Sage Journals: Discover world-class research

Abstract

Study Design

Retrospective study.

Objectives

To investigate the impact of diabetes mellitus on neurological recovery and determine the relationship between moderate-severe diabetes and neurological recovery in patients with cervical spinal cord injury (CSCI) without bone injury.

Methods

A retrospective study was conducted on 389 consecutive patients aged ≥65 years with CSCI without bone injury across 33 medical institutes. The patients were divided into a nondiabetic group (n = 270) and a diabetic group (n = 119). Neurological outcomes were compared between the two groups through propensity score matching. The impact of moderate-severe diabetes (defined as hemoglobin A1c ≥ 7.0% or requiring insulin treatment) on neurological recovery was evaluated through multiple linear regression analysis.

Results

Propensity score matching revealed no significant differences between the diabetic and nondiabetic groups in terms of American Spinal Injury Association (ASIA) impairment scale grade and mean total ASIA motor scores (AMS) at 6 months post-injury. Multiple linear regression analysis indicated that age on admission (B = −0.34; 95% confidence interval [CI], −0.59 to −0.08; P = 0.01), dementia (B = −16.50; 95% CI, −24.99 to −8.01; P < 0.01), and baseline total AMS (B = −0.62; 95% CI, −0.72 to −0.51; P < 0.01) were negative predictors of neurological recovery at 6 months post-injury. The presence of moderate-severe diabetes did not influence neurological recovery at 6 months post-injury.

Conclusions

Diabetic patients with CSCI without bone injury achieved improvements in neurological function comparable to those of nondiabetic patients. Moderate-severe diabetes did not affect neurological recovery in patients with CSCI without bone injury.

Keywords

cervical spinal cord injury without bone injury diabetes mellitus moderate-severe diabetes mellitus blood glucose level American spinal injury association impairment scale grade American spinal injury association motor scores older patients propensity score matching multiple linear regression analysis prognostic factors dementia

Introduction

Cervical spinal cord injury (CSCI) is one of the most devastating spinal injuries, and can cause severe permanent neurological deficits. A nationwide survey in Japan demonstrated that the overall rate of CSCI among cases of traumatic spinal cord injury (SCI) was 88.1%.¹ Among the CSCI cases, 70.7% did not suffer a bone injury resulting from minimal trauma, the older population. The incidence of CSCI without bone injury is anticipated to increase with the growing aging population.

Although several prognostic factors for neurological recovery in patients with CSCI without bone injury have been identified, most studies have focused on imaging features or surgical interventions. Preexisting conditions such as cervical spondylosis, ossification of the posterior longitudinal ligament (OPLL), and signal intensity (SI) changes in the spinal cord on magnetic resonance imaging (MRI) are relevant prognostic factors for neurological recovery in patients with CSCI without bone injury.^2-4 Surgical intervention is also reported as a prognostic factor for neurological recovery in patients with CSCI without bone injury.^5,6 Recently, Nakajima et al revealed that factors such as body mass index (BMI), OPLL, SI changes on MRI, American Spinal Injury Association (ASIA) impairment scale (AIS) on admission, comorbidity of dementia/delirium, and post-injury pneumonia were independent prognostic factors for the recovery of walking ability in patients with CSCI without major bone injury.⁷

Diabetes mellitus is one of the most common comorbidities which is well known to negatively affect perioperative complications in spine surgery, including decreased spinal fusion rates and increased surgical site infections.^8-10 Additionally, numerous studies have suggested that diabetes is a risk factor for poor surgical outcomes in patients with lumbar disc degenerative disease.^11,12 Regarding cervical spondylotic myelopathy (CSM), some studies have reported preferable surgical outcomes in diabetic,^13-15 while others diabetes contributes to poor surgical outcomes.^16,17 However, the influence of diabetes on CSCI without bone injury also remains unclear because of the few studies.⁷ Several animal studies supported that diabetes negatively affected neurological outcomes in patients with CSCI without bone injury.^18,19 Furthermore, no previous studies have evaluated the relationship between diabetes severity and neurological recovery in patients with CSCI without bone injury.

