial Transplant, 35: 1602–1608, 2020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Park J, Rhee C, Sim J, KIm Y, Ricks J, Streja E, et al. : A comparative effectiveness research study of the change in blood pressure during hemodialysis treatment and survival. Kidney Int, 84: 795–802, 2013. 10.1038/ki.2013.237 [DOI] [PMC free article] [PubMed] [Google Scholar]