Preoperative Hypovitaminosis D Can Predict Development of Postoperative Cognitive Dysfunction

Increasing evidence links POCD to surgery and anesthesia and is recognized as an important neuropsychological adverse outcome in surgical patients [4]. Diagnosis of POCD requires complicated neuropsychological testing and is often delayed, so possible predictor for early detection or prediction of POCD is essential for its prevention and treatment [5]. Also, increasing evidence defined age as a major risk factor for POCD [2]. Multiple studies tried to evaluate biomarkers for screening patients for an increased risk of POCD; preoperative neuronspecific enolase levels are more suitable for assessment of incidence of POCD [6]. The ratio of β-amyloid and Tau proteins, biomarkers of Alzheimer’s disease, can be used in forecasting the occurrence of POCD, especially in elderly patients [7]. Preoperative salivary cortisol AM/PM ratio was significantly associated with the presence of early POCD [5].


Introduction
Postoperative cognitive dysfunction (POCD) is a neurologic dysfunction characterized by impairment of memory function and intellectual ability [1]. POCD is also associated with greatly reduced speech processing after surgery [2] as well as increased use of healthcare resources [1]. POCD may persist for a period ranging from weeks to months and affects patient's quality of life [3].
Increasing evidence links POCD to surgery and anesthesia and is recognized as an important neuropsychological adverse outcome in surgical patients [4]. Diagnosis of POCD requires complicated neuropsychological testing and is often delayed, so possible predictor for early detection or prediction of POCD is essential for its prevention and treatment [5]. Also, increasing evidence defined age as a major risk factor for POCD [2].
Multiple studies tried to evaluate biomarkers for screening patients for an increased risk of POCD; preoperative neuronspecific enolase levels are more suitable for assessment of incidence of POCD [6]. The ratio of β-amyloid and Tau proteins, biomarkers of Alzheimer's disease, can be used in forecasting the occurrence of POCD, especially in elderly patients [7]. Preoperative salivary cortisol AM/PM ratio was significantly associated with the presence of early POCD [5].

Patients & Methods
The current study intended to include >250 patients assigned for surgical procedures extending for >60 min under sevoflurane anesthesia. Exclusion criteria included duration of surgery for <60 min, age <18 years, endocrinopathy inducing HVD, orthopedic diseases or surgery, maintenance on VD supplemental therapy, pregnancy and/or lactation, chronic kidney diseases. The study protocol was approved by the Local Ethical Committee and all enrolled patients signed written fully informed consents. Group A included native Kuwaiti and Group B included native Egyptian patients and immigrants were excluded from either group. Preoperative preparation included general examination to determine demographic data, baseline clinical and hemodynamic data and to take blood samples for baseline investigations and ELISA estimation of serum 25OHD by a clinical chemist who is blinded about the hypothesis of the study or its outcome.

Sampling & investigations
Venous blood samples (5ml) were collected from the antecubital vein under complete aseptic conditions at booking time. The obtained sample was kept in a plane container and allowed to clot then serum was separated by centrifugation at 3000 rpm for 10 min. Serum was removed and placed in pyrogenfree Eppendorf tubes and stored at -70°C until ELISA assayed by Spectrophotometer for estimation of 25OHD using the Calbiotech Vitamin D Kit (Calbiotec, A Life Science Co, USA; Catalog No. VD220B) which is a solid phase enzyme-linked immunoassay (ELISA) based on the principal of competitive binding [17].

Clinical evaluation
a) Vitamin D sufficiency status was defined according to 25-OHD concentration as follows: ≥75 nmol/L sufficient level, 50-75nmol/L insufficient level and <50nmol/L deficient level. Vitamin D deficiency was categorized as mild, moderate and severe if 25-OHD concentration was 25-50nmol/L, 12.5-25nmol/L and <12.5nmol/L, respectively [18]. b) Cognitive function was assessed preoperatively, 48hr, 1-wk and 2-wk postoperatively (PO) using the Digit span (DS) test: Forward DS tests are used to measure short-term memory and general attention. Backward DS tests are used to measure verbal working memory [19][20][21].

Preoperative preparation
Patients were preoperatively evaluated for ASA grade and presence of associated co-morbidities. Patients with comorbidities were adjusted and were maintained on their preoperative therapies till day of surgery and postoperatively to maintain the control state. Patients were premedicated by midazolam 0.02mg/kg.

Anesthetic procedure
Anesthesia was induced, in studied groups, using propofol 2mg/kg, fentanyl 1-2ug/kg, and rocuronium 0.6mg/kg, and was maintained with sevoflurane, fentanyl and rocuronium. After tracheal intubation, the lungs were ventilated with 100% O 2 in air using a semi-closed circle system for a tidal volume of 6-8ml/ kg, and end-tidal carbon dioxide (paCO 2 ) of 32-35mmHg. Patients were continuously non-invasively monitored for Mean Arterial Pressure (MAP) and Heart Rate (HR).