This multicenter large-cohort study aimed to investigate the impact of diabetes on post-injury complications and neurological recovery, and the relationship between moderate-severe diabetes and neurological recovery in patients with CSCI without bone injury.

Materials and Methods

Participants

The multicenter study was conducted by the Japan Association of Spine Surgeons with Ambition and included 1512 consecutive patients aged ≥65 years with cervical spine/SCIs from 33 medical institutes between 2010 and 2020. The original dataset used in the study has also been utilized in other studies.^7,20,21 Of the 1512 patients, 614 (40.6%) were diagnosed with CSCI without bone injury. This condition was defined as a CSCI with no evidence of spinal fractures or dislocations on radiography or computed tomography (CT).²² Any patients with missing values in diabetes, baseline AIS grade, baseline ASIA motor score (AMS), AIS grade at 6 months after injury, or AMS at 6 months after injury were excluded from the study. Patients with baseline AIS grade A were also excluded because the probability of the AIS grade converting to grades C–E is very low.²³ A total of 389 patients followed up for at least 6 months were included in the present study (Figure 1). The 6-month time period was chosen based on previous clinical trials that demonstrated that neurological recovery after traumatic SCI mainly occurs within the first 6-9 months.^24,25 Patients were divided into two groups: nondiabetic (n = 270) and diabetic group (n = 119). The diabetic group consisted of patients with hemoglobin A1c (HbA1c) ≥6.5%, those undergoing diabetes treatment with oral agents or insulin or both, or those previously diagnosed with diabetes by diabetologists by the clinical practice guidelines for diabetes.⁵ National Glycohemoglobin Standardization Program method was used for measuring HbA1c. Patients with an HbA1c level ≥7.0% or those requiring insulin treatment were categorized into the moderate-severe diabetic group (n = 55).^26,27

Figure 1.

Flowchart of the multicenter cohort. ASIA, American Spinal Injury Association; AIS, American Spinal Injury Association impairment scale.

Data Collection

Demographic variables, including age at injury, sex, BMI, and medical comorbidities, were collected. Medical comorbidities included diabetes, hypertension, cardiovascular disease, cerebrovascular disease, rheumatoid arthritis, osteoporosis, respiratory disease, renal disease, Parkinson’s disease, dementia, and a history of surgery for musculoskeletal disorders. For diabetic patients, the HbA1c levels at admission and medication details, including insulin, were also documented. Cervical OPLL and diffuse idiopathic skeletal hyperostosis (DISH) were assessed by spinal radiography and CT. SI changes in the cervical spinal cord were evaluated using T2-weighted sagittal and axial MRI at the time of injury. Senior spinal surgeons and physical therapists at each center evaluated the neurological status on admission, at discharge, and 6 months post-injury using the AIS and AMS. The indications for surgery and steroid therapy were determined by the attending spinal surgeons at each institute. Complications during hospitalization, such as motor or sensory neurological deterioration, cerebral infarction, delirium, dysphagia, respiratory failure, pulmonary embolism, pneumonia, and renal infection, were also documented.

Statistical Analysis

The level of statistical significance was set at P < 0.05. The R Statistical Package version 2.6.0 (R Foundation for Statistical Computing, Vienna, Austria) was used for all statistical analyses. Continuous variables were analyzed using the unpaired t-test, Welch’s t-test, or Wilcoxon rank-sum test, as appropriate. Categorical variables were analyzed using the χ² test or Fisher’s exact test. Propensity score matching was conducted to compare neurological outcomes between the nondiabetic and diabetic groups. A multivariate logistic regression model was used to calculate the propensity scores. The moderator variables were age, sex, BMI, cervical OPLL, cervical DISH, SI changes on MRI, medical comorbidities, baseline AIS grade, baseline AMS score, steroid therapy, and surgical intervention for CSCI. To adjust for baseline characteristics and comorbidities, 1-to-1 matching with fixed caliper widths (0.15) was performed without replacement. Each nondiabetic case was matched with a corresponding case in the diabetic group with the same propensity score. Standardized differences were used to measure covariate balance, with a standardized difference <10% indicating a negligible difference between both groups. Additionally, patients with an HbA1c level ≥7.0% or those requiring insulin treatment were classified as the moderate-severe diabetes group. Clinically relevant variables (age, sex, BMI, cervical DISH, SI changes on MRI, baseline AIS grade, baseline AMS, and surgical intervention), along with variables with significance level <0.05, as determined by univariate analysis (blood glucose level on admission, diabetes, rheumatoid arthritis, dementia, post-injury complication of dysphagia, cervical OPLL, and steroid therapy for CSCI), were included in the multiple linear regression analysis with listwise deletion of missing data. This was performed to assess the influence of moderate-severe diabetes on the degree of improvement in total AMS at 6 months post-injury.

Results

Comparison of Baseline Characteristics and Medical Comorbidities Between the Nondiabetic and Diabetic Groups

A total of 389 patients were included in the present study (male, n = 279, 71.7%; female, n = 110, 28.3%; mean age at the time of injury, 74.7 ± 6.4 years). The blood glucose level on admission in the diabetic group was significantly higher than that in the nondiabetic group (Table 1). The prevalence of dementia was significantly higher in the diabetic group than in the nondiabetic group. The mean baseline total AMS and lower extremity AMS scores were comparable between the two groups, but both were slightly lower in the diabetic group.

Table 1.

Comparison of Baseline Characteristics and Medical Comorbidities.

Variables	Non-diabetic (n = 270)	Diabetic (n = 119)	OR [95% CI]	P Value
Demographic variables
Age (years)	74.4 ± 6.4	75.5 ± 6.3	-	0.08**
Sex (male, %)	188 (69.6)	91 (76.5)	0.8 [ 0.8-2.4]	0.17**
BMI	22.3 ± 3.3	22.3 ± 3.7	-	0.98**
Blood glucose level (mg/dl)	125.1 ± 36.3	172.6 ± 71.9	-	9.8 × 10⁻¹⁴**
HbA1c (%)	-	7.1 ± 1.1	-	-
Medical comorbidities (%)
Hypertension	128 (47.4)	66 (55.5)	1.4 [0.9-2.2]	0.15′
Cardiovascular disease	37 (13.7)	16 (13.4)	1.0 [0.5-1.9]	0.93′
Cerebrovascular disease	20 (7.4)	7 (5.9)	0.8 [0.3-2.0]	0.58′
Rheumatoid arthritis	1 (0.4)	3 (2.5)	6.9 [0.5-364.4]	0.09″
Osteoporosis	15 (5.6)	5 (4.2)	0.7 [0.2-2.2]	0.57″
Respiratory disease	12 (4.4)	2 (1.7)	0.4 [0.04-1.7]	0.24″
Renal disease	12 (4.4)	6 (5.0)	1.1 [0.3-3.4]	0.80′
Parkinson Disease	5 (1.9)	1 (0.8)	0.4 [ 0.01-4.1]	0.67″
Dementia	8 (3.0)	10 (8.4)	3.0 [1.0-9.0]	0.02′
Surgical histroy for musculoskeletal disorders	34 (12.6)	17 (14.3)	1.2 [0.6-2.2]	0.66′
Cervical OPLL (%)	84 (31.1)	43 (36.1)	1.3 [0.8-2.0]	0.33′
Cervical DISH (%)	23 (8.5)	14 (11.8)	1.4 [0.7-3.0]	0.31′
SI Changes on MRI (%)	219 (81.1)	101 (84.9)	1.5 [0.8-3.0]	0.22′
Steroid therapy for SCI (%)	60 (22.2)	19 (16.0)	0.7 [0.4-1.2]	0.16′
Surgical intervention (%)	136 (50.4)	71 (60.0)	1.5 [0.9-2.3]	0.09′
Baseline AIS grade (%)
A	Omit	Omit
B	7 (2.6)	7 (5.9)
C	91 (33.7)	44 (37.0)
D	172 (63.7)	68 (57.1)	-	0.19″″
Total AMS	70.9 ± 26.2	66.2 ± 27.5	-	0.10**
Upper extremity AMS	32.2 ± 13.6	30.6 ± 13.5	-	0.28**
Lower extremity AMS	38.9 ± 15.1	35.5 ± 16.5	-	0.06**

Means and standard deviations.

*unpaired t-test, **Wilcoxon rank-sum test, ***Welch’s t-test.

‘chi-squared test, “Fisher’s exact test.

BMI body mass index, HbA1c glycated hemoglobin levels, OPLL ossification of posterior longitudinal ligament, DISH diffuse idiopathic skeletal hyperostosis, MRI magnetic resonance imaging, SI signal intensity, SCI spinal cord injury, AIS American Spinal Injury Association impairment scale, AMS American spinal injury association motor score, OR odds ratio, CI confidence interval.

Comparison of Post-Injury Complications and Neurological Outcomes Between the Nondiabetic and Diabetic Groups

The mean total AMS at discharge in the diabetic group was significantly lower than that of the nondiabetic group (Table 2). Post-injury complications during hospitalization did not differ significantly between the groups. The AIS grade at 6 months post-injury was significantly poorer in the diabetic group than that in the nondiabetic group. The mean total, upper extremity, and lower extremity AMS in the diabetic group were significantly lower than those in the nondiabetic group.

Table 2.

Comparison of Post-injury Complications and Neurological Outcomes.

Variables	Non-diabetic (n = 270)	Diabetic (n = 119)	OR [95% CI]	P Value
Hospitalization (Days)	50.0 ± 68.3	65.3 ± 91.6	-	0.65**
Complications (%)
Neurological deterioration (motor)	5 (1.9)	1 (0.8)	0.4 [0.01-4.1]	0.67″
Neurological deterioration (sensory)	3 (1.1)	1 (0.8)	0.7 [0.01-9.4]	0.75″
Cerebral infarction	1 (0.4)	1 (0.8)	2.3 [0.03-178.0]	0.52″
Delirium	9 (3.3)	7 (5.9)	1.8 [0.6-5.6]	0.25′
Dysphagia	2 (0.7)	4 (3.4)	4.6 [0.6-51.6]	0.07″
Respiratory failure	0 (0)	1 (0.8)	Inf [0.1-Inf]	0.31″
Pulmonary embolism	0 (0)	1 (0.8)	Inf [0.1-Inf]	0.31″
Pneumonia	8 (3.0)	2 (1.7)	0.6 [0.1-2.8]	0.73″
Renal infection	23 (8.5)	8 (6.7)	0.8 [0.3-1.8]	0.53′
Time point of discharge
Total AMS	84.8 ± 17.3	78.1 ± 23.1	-	0.01**
6 months after the injury
AIS grade (%)
A	0 (0)	1 (0.8)
B	0 (0)	2 (1.7)
C	21 (7.8)	17 (14.3)
D	206 (76.3)	86 (73.1)
E	43 (15.9)	12 (10.1)	-	0.01″
Total AMS	88.9 ± 15.6	81.9 ± 21.8	-	1.1 × 10⁻³**
Improvement of total AMS	17.9 ± 20.9	15.7 ± 16.2	-	0.96**
Upper extremity AMS	43.0 ± 8.5	40.3 ± 10.3	-	0.01**
Improvement of upper extremity AMS	10.8 ± 11.6	9.6 ± 8.7	-	0.96**
Lower extremity AMS	45.9 ± 8.2	41.6 ± 12.6	-	7.8 × 10⁻⁵**
Improvement of lower extremity AMS	7.0 ± 12.3	6.1 ± 10.0	-	0.9**

Means and standard deviations.

*unpaired t-test, **Wilcoxon rank-sum test, ***Welch’s t-test.

‘chi-squared test, “Fisher’s exact test.

AIS American spinal injury association impairment scale, AMS American spinal injury association motor score, OR odds ratio, CI confidence interval.

Propensity Score–Matched Comparison of Baseline Characteristics and Comorbidities Between the Nondiabetic and Diabetic Groups

After propensity score matching, 96 patients were included in each group. The standardized difference in the moderator variables in the matched cohort was <10%. No significant differences in baseline characteristics or comorbidities were identified between the groups (Table 3).

Table 3.

Propensity Score-matched Comparison of Baseline Characteristics and Medical Comorbidities.

Variables	Non-diabetic (n = 96)	Diabetic (n = 96)	P Value	Standardized Difference
Demographic variables
Age (years)	74.7 ± 6.6	74.9 ± 6.2	0.70**	0.04
Sex (male, %)	73 (76.0)	74 (77.1)	1′	0.03
BMI	22.5 ± 3.1	22.3 ± 3.8	0.58***	0.05
Blood gluicose level (mg/dl)	125.6 ± 33.7	171.4 ± 64.9	8.8 × 10⁻⁸**	-
HbA1c (%)	-	7.1 ± 1.1	-
Medical comorbidities (%)
Hypertension	48 (50.0)	49 (51.0)	1′	0.02
Cardiovascular disease	12 (12.5)	13 (13.5)	1′	0.03
Cerebrovascular disease	6 (6.2)	6 (6.2)	1′	<0.001
Rheumatoid arthritis	1 (1.0)	1 (1.0)	1″	<0.001
Osteoporosis	5 (5.2)	5 (5.2)	1″	<0.001
Respiratory disease	2 (2.1)	2 (2.1)	1″	<0.001
Renal disease	6 (6.2)	4 (4.2)	0.75″	0.09
Parkinson Disease	2 (2.1)	1 (1.0)	1″	0.08
Dementia	6 (6.2)	4 (4.2)	0.75″	0.09
Surgical histroy for	12 (12.5)	12 (12.5)	1′	<0.001
Musculoskeletal disorders
Cervical OPLL (%)	33 (34.4)	35 (36.5)	0.88′	0.04
Cervical DISH (%)	11 (11.5)	11 (11.5)	1′	<0.001
SI Changes on MRI (%)	85 (88.5)	83 (86.5)	0.83′	0.06
Steroid therapy for SCI (%)	16 (16.7)	18 (18.8)	0.85′	0.06
Surgical intervention (%)	56 (58.3)	58 (60.4)	0.88′	0.04
Baseline AIS grade (%)
A	-	-
B	5 (5.2)	7 (7.3)
C	34 (35.4)	32 (33.3)
D	57 (59.4)	57 (59.4)	0.87″	0.09
Total AMS	65.8 ± 27.9	67.2 ± 28.0	0.7**	0.05
Upper extremity AMS	30.4 ± 13.3	31.3 ± 13.7	0.58**	0.07
Lower extremity AMS	35.4 ± 16.9	35.9 ± 16.6	0.8**	0.03

Means and standard deviations.

*unpaired t-test, **Wilcoxon rank-sum test, ***Welch’s t-test.

‘chi-squared test, “Fisher’s exact test.

Propensity Score–Matched Comparison of Post-Injury Complications and Neurological Outcomes Between the Nondiabetic and Diabetic Groups

No significant differences between the groups in terms of post-injury complications during hospitalization were observed. Similarly, the mean total AMS score at the time of discharge was not significantly different between the groups (Table 4). The AIS grade, mean total AMS, and upper and lower extremity AMS at 6 months post-injury were also not significantly different between the groups. The degree of improvement in the mean total, upper extremity, and lower extremity AMS at 6 months post-injury demonstrated no significant differences between the two groups.

Table 4.

Propensity Score-matched Comparison of Post-injury Complications and Neurological Outcomes.

Variables	Non-diabetic (n