Acute fasting reduces tolerance to progressive central hypovolemia in humans

Abstract

Potential health benefits of an acute fast include reductions in blood pressure and increases in vagal cardiac control. These purported health benefits could put fasted humans at risk for cardiovascular collapse when exposed to central hypovolemia. The purpose of this study was to test the hypothesis that an acute 24-h fast (vs. 3-h postprandial) would reduce tolerance to central hypovolemia induced via lower body negative pressure (LBNP). We measured blood ketones (β-OHB) to confirm a successful fast (n = 18). We recorded the electrocardiogram (ECG), beat-to-beat arterial pressure, muscle sympathetic nerve activity (MSNA; n = 7), middle cerebral artery blood velocity (MCAv), and forearm blood flow. Following a 5-min baseline, LBNP was increased by 15 mmHg until –60 mmHg and then increased by 10 mmHg in a stepwise manner until onset of presyncope. Each LBNP stage lasted 5-min. Data are expressed as means ± SE β-OHB increased (β-OHB; 0.12 ± 0.04 fed vs. 0.47 ± 0.11, P < 0.01 mmol/L fast). Tolerance to central hypovolemia was decreased by ∼10% in the fasted condition measured via total duration of negative pressure (1,370 $\pm$ 89 fed vs. 1,229 ± 94 s fast, P = 0.04), and was negatively associated with fasting blood ketones (R = –0.542, P = 0.02). During LBNP, heart rate and MSNA increased similarly, but in the fasted condition forearm vascular resistance was significantly reduced. Our results suggest that acute fasting reduces tolerance to central hypovolemia by blunting increases in peripheral resistance, indicating that prolonged fasting may hinder an individual’s ability to compensate to a loss of blood volume.

NEW & NOTEWORTHY An acute 24 h fasting reduces tolerance to central hypovolemia, and tolerance is negatively associated with blood ketone levels. Compared with a fed condition (3-h postprandial), fasted participants exhibited blunted peripheral vasoconstriction and greater reductions in stroke volume during stepwise lower body negative pressure. These findings suggest that a prolonged fast may lead to quicker decompensation during central hypovolemia.

INTRODUCTION

Lower body negative pressure (LBNP) is a laboratory model to study hemorrhage and has been applied extensively to investigate compensatory mechanisms responsible for maintaining perfusion pressure during progressive central hypovolemia (1, 2). Previous studies have reported that factors such as sex, age, hydration status, and environmental conditions alter individual tolerances to LBNP-induced central hypovolemia (3–6). Reduced tolerance to hypovolemia might also be related to caloric restriction, but the influence of energy balance on tolerance to progressive central hypovolemia is unknown.

Fasting quickly induces a negative energy balance leading to ketogenesis once glycogen stores are depleted (7). Intermittent fasting is a growing dieting concept and may convey other health benefits such as reductions of blood pressure and increases of vagal-cardiac control (8–10). However, these health benefits mediated by enhanced parasympathetic activation might challenge compensatory maintenance of adequate organ perfusion during hypovolemia. To maintain adequate organ perfusion pressure during progressive central hypovolemia, the sympathetic nervous system (SNS) coordinates increases in peripheral resistance, cardiac contractility, and heart rate. Moreover, the unloading of cardiopulmonary and arterial baroreceptors during central hypovolemia stimulates sympathetic neural activity, the release of vasoactive hormones (i.e., epinephrine and norepinephrine), and volume-regulating hormones (i.e., vasopressin and renin; 11, 12). In rats, fasting for 48-h significantly reduces SNS activity measured directly via norepinephrine turnover at the heart (13). Suppression of the SNS may hinder the effectiveness of the acute compensatory mechanisms during hemorrhagic insult and reduce tolerance to central hypovolemia. It is difficult to assess SNS-mediated compensatory mechanisms during actual experimental hemorrhage in humans, but it is not difficult to simulate hemorrhage by inducing progressive central hypovolemia with lower body negative pressure (1).

The only LBNP study conducted in fasted humans was performed by Bennet et al. (14), and research subjects were not taken to the point of tolerance (presyncope). In their study, systolic pressure and vasoconstriction decreased and heart rate and forearm blood flow increased to a greater extent during moderate LBNP after a 48-h fast compared with a 12-h fast (14). Tilt-table studies have demonstrated that participants who have fasted for 24-, 48-, and 72-h exhibit greater increases in heart rate and reductions in metrics of heart rate variability (HRV) when exposed to brief head-up tilt compared with the fed state (15, 16). Tilt-table and LBNP studies seem to provide converging evidence suggesting that fasting may reduce tolerance to an orthostatic challenge by potentially reducing vasoconstriction and altering cardiovascular autonomic control. However, direct evidence of a reduced tolerance to central hypovolemia with an acute fast has not been established. Therefore, the purpose of this study was to test the hypothesis that fasting for 24-h reduces tolerance to progressive central hypovolemia in conjunction with greater reductions of heart rate variability and blunted peripheral vasoconstriction during progressive LBNP.

METHODS

Participants

Eighteen healthy young adults (7 F, age of 23 ± 0.7, BMI 25 ± 0.7 kg/m²) participated. The participants in the current study are a subset of a larger sample population of a previously published work (8). To mitigate the potential effect on vascular volume and autonomic control, participants were asked to refrain from exercise and stimulants such as caffeine, cold medications that might alter autonomic function (e.g., those containing diphenhydramine), and alcohol 24 h before autonomic testing. Because of the potential influence of ovarian hormones on autonomic function (17), all female participants were tested in the early to mid-late follicular phase of their menstrual cycles (days 3–8). Participants visited the laboratory for a familiarization session before the first day of experimentation. Procedures and equipment were explained, and they had the opportunity to ask questions. Participants then read and signed an informed consent document that had been approved by the Committee for the Protection of Human Subjects in Research at Michigan Technological University (MTU IRB: M1785).

Measurements

Blood biomarkers and hydration status.

Upon arrival at the laboratory for autonomic testing, blood glucose, ketones, and lipids were measured. Whole blood was utilized to measure total cholesterol, high-density lipoproteins, low-density lipoproteins, triglycerides, and glucose using a Cholestech LDX analyzer (Alere Cholestech, San Diego, CA). Whole blood was also utilized to measure blood ketones (β-OHB) using a Precision Xtra (Abbott Laboratories, MediSense Products Inc., Bedford, MA) ketone monitor. Urine was also collected to assess hydration status via urine-specific gravity using a PALS-10S urine refractometer (Atago, Tokyo, Japan).

Blood pressure and heart rate.

Beat-to-beat arterial blood pressure was recorded continuously throughout all time points of the autonomic function test using a NOVA Finometer (Finapres Medical Systems, Amsterdam, The Netherlands). Heart rate was recorded via a three-lead electrocardiogram (ECG) (Finapres Medical Systems, Amsterdam, The Netherlands), and respiratory rate was continuously measured using a pneumobelt (Harvard Apparatus, Holliston, MA). Three brachial blood pressures (HEM-907XL, Omron, Kyoto, Japan) were used to calibrate finger plethysmography.

Microneurography.

Using a nerve traffic analysis system (Model 662 C-1, Dept. of Bioengineering, University of Iowa, Iowa City, IA), multifiber efferent sympathetic nerve traffic was recorded from peroneal nerve muscle fascicles at the popliteal fossa by inserting a tungsten microelectrode (Frederick Haer and Co., Bowdoinham, ME). A reference electrode was inserted subcutaneously 2–3 cm from the recording electrode. Both electrodes were connected to a differential preamplifier and then to an amplifier (total gain 80,000) where the nerve signal was band-pass filtered (700–2,000 Hz) and integrated (time constant, 0.1 s) to obtain a mean voltage display of nerve activity. Satisfactory recordings of MSNA were defined by spontaneous pulse-synchronous bursts that did not change during tactile or auditory stimulation and increased during end-expiratory apnea.

Transcranial Doppler ultrasound.

Cerebral blood velocity was recorded from the left middle cerebral artery (MCAv) using a 2-MHz Doppler probe (Viasys Healthcare, Conshohocken, PA), positioned at a constant angle over the temporal window, located above the zygomatic arch. The probe was placed in the approximated same position for both the fed and fasted conditions. Depth of insonation was recorded during the initial session for repeated measurements. Using cross-spectral analysis of mean arterial pressure (MAP) and MCAv, we calculated coherence within the low-frequency range (0.07–0.20 Hz) at baseline (18). Transfer function calculations were not performed, as a majority of coherence values were <0.5.

Venous occlusion plethysmography.

Limb blood flow was measured from the forearm using venous occlusion plethysmography (D.E. Hokanson, Bellevue, WA). A mercury-in-silastic strain gauge was placed around the maximal circumference of the forearm. A wrist cuff was inflated to 220 mmHg to arrest circulation to the hand. An arm cuff was inflated (∼70 mmHg) and deflated (0 mmHg) in 7–8 s intervals (15 s/cycle). This technique allows for occlusion of venous blood flow but still allows arterial blood flow. A strain gauge was used to measure circumference changes during these inflation/deflation cycles. Forearm blood flow was calculated by the rate of increase in forearm circumference over time during venous occlusion. Vascular resistance was calculated as mean arterial pressure divided by forearm blood flow. Vascular conductance was calculated (FBF/MAP) × 100 and is expressed as arbitrary units. Arbitrary units were used because they are quantitatively similar to the standard units for forearm blood flow (19). Forearm blood flow measurements were taken during the last 3-min of each 5-min LBNP stage and averaged. We allowed a 2-min recovery from hand ischemia between each LBNP stage. Venous occlusion plethysmography is a time-honored, safe, and noninvasive method for estimating limb blood flow (20). Due to limb movement artifact, forearm blood flow was not available in four individuals. Thus, we have an n = 14 for forearm blood flow. Unfortunately, due to differing levels of LBNP tolerance and movement artifact, we were unable to measure forearm blood flow at the −80 mmHg LBNP stage.

Experimental Design

This study was a randomized controlled crossover design with repeated measures. Each participant was taken to the point of presyncope twice, once in the fed condition (3-h postprandial) and again in the fasted condition (24-h postprandial). LBNP trials were performed ∼4 wk apart. A standardized caloric-controlled meal was provided for each condition. The caloric intake of the standardized meal was estimated to be one-third of the total caloric intake needed to maintain the participants’ weight. Macronutrient composition of the meal was estimated to be ∼55% carbohydrate, ∼20% fats, and 25% protein. To estimate total daily energy needs resting metabolic rate was estimated using the Mifflin-St. Jeor equation and multiplied by a physical activity factor (21, 22). The experimental design for the present study is part of a completed randomized control trial, where details for the fed and fasted conditions are described previously (8).

Upon arrival at the laboratory, blood and urine samples were collected and analyzed. Participants were then instrumented and positioned supine within an airtight chamber that was sealed at the iliac crest. The LBNP protocol consisted of a 5 min baseline period followed by a stepwise chamber decompression, which consisted of 5 min at −15, −30, −45, and −60 mmHg, and then additional increments of −10 mmHg every 5 min until the onset of presyncope. To generate negative pressure, a vacuum regulated by a variable autotransformer was manually adjusted until the desired pressure was achieved. Presyncope was identified in real time by two investigators (Gonzalez and Cooke), one of which was blinded to the condition (Cooke), by a precipitous fall in systolic pressure greater than 15 mmHg, persistent systolic pressure below 90 mmHg, sudden bradycardia, and/or voluntary subject termination due to discomfort from symptoms such as dizziness, nausea, or tunnel vision. We quantified tolerance by calculating the total duration of negative pressure (including stage transitions) and cumulative stress index (CSI; product of the negative pressure stage and time spent in that stage, mmHg × min) (23, 24).

Data Analysis

Data were sampled at 500 Hz (WINDAQ, Dataq Instruments, Akron, Ohio) and analyzed with specialized software (WINCPRS, Absolute Aliens, Turku, Finland). R waves generated from the ECG signal were automatically detected and marked. Each R wave was manually confirmed and used to determine the R-R interval (RRI). Time domain heart rate variability metrics were expressed using the percentage of R-R intervals that varied by 50 ms or more (pNN50) and the root mean squared of successive differences of RRI (RMSSD). Systolic and diastolic arterial pressures were marked from the Finometer tracings. Using the arterial pressure waveform as an input, stroke volume (SV), cardiac output, and total peripheral resistance were automatically estimated on a beat-to-beat basis using the pulse contour method (25). Spontaneous cardiovagal baroreflex sensitivity (cvBRS) was assessed via the sequence method (26). Specifically, baroreflex sensitivity was assessed by identifying sequences of three or more consecutively increasing systolic arterial pressures to three or more consecutive lengthening R-R intervals (up-up sequences) and systolic pressure that exhibited three or more decreases corresponding with three or more shortenings of R-R interval were identified (down-down sequences). Systolic arterial pressure (SAP) that changed by at least 1 mmHg/beat and R-R interval that changed by at least 4 ms were identified as a sequence. The minimum r value of >0.7 was used as criteria for accepting a sequence. Maximum (systolic) and minimum (diastolic) blood flow velocities from the middle cerebral artery were marked from the Doppler tracings. Muscle sympathetic nerve bursts were automatically detected based on their amplitude and a 1.3 s expected burst peak latency from the previous R wave. All automated detection results were checked manually. Sympathetic bursts of activity were expressed as burst frequency (bursts/min), burst incidence (bursts/100 heartbeats), and total MSNA (i.e., the sum of the normalized burst areas per minute). Baseline data were averaged over the entire 5-min. For each LBNP stage, the last 60 s of data were averaged, and the last 30 s leading to presyncope was averaged for the presyncope time point. Due to variability in the time to decompensation between conditions, we normalized time to decompensation for relative comparisons of variables of interest between conditions. For example, if a participant experienced decompensation during the −60 mmHg stage, −60 mmHg was designated as 100% tolerance. LBNP stages were then designated as percent to presyncope bins (100, 80, 40, and 0). If a participant decompensated before forearm blood flow data was obtained, the latter stage was designated as 100% tolerance. In addition, we examined the magnitude of changes in variables of interest from baseline for each LBNP stage by condition (LBNP stage × condition) and by condition by percent to presyncope (percent to presyncope × condition). We utilized a mixed effects model to account for missing data points because of variable tolerance to LBNP (GraphPad Prism 10, Graphpad Software Inc, La Jolla, CA). If a significant interaction (LBNP stage × condition) or (percent to presyncope × condition) was detected, we performed post hoc multiple comparisons using Šídák’s multiple comparisons tests. Normality was assessed using a Shapiro–Wilk test. If data were normally distributed, dependent variables were assessed with a paired t test. If data were not normally distributed a Wilcoxon matched pairs signed rank test was used. Biomarkers were analyzed via two-tailed statistical tests, while one-tailed statistical tests were utilized for directional hypotheses related to LBNP tolerance. Data are presented as means ± SE. A probability value of ≤ 0.05 was considered statistically significant.

RESULTS

Participants were weight-stable and similarly hydrated between conditions. Participant characteristics and blood biomarkers are displayed in Table 1.

Table 1.

Participant blood biomarkers in the fed and fasted condition

Variable	Fed	Fasted	t Test P Value	Wilcoxon P Value
Weight, kg	77.3 ± 3.6	76.9 ± 3.4	0.271
Urine specific gravity, A.U. (n = 16)	1.017 ± 0.002	1.012 ± 0.002		0.126
Total cholesterol, mg/dL	174.0 ± 8.93	172.9 ± 7.3	0.814
Low-density lipoprotein, mg/dL	100.9 ± 7.9	107.2 ± 6.6	0.342
High-density lipoprotein, mg/dL	52.7 ± 4.5	54.4 ± 4.2	0.398
Triglycerides, mg/dL	130.8 ± 15.8	78.4 ± 9.2		0.0154
Glucose, mg/dL	100.2 ± 2.9	81 ± 1.9	<0.001
Ketones, mmol/L (β-hydroxybutyrate)	0.122 ± 0.04	0.472 ± 0.11		<0.001

Values are means ± SE. n = 18, unless specified in table; paired t tests were used to compare variables between the fed and fasted conditions. Wilcoxon matched-pairs signed rank test was used to compare variables not normally distributed. Two-tailed P values are displayed.

Previous communications have reported baseline autonomic and cardiovascular variables (8). The primary outcome measure of LBNP tolerance was reduced in the fasted condition, measured via total duration of negative pressure (1,370 ± 89 fed vs. 1,229 ± 94 s fasted; P = 0.04, Fig. 1A). Cumulative stress index (833 ± 87 fed vs. 725 ± 89 mmHg min fasted; P = 0.08) was not significantly different between conditions (Fig. 1B). In the fed condition, 16 of the 18 LBNP trials were investigator initiated due to low systolic blood pressure and in the fasted condition 17 of the 18 trials were investigator initiated due to low systolic blood pressure. Otherwise, LBNP termination was initiated by the participant. We detected a significant inverse correlation between ketone levels in the fasted condition and LBNP tolerance measured via duration of negative pressure (r = −0.542, P = 0.02, Fig. 2A). In addition, we detected a significant inverse correlation between the change in blood ketone levels between the fed and fasted condition and change in maximal forearm vascular resistance (FVR) at the point of presyncope between the fed and fasted condition (r = −0.639, P = 0.02, Fig. 3).

Figure 1.

LBNP tolerance in the fed and fasted condition. Duration of negative pressure (A) and cumulative stress index (B) represented as boxplots. The line in the boxplots represents the median and the box represents the interquartile rang (IQR; the difference between the 25th and 75th percentile). The whiskers extend from the upper and lower edge of the box to the highest and lowest values. Circles, male participants, triangles, female participants; solid symbols, fed condition; open symbols, fasted condition. *P <0.05.

Figure 2.

The relationship between LBNP tolerance and fasting ketone levels and glucose levels. A: bivariate correlations were used to assess the relationship between fasted ketone levels and tolerance to LBNP measured via duration of negative pressure. B: in addition, bivariate correlations were used to assess the relationship between fasted glucose levels and tolerance to LBNP measured via duration of negative pressure. ●, male participants; ▴, female participants. LBNP, lower body negative pressure.

Figure 3.

The relationship between changes in ketone levels between the fed and fasted condition and changes in forearm vascular resistance during maximal LBNP. Bivariate correlations were used to assess the relationship between changes in ketone levels between the fed and fasted conditions and changes in maximal forearm vascular resistance during the fed and fasted conditions. ●, male participants; ▴, female participants. LBNP, lower body negative pressure.

There was no significant condition by percent to presyncope interactions detected for heart rate, heart rate variability, blood pressure, pulse pressure, cardiac output, respiratory rate, or muscle sympathetic nerve activity, as displayed in Table 2. A significant condition by percent to presyncope interaction was detected for forearm blood flow (P < 0.01), forearm vascular resistance (P < 0.01, Fig. 4A), and stroke volume (P = 0.02 Fig. 4C). Post hoc multiple comparisons revealed significantly lower forearm vascular resistance for the fasted condition at the point of presyncope (100%) and elevated stroke volume at baseline (0%) in the fasted condition. Although an interaction was identified for forearm blood flow and forearm vascular conductance, no significant differences were detected with post hoc multiple comparisons. To quantify reactivity to relative central hypovolemia between conditions, we calculated delta changes from baseline for variables of interest by percent to presyncope. In the fasted condition, individuals exhibited a lower magnitude of increase in forearm vascular resistance at 80% to presyncope and at the point of presyncope 100% compared with the fed condition.

Table 2.

Cardiovascular and sympathetic measurements by percent to presyncope

	0%		40%		80%		100%		P Values
	Fed	Fasted	Fed	Fasted	Fed	Fasted	Fed	Fasted	Interaction	Condition	%
Heart rate, beats/min	63 ± 2	59 ± 3	67 ± 2	63 ± 3	86 ± 3	83 ± 4	99 ± 3	94 ± 4	0.87	0.19	<0.01
RMSSD, ms	80 ± 9	86 ± 9	57 ± 7	69 ± 9	23 ± 3	27 ± 4	18 ± 3	30 ± 7	0.75	0.09	<0.01
PNN50, %	39 ± 4	38 ± 5	27 ± 5	33 ± 5	5 ± 2	9 ± 3	3 ± 1	5 ± 2	0.57	0.32	<0.01
SAP, mmHg	115 ± 2	114 ± 2	116 ± 3	113 ± 2	116 ± 3	113 ± 3	104 ± 3	101 ± 3	0.93	0.43	<0.01
DAP, mmHg	64 ± 2	61 ± 2	66 ± 3	64 ± 2	74 ± 2	72 ± 2	67 ± 3	65 ± 3	0.99	0.37	<0.01
MAP, mmHg	82 ± 2	82 ± 2	84 ± 3	84 ± 2	89 ± 2	88 ± 2	80 ± 3	79 ± 3	0.97	0.57	<0.01
cvBRS up-up, ms/mmHg (n = 16)	25 ± 2	25 ± 3	22 ± 2	22 ± 3	11 ± 2	13 ± 2	7 ± 1	8 ± 2	0.93	0.68	<0.01
cvBRS down-down, ms/mmHg	20 ± 2	22 ± 3	16 ± 2	17 ± 2	8 ± 1	9 ± 1	5 ± 1	7 ± 1	0.83	0.27	<0.01
Pulse pressure, mmHg	51 ± 3	52 ± 2	50 ± 2	48 ± 2	42 ± 2	41 ± 2	37 ± 2	36 ± 2	0.87	0.69	<0.01
Q, L/min	5.5 ± 0.3	5.7 ± 0.3	5.1 ± 0.3	5.2 ± 0.3	4.9 ± 0.3	4.6 ± 0.2	4.9 ± 0.2	4.9 ± 0.3	0.26	0.81	<0.01
Respiratory rate, breaths/min	15 ± 0.8	15 ± 1	15 ± 0.8	14 ± 1	16 ± 0.7	15 ± 0.7	16 ± 1	17 ± 1	0.24	0.92	0.05
MSNA burst frequency, bursts/min (n = 7)	16 ± 3	15 ± 3	26 ± 6	22 ± 5	43 ± 5	41 ± 7	44 ± 6	41 ± 8	0.85	0.57	<0.01
MSNA burst incidence, bursts/100 hb (n = 7)	26 ± 4	29 ± 6	39 ± 7	36 ± 7	49 ± 4	48 ± 7	44 ± 4	43 ± 7	0.65	0.80	<0.01
MSNA total, arb units (n = 7)	0.7 ± 0.1	1.5 ± 0.5	3.0 ± 1.4	2.9 ± 0.7	7.4 ± 2.1	7.2 ± 1.9	8.3 ± 2.9	9.1 ± 4.5	0.98	0.89	<0.01
Forearm blood flow, mL·100 mL−1·min−1	2.4 ± 0.2	2.1 ± 0.2	1.7 ± 0.2	1.6 ± 0.1	1.5 ± 0.2	1.5 ± 0.1	1.2 ± 0.1	1.3 ± 0.1	0.02	0.95	<0.01
Forearm vascular conductance, arb units	2.9 ± 0.3	2.6 ± 0.3	1.9 ± 0.3	1.9 ± 0.2	1.7 ± 0.2	1.7 ± 0.2	1.4 ± 0.2	1.6 ± 0.2	0.03	0.78	<0.01

Values are means ± SE. [0% n = 18, 40% n = 18, 80% n = 16, 100% n = 18) Conditions were compared using a mixed effect model ANOVA. If a significant interaction was detected post hoc multiple comparisons using Šídák’s multiple comparisons tests was performed. Forearm blood flow and forearm vascular conductance: [0% n = 14; 40% fed: n = 14, fast: n = 12; 80% fed: n = 14, fast: n = 12; 100% fed n = 14, fast n = 12). arb, Arbitrary; cvBRS, cardiovagal baroreflex sensitivity; DAP, diastolic arterial pressure; hb, heart beats; MSNA, muscle sympathetic nerve activity; PNN50, percentage of R-R intervals that varied by 50 ms; Q, cardiac output; RMSSD, root mean squared of successive differences; SAP, systolic arterial pressure.

Figure 4.

FVR, SV, MCAv by percent to presyncope in the fed and fasted conditions. Individual data points and group means ± SD data depicting forearm vascular resistance (FVR; A and B), stroke volume (SV; C and D), and blood flow velocity of the middle cerebral artery (MCAv; E and F) during progressive central hypovolemia in both the fed and fasted conditions. A and B: depicts FVR and FVR reactivity for a given percent to presyncope for both the fed and fasted conditions. C and D: depicts stroke volume and stroke volume reactivity for a given percent to presyncope for both the fed and fasted conditions. E and F: depict MCAv and MCAv reactivity for a given percent to presyncope for both the fed and fasted conditions. Conditions were compared using a mixed-effect model analysis. If a significant interaction was detected post hoc multiple comparisons using Šídák’s multiple comparisons tests was performed. Circles, male participants, triangles, female participants; solid symbols, fed condition; open symbols, fasted condition. *P < 0.05.

To provide information on the effect of the LBNP stimulus on variables of interest, Fig. 5 contains absolute values by LBNP stage. To quantify reactivity to progressive central hypovolemia between conditions, we calculated delta changes from baseline for variables of interest at each LBNP stage relative to baseline. Using a mixed effects model, a significant condition × LBNP stage interaction was identified for stroke volume (P < 0.01, Fig. 5D) and forearm vascular resistance (P < 0.01, Fig. 5B). A significant interaction was detected for absolute FVR by LBNP stage and changes in FVR by LBNP stage but no significant post hoc comparisons were observed for absolute values of FVR by LBNP stage (Fig. 5A). Post hoc multiple comparisons revealed that fasting attenuated the magnitude of increase in forearm vascular resistance at LBNP stage −15, −30, −45, and −60 mmHg (Fig. 5B). A significant interaction was detected for absolute stroke volume by LBNP stage and changes in stroke volume by LBNP stage. Stroke volume exhibited a greater magnitude of decrease in the fasted condition at LBNP stages −30, −45, −60, and −70 mmHg (Fig. 5D). A significant interaction was detected for absolute MCAv by LBNP stage (Fig. 5E). However, the only significant post hoc comparison was at −80 mmHg, which contained only two data points per condition. MCAv decreased at a similar rate between conditions. To investigate if fasting influenced cerebral autoregulation, we calculated the coherence between MAP and MCAv using cross spectral analysis in the low-frequency range (0.07–0.20 Hz). We did not find significant differences at baseline between the fed and fasted conditions (0.42 ± 0.03 fed vs. 0.39 ± 0.21 fasted; P = 0.44, Wilcoxon matched-pairs signed rank test). Due to the low coherence between MAP and MCAv at baseline during both conditions, we did not calculate transfer function gain.

Figure 5.

FVR, SV, MCAv by LBNP stage in the fed and fasted conditions. Individual data points and group means ± SD data depicting forearm vascular resistance (FVR; A and B), stroke volume (SV; C and D), and blood flow velocity of the middle cerebral artery (MCAv; E and F) during progressive central hypovolemia in both the fed and fasted conditions. A and B: depicts FVR and FVR reactivity for a given lower body negative pressure (LBNP) stage for both the fed and fasted conditions. C and D: depicts stroke volume and stroke volume reactivity for a given LBNP stage for both the fed and fasted condition. E and F: depict MCAv and MCAv reactivity for a given percent to presyncope for both the fed and fasted conditions. Conditions were compared using a mixed-effect model analysis. If a significant interaction was detected post hoc multiple comparisons using Šídák’s multiple comparisons tests was performed. Circles, male participants, triangles, female participants; solid symbols, fed condition; open symbols, fasted condition. *P < 0.05.

DISCUSSION

We tested tolerance to progressive central hypovolemia induced by severe LBNP in healthy individuals twice, once in a fasted condition and once in a fed condition. The primary novel finding of this study is that an acute 24 h fast reduced tolerance to a simulated hemorrhagic insult (by ∼10%) compared with a fed (3-h postprandial) condition. In addition, we observed that higher blood ketone levels (β-OHB) were negatively associated with the duration of negative pressure tolerated and that greater changes in β-OHB between the fed and fasted condition were associated with a larger reduction in maximal FVR at the point of presyncope. In the fasted condition, we observed reduced forearm vascular resistance at the point of presyncope and an attenuation in the magnitude of the increase in vascular resistance at most LBNP stages (−15, −30, −45, −60 mmHg). In addition, in the fasted condition participants exhibited greater reductions in stroke volume from baseline. We did not observe any differences between conditions during LBNP in autonomic variables of interest (i.e., MSNA and HRV).

In both civilian and military patients, hemorrhage is the primary cause of death within the first hour of traumatic injury (27). Our study suggests that extended fasting of 24 h modestly reduces an individual’s tolerance to central hypovolemia (by ∼10%), compared with a 3-h postprandial fed state. The reductions in tolerance to central hypovolemia we observed due to fasting are modest compared with the reported 64% decrease in tolerance due to dehydration (5) and 30% decrease in tolerance during low-dose morphine infusion (28). However, even a mild reduction in tolerance due to prolonged fasting is important to understand. In animals, it is well established that food deprivation decreases survival to hemorrhagic hypotension (23). Our study is the first to provide evidence that food deprivation reduces tolerance to central hypovolemia in humans. In addition, a previous study has demonstrated that in food-deprived (24 h) rats that undergo standardized hemorrhage, rats given a glucose infusion before the hemorrhagic insult recover and survive the 7-day observation period (24). Comparatively, rats that received an identical volume of saline developed irreversible shock and died. This study in animals provides evidence to suggest that glucose infusion may increase tolerance to central hypovolemia in humans. Indeed, a study conducted in seriously injured combat casualties, with blood pressure <90 mmHg and heart rate >120, showed that acute infusion of hypertonic glucose produced a rapid and sustained rise in systolic arterial pressure, diastolic arterial pressure, and heart rate compared with saline infusion, mannitol infusion, and a control (29). However, the question as to whether this glucose infusion improved survival rates was not discussed.

Bennet et al. (14) reported that after 48-h of fasting (compared with a 12-h fast) participants exhibited impaired forearm vasoconstriction and a reduced ability to maintain systolic pressure when exposed to brief bouts of LBNP. Notably, after 48-h of fasting significant differences were detected at baseline in resting forearm blood flow and diastolic blood pressure compared with the 12-h fast. In the present study, we report that after an acute 24-h fast, participants exhibited blunted peripheral vasoconstriction (measured from the forearm) during LBNP exposure and had lower peripheral resistance at the point of presyncope. Our observations suggest that a 24-h fast does not alter baseline forearm blood flow but is a sufficient stimulus to blunt maximal peripheral vasoconstriction during central hypovolemia. Peripheral vasoconstriction is a key compensatory mechanism for maintaining arterial pressure and venous return, as inappropriate volume shunting would reduce central venous pressure leading to hemodynamic decompensation (1). A greater capacity to increase vascular resistance is a characteristic of individuals with a high tolerance to LBNP compared with low tolerance individuals, despite having similar resting values (30). Bennet et al. (14) hypothesized that the blunting of peripheral vasoconstriction may be due to fasting-induced inhibition of the SNS, but we did not detect any differences between conditions (in a limited sample population, n = 7) in directly recorded MSNA or in cardiac indices of autonomic control (HRV). Comparable peripheral sympathetic outflow in the fed and fasted conditions during LBNP suggests that the observed attenuation in peripheral vasoconstriction may be mediated by another mechanism—perhaps elevated ketone bodies.

Two recent reviews have highlighted the impacts of ketone bodies (β-OHB in particular) on the cardiovascular system (31, 32). In rats, it has been demonstrated that β-OHB is an autophagy-sensitive metabolite synthesized in the liver that can cause nitric oxide independent vasodilation via potassium channels (33). In patients with heart failure, acute β-OHB infusion (blood ketone concentrations of ∼3.3 mmol/L) significantly reduced systemic vascular resistance (by ∼30%) and increased stroke volume (34). Moreover, lower systemic vascular resistance (a ∼16% reduction) and increased stroke volume were also observed in a single-arm study that had participants ingest ketone ester (blood ketone concentrations of ∼3.23 mmol/L; 35). Ketone bodies may also contribute to energy conservation through direct suppression of the SNS. In rats, ketone body (3-OHB) infusion decreases plasma norepinephrine and reduces the thermogenic response to norepinephrine (36). In mice, β-OHB supplementation has been reported to decrease sympathetic outflow and heart rate (37). In our sample population, we measured a mild but significant increase in blood ketone levels (β-OHB ∼0.47 mmol/L) after a 24-h fast. Ketone therapies have demonstrated some potential benefits in the treatment of cardiovascular diseases such as heart failure (32, 34). However, it is possible that the presence of increased circulating β-OHB in the fasted condition acts as a sympatholytic attenuating peripheral vasoconstriction and thus contributed to earlier decompensation during central hypovolemia.

Lower central volume leads to decreased cardiac filling and thus reduced stroke volume in accordance with the Frank–Starling relationship. We estimated stroke volume on a beat-to-beat basis using the pulse contour method (25). This method cannot accurately measure absolute values of stroke volume but reliably measures changes in stroke volume (38). Our data suggest that in the fasted condition stroke volume decreases at a greater rate compared with the fed condition during LBNP. Using a similar method as the current study, a significant linear regression between blood volume loss and reductions in stroke volume has been previously demonstrated (39). Therefore, due to lower peripheral resistance in the fasted condition, more blood may have been sequestered into the lower body regions, away from central circulation.

Limitations

Our study suggests that extended fasting of 24 h may reduce individuals’ tolerance to central hypovolemia by ∼10%, compared with a 3-h postprandial fed state as measured via duration of negative pressure. We did not detect a significant difference (P = 0.079) in CSI between conditions, likely due to the manual adjustments made during stage transitions. In an attempt to maintain stable muscle sympathetic nerve recordings, stage transitions were done carefully with transition time varying from 20 s to 1 min. The CSI calculation does not account for LBNP stage transition.

The cerebral blood flow velocity data should be interpreted with caution, as we did not measure end-tidal CO₂ and thus cannot determine if partial pressure of arterial CO₂ was a driving factor in the cerebral blood velocity response.

Perspectives and Significance

These data suggest that an acute fast can reduce an individual’s tolerance to central hypovolemia potentially by blunting peripheral vasoconstriction. In addition, we report that higher blood ketone levels are associated with reduced tolerance to LBNP and lower maximal forearm vascular resistance. Previous studies in animals have demonstrated that food deprivation increases mortality from hemorrhage-induced hypotension and that glucose infusion beforehand prevents this fatal outcome in food-deprived rats (40, 41). This study in animals provides evidence that acute glucose infusion may bolster tolerance to central hypovolemia if an individual is experiencing a prolonged negative energy balance. Assessing an individual’s energy balance may be an important consideration in the prehospital setting during a hemorrhagic insult as it may compound with other factors (such as dehydration or an analgesic) to reduce an individual’s compensatory capacity to central hypovolemia. In addition, future experimental studies utilizing lower body negative pressure should consider reporting their participants last caloric intake or assessing blood ketone levels to ensure individuals are not in a fasted state, as this may confound tolerance tests.

Conclusions

In conclusion, an acute 24-h fast reduces tolerance to progressive central hypovolemia compared with a 3-h postprandial fed condition. In addition, duration of negative pressure tolerated was associated with blood ketone levels. Blunted peripheral vasoconstriction in the fasted condition is a potential contributing factor to the observed reduction in tolerance to LBNP. We did not observe any differences in peripheral sympathetic outflow (n = 7), heart rate, or heart rate variability between conditions during LBNP, suggesting that the responsiveness of the SNS to a simulated hemorrhagic insult was not significantly altered between conditions.

DATA AVAILABILITY

Data will be made available upon reasonable request.

GRANTS

This work was supported by an Oregon Health and Science University Fellowship for Diversity in Research (to J. E. Gonzalez), a grant from the Michigan Space Grant Consortium, an endowment from the Portage Health Foundation (Houghton, Michigan), and by the National Heart, Lung, and Blood Institute of the National Institutes of Health under Award No. T32HL083808.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.E.G. and W.H.C. conceived and designed research; performed experiments; analyzed data; interpreted results of experiments; prepared figures; drafted manuscript; edited and revised manuscript; approved final version of manuscript.