A prospective observed study was conducted. The sedative, either propofol (Group A), midazolam (Group B), or dexmedetomidine (Group C), was used according to the patients’s needs, and light sedation was applied. Gastric motility was measured with gastric antrum ultrasound before and after the administration of sedatives. The effects of different sedatives on gastric antrum contraction, which includes frequency (ACF), amplitude (ACA), and motility index (MI), were analyzed and compared. In addition, 18 adult healthy volunteers were recruited as normal controls to compare with patients with ABI.

This research was approved by the Ethics Committee of Clinical Research and Experimental Animals of the First Affiliated Hospital of Sun Yat-sen University (Ethics No. [2018] 161). The institutional review board waived the requirement for informed consent since no intervention was performed and no personally identifiable information appeared.

Patients hospitalized in the neurosurgery ICU of the First Affiliated Hospital, Sun Yat-sen University, from July 2018 to November 2018 were recruited (Figure 1).

Inclusion criteria: 18–80 years old. Patients underwent ABI including intracranial hemorrhage, post-selective operation with brain tumors, or traumatic brain injury. Need therapy of short-acting sedatives like midazolam, propofol, or dexmedetomidine, and no oral hypnotic drugs were used.

Exclusion criteria: known upper gastrointestinal anatomical problems. Pregnancy.

The Richmond Agitation-Sedation Scale (RASS) was maintained at −2 ~ 1 points, meaning the depth of sedation was maintained at light sedation. A daily wake-up procedure was performed on each patient at 6–9 a.m.

Gastric antrum ultrasound imaging was performed before and 20 min after sedation. All ultrasonic gastric motility monitoring in this study was completed between 9 and 10 a.m., which is 2 h later after the daily wake-up procedure. Enteral nutrition was performed at 60–80 mL/h through a nasogastric tube. Gastric antrum images were obtained with a 2.5–6 MHz curvilinear probe (SONIMAGE HS1 portable ultrasonic, Konica Minolta). Ultrasound examination was performed at 30-degree head-of-bed elevation and supine position. The antrum was located after identifying the liver’s left anterior lobe, the pancreas’s head, and the abdominal aorta. ACF was measured for 6 min, and an average number of gastric antrum contractions were observed every 2 min. The maximum gastric antral diastolic area (S_max) and minimum contraction area (S_min) were measured thrice. ACA was calculated as follows: ACA = (S_max - S_min)/ S_max. MI was calculated as follows: MI = ACF × ACA. The same experienced sonographer performed all of the scans, as the force of ultrasound probe placement may affect the interpretation of the cross-section.

Bowel sound auscultation was conducted before and 20 min after sedation. Bowel sound auscultation was for 3 min and the frequency of bowel sounds was calculated per minute.

The following clinical data were collected: age, gender, disease diagnosis, complications such as diabetes and coronary heart disease, body mass index (BMI), GCS score, and the utilization of basic medications such as proton pump inhibitors, or nonsteroidal anti-inflammatory drugs, or opioids, enrollment time after admission to ICU, length of ICU stay, and death.

The SPSS software (version 23.0) was used for the statistical tests. Normally distributed data were expressed as mean ± standard deviation, non-normal distribution data median (interquartile ranges, IQR). The paired student’s test or the paired rank-sum test was adopted to analyze the difference before and after the sedative therapy. The least significant difference t-test (LSD-t) or Bonferroni method was applied for multiple comparisons. p < 0.05 was considered to be statistically significant.

This study included 37 patients. Among them, 30 cases were with intracranial hemorrhage, 5 cases were with brain tumor, and 2 cases were with traumatic brain injury. In addition, 18 healthy volunteers were recruited for this study to compare gastric motility in patients with ABI. The sedative dosage was recorded. In this study, the dosage of propofol (n = 13) was 0.036–0.043 mg/kg/h, the dosage of midazolam (n = 12) was 0.031–0.067 mg/kg/h, and the dosage of dexmedetomidine (n = 12) was 0.171–0.230 mg/kg/h. It is shown in Table 1.

In this study, patients with ABI showed lower ACF (2.41 ± 0.89 times/2 min) and MI (1.25 ± 0.57) than health volunteers (ACF: 4.5 ± 0.39 times/2 min, p = 0.001; MI: 3.59 ± 0.24, p = 0.001). There was no significant inhibition of ACA (50.80 ± 11.60%), as shown in Table 2.

In patients receiving propofol (n = 13), the ACA decreased significantly (53.31 ± 9.76% before vs. 26.41 ± 18.01% after, p = 0.002), ACF decreased significantly (2.27 ± 0.78 times/2 min before vs. 1.08 ± 0.74 times/2 min after, p = 0.002), and MI decreased significantly (1.14(0.59, 1.44) before vs. 0.84(0.09, 0.83) after, p = 0.002). The bowel sounds had no significant change after using propofol (3 (2.5, 3) times/min before vs. 3 (2, 3) times/min after, p = 0.317). It is shown in Table 3.