Sample size calculation
Review of literature showed that the overall incidence of HVD in Middle East ranged from 37.5% [22] to 26% [23], to get study

Statistical analysis
Obtained data were presented as mean ± SD, numbers and percentages. Results were analyzed using paired t-test, One-way ANOVA Test and Chi square test. Statistical analysis was conducted using the IBM SPSS (Version 23, 2015) for Windows statistical package. P value <0.05 was considered statistically significant.

Results
The study included 284 patients eligible for evaluation, 30 patients were excluded for not fulfilling the inclusion criteria and 254 patients were enrolled; 127 Kuwaiti (Group A) and 127 Egyptian patients (Group B) ( Figure 1). Interestingly, mean educational duration was significantly longer and the frequency of employed patients was significantly higher among Egyptian than Kuwaiti patients. Patients collected from both communities showed non-significant (p>0.05) difference as regards preoperative data (Table 1).  Data are presented as numbers and mean ± SD; percentages are in parenthesis; VD: Vitamin D; P value indicates the significance of difference between both groups.

Journal of Anesthesia & Intensive Care Medicine
At time of enrolment; 31 patients (12.2%) had sufficient and 88 patients (34.6%) had insufficient, while 135 patients (53.2%) had deficient VD level with significant (p=0.00001) in favor of Group B. Among patients had deficient VD level, 48 patients (35.6%) had mild, 55 patients (40.7%) had moderate and 32 patients (23.7%) had severe HVD with significant (p=0.0001) difference in favor of Group B. Mean serum VD levels were non-significantly (p=0.262) higher in patients of group B than those of group A. Despite of the significantly higher level of VD in patients of group B who had insufficient and deficient VD than corresponding patients of group A, patients of group B having sufficient VD had non-significantly (p=0.515) higher serum VD levels than those of group A ( Table 2). Data are presented as numbers, percentages and mean ± SD; VD: Vitamin D; P value indicates the significance of difference between both groups All surgeries were conducted uneventfully within a mean operative time of 154.4 ± 37; range 70-215 minutes. Mean intraoperative blood loss was 415±158ml, but no patient required blood transfusion. HR and MAP measures recorded at end of surgery were significantly lower in patients of group A compared to that recorded at time of induction of anesthesia. In group B, at end of surgery, MAP measures were significantly lower, while HR measures were non-significantly lower than them at-induction measures. MAP measures at end of surgery were non-significantly higher, while HR measures were significantly (p=0.036) higher in patients of group B compared to patients of group A (Figure 1).
Thirty-eight (15%) patients required transfer to ICU for hemodynamic and respiratory stabilization; 25 patients of group A and 13 patients of group B with significantly (p=0.035) higher frequency among patients of group A. Mean ICU stay duration was 31.7±17.8; range: 12-72hr with non-significant (p=0.302) difference between both groups. The remaining patients were transferred to post-anesthetic care unit (PACU) and shifted to surgical ward after a mean duration of 87±45.8; range: 30-240min, with non-significant (p=0.363) difference between patients of both groups (Table 3).

Journal of Anesthesia & Intensive Care Medicine
patients had DS score <6 on forward testing was significantly higher in patients of group A than in group B. However, on backward testing the difference was non-significantly higher between both groups, but in favor of group B (Table 4). Mean DS score at 48-hr and 1-wk PO was significantly lower on both forward and backward testing of all patients in comparison to preoperative scores with significantly lower score for patients of group A than group B at 48-hr and 1-wk PO for forward and at 1-wk PO on backward testing. At 2-wk PO, mean DS score was non-significantly lower in patients of group B on forward and backward testing on comparison to preoperative score, while in patients of group A, the difference in score on backward testing was still significantly lower than their preoperative score ( Table  4).
In total studied population, preoperative serum 25OHD level showed negative significant correlation with age and BMI, while showed positive significant correlation with male gender, duration of education. Moreover, median value of estimated PO score on DS testing showed positive significant correlation with preoperative serum 25OHD on both forward and backward DS test evaluation. These correlations with preoperative serum 25OHD levels were more pronounced in Kuwaiti patients than in Egyptian patients (Table 5). Regression analysis defined preoperative low serum 25OHD (β=0.379, p<0.001), short duration of education (β=0.203, p=0.002), high BMI (β=0.173, p=0.006) and old age (β=0.171, p=0.007) as positive predictors for POCD. ROC curve analysis for

Discussion
The current study included 254 patients underwent major surgical procedures under sevoflurane anesthesia within a mean operative time of 154.4 (±37) min and 38 patients had admitted to ICU for a mean duration of 31.7 (±17.8) hr. All patients showed significantly lower score on DS test in relation to their preoperative score and despite of the progressively regained Cognitive Function (CF), the 2-wk PO scores were still significantly lower than their preoperative scores.
Other studies reported non-significant difference between frequency and severity of POCD after either of sevoflurane or propofol anesthesia [29][30][31]. The situation was made worse by Guo et al. [32] who found sevoflurane could improve cerebral oxygenation even in patients with impaired cerebral oxygenation. Moreover, Shoair et al. [25] found using highly anticholinergic or sedative-hypnotic drugs prior to surgery are risk factors for POCD and Sato et al. [33] suggested that the risk of early PO attention dysfunction may be magnified in patients experiencing other risk factors of POCD, as severity of surgery, occurrence of complications, and pre-existing cognitive impairments. Furthermore, Fan et al. [34] found the use of LMA reduced the prevalence of POCD during pediatric neurosurgery than with endotracheal intubation.
The reported continued CD for 2-wk after surgery could be explained depending on the experimental findings of Huang et al. [35] who detected neuro-inflammation, glia activation, aberrant tau phosphorylation and DNA damage in the hippocampus of aged rats underwent a midline laparotomy or exposed to sevoflurane anesthesia alone up to 14 days PO.
About 88% of studied patients had insufficient-deficient serum 25OHD levels, a prevalence which coincided with Chakhtoura et al. [36] who documented that the Middle East and North Africa region is of the lowest serum 25OHD concentrations, worldwide. However, HVD was more evident in Kuwaiti than Egyptian patients as manifested by significantly higher frequency of patients had HVD and significantly lower serum 25OHD among Kuwaiti patients. These findings spot light on the impact of cultural, environmental and educational factors on VD status. In support of this assumption, the reported significantly higher educational duration and frequency of employee among Egyptian patients and this most probably provided greater opportunity for Journal of Anesthesia & Intensive Care Medicine sunlight exposure with subsequent increased VD synthesis. In line with these data and explanation Al-Taiar et al. [37] reported that despite the abundant sunshine in Kuwait, prevalence of VD deficiency and severe deficiency was 81.21% and 39.48%, respectively and this reflects strong sun avoidance behavior, so adequate outdoor daytime activities should be encouraged.
Preoperative serum VD levels of studied patients showed negative significant correlation with age and BMI, while showed positive significant correlation with male gender, duration of education. Similarly, Golan-Cohen et al. [38] in multivariate analysis of results of cross-sectional study found older age, female gender, higher BMI, sunscreen use, and low sun exposure were associated with lower VD levels and Hashem et al. [39] found most Kuwaiti adolescents are physically inactive and attributed this to low/medium maternal educational levels.
Collectively, median value of PO score on DS test was positively correlated with preoperative serum VD. This finding supported previous reports concerning the association between HVD and poorer mental flexibility [40] or higher likelihood of functional deficits among patients with memory complaint [41], CD in specific domains in old adults [42], and increased problems of memorizing new information [43]. Moreover, these data are consistent with results of VD supplemental intake and slowing, prevention, or improvement of neurocognitive decline [44], decline in verbal fluency [45], or memory loss (46). Moreover, statistical analyses defined preoperative HVD as positive specific predictor for the possibility for development of POCD. These findings point to a close relationship between VD and cognitive function (CF), so HVD could give an objective predictor for the possibility of development of POCD. Similarly, Sakuma et al. [47] found HVD was independently associated with a higher prevalence of CD and De Oliveira et al. [48] documented that the independent and inverse relation between low 25OHD levels and elevated depressive symptoms suggests that HVD may be a risk factor for late-life depression, particularly among women.
Few studies evaluated the relation between HVD and CF disturbances in middle aged populations; the current study included patients in age range of 28-79 years and all these patients showed POCD. Moreover, Similarly, Shu et al. [49] reported that the depth of anesthesia under combined intravenous-inhalational anesthesia yielded milder influence on POCF in young and middleaged patients.
Unfortunately, previous studies dealt with the relation between HVD and POCD were deficient. However, the obtained results go in hand with Zhang et al. [50] who documented that after adjusting for confounding factors, preoperative VD deficiency was associated with an increased risk of POCD. In trial to explore the association between HVD and disturbed CF, Annweiler et al. [51] suggested an association between HVD and dysfunction of the frontal-subcortical neuronal circuits, particularly the dorsolateral circuit in patients with disturbed CF and HVD. Thereafter, Annweiler [51] documented that VD is essential to neurotransmitters and neurotrophins regulation with anti-inflammatory and antioxidant neuroprotective action, so HVD may result in neurological dysfunction that may explain part of the cognitive disorders.

Conclusion
POCD is a common incident event among patients undergoing major surgical procedures consuming long operative time under sevoflurane anesthesia. Sevoflurane anesthesia could not be accused as the sole precipitating factor. Old age, high BMI, duration and severity of surgery and development of complications are cofactors for upcoming CD. Preoperative HVD has a pronounced effect and correlated with all of these precipitating factors and could be used as an independent significant predictor for POCD.