In patients receiving midazolam (n = 12), the ACA decreased significantly (53.35 ± 13.87% before vs. 21.04 ± 12.47% after, p = 0.002), ACF decreased significantly (2.23 ± 1.04 times/2 min before vs. 1.32 ± 1.04 times/2 min after, p = 0.007), MI decreased significantly (1.48 (0.73, 1.62) before vs. 0.31 (0.04, 0.58) after, p = 0.007). The bowel sounds decreased after using midazolam (3 (2.25, 3) before vs. 2 (2, 3) after, p = 0.034). It is shown in Table 4.

In patients receiving dexmedetomidine (n = 12), ACF decreased significantly (2.71 ± 0.86 times/2 min before vs. 1.85 ± 0.86 times/2 min after, p = 0.01), MI decreased significantly (1.37 (0.85, 1.5) before vs. 0.9 (0.35, 1.14) after, p = 0.01), and no significant difference in ACA (45.54 ± 9.98% before vs. 34.44 ± 16.04 after, p = 0.099) and bowel sounds (3 (2, 3) before vs. 3 (2.25, 3) after, p = 1). It is shown in Table 5.

Dexmedetomidine showed less inhibitory effects on ACA compared with midazolam and propofol. D-value of ACA was −11.10 ± 19.96% in dexmedetomidine, −32.21 ± 14.03% in midazolam, −26.90 ± 16.77% in propofol, p = 0.003. ACA suppression ratio was 20.67 ± 33.59% in dexmedetomidine, 60.43 ± 22.40% in midazolam, 51.50 ± 32.83% in propofol, p = 0.002 (Figures 2A,D).

Dexmedetomidine also showed less inhibitory effects on MI. MI suppression ratio was 36.00 ± 34.77% in dexmedetomidine, 68.81 ± 20.84% in midazolam, 60.69 ± 27.49% in propof ol, p = 0.012. D-value of MI was −0.28 ± 0.90 in dexmedetomidine, −0.89 ± 0.51 in midazolam, −0.4 ± 0.59 in propofol, p = 0.066 (Figures 2C,D).

There was no statistical difference in ACF. D-value of ACF was −1.08 ± 0.74 times/2 min in propofol, −0.96 ± 1.00 times/2 min in midazolam, 0.86 ± 0.94 times/2 min in dexmedetomidine, p = 0.811. ACF suppression ratio 49.74 ± 34.81% in propofol, 37.87 ± 35.24% in midazolam, and 29.0 ± 32.8% in dexmedetomidine, p = 0.518 (Figures 2B,D).

Sedation is a common treatment for brain injury. Previous studies that explored the effects of sedation on gastric motility in patients with ABI have focused on deep sedation (13, 14). However, light sedation is the current trend in ICU sedation management (5, 15). This study found that light sedation can also reduce gastric motility in patients with ABI.

This study also compared the inhibitory effects of different sedatives on gastric motility. We found dexmedetomidine had less inhibitory effects on gastric motility than propofol and midazolam. Midazolam, a short-acting benzodiazepine, acts directly on the γ-aminobutyric ABId receptor (16). Propofol acts on the subtype A of the γ-aminobutyric ABId receptor (17). Different from propofol or midazolam, dexmedetomidine is a highly selective α² agonist, having mild sedative, mild analgesic, and antisympathetic properties (18). Dexmedetomidine showed a less inhibitory effect than midazolam and propofol, possibly due to its lighter sedation effect and sympathoexcitatory depressant function (19). In addition, dexmedetomidine has the ability to attenuate intestinal ischemia–reperfusion injury, inhibit the inflammatory response, and ameliorate the stress response (20). This could also rationalize the weaker inhibition of gastric antral contraction by dexmedetomidine.

This study also observed bowel sounds to evaluate intestinal motility. We found that midazolam suppresses bowel sounds, but propofol and dexmedetomidine had no significant inhibition of bowel sounds. However, there are no studies to explore the effects of sedative medications on bowel sounds in patients with ABI. We conjecture that this might be related to the influence of sedatives on gastrointestinal hormones. Narchi et al. (21) found that sedatives, including midazolam and propofol, affect gastrin secretion. Early animal experiments revealed that midazolam intervention could cause a decrease in the secretion of motilin (22, 23). Xu et al. (24) also found that propofol can increase the levels of gastrin and vasoactive intestinal peptide in patients undergoing gastrointestinal endoscopy. Moreover, in a clinical study of open surgery for colon cancer, patients undergoing intravenous anesthesia involving dexmedetomidine had a quicker recovery of gastrointestinal motility, and the levels of prokinetic gastrointestinal hormones such as motilin and gastrin in the plasma increased (25). Thus, It could be explained by the different mechanisms of midazolam, propofol, and dexmedetomidine affecting gastrointestinal motility and bowel sound.

This is a prospective observational study. There are some limitations in this study. First, the clinical application of basic drugs such as opioids (12), proton pump inhibitors, and non-steroidal anti-inflammatory drugs (26) may influence gastric motility. However, the before-and-after comparison was employed in this study, which mitigated the errors brought by other therapies. Second, in this study, radioactive nuclides, gastrointestinal electrical examination, and capsule endoscopy were not utilized for the detection of gastrointestinal motility because these methods are not suitable for before-and-after comparison of short-acting sedatives. By contrast, gastric antrum ultrasound can be operated non-invasively and enables dynamic re-examination. Nevertheless, further studies were needed to explore the specific mechanisms of different sedatives affecting gastric motility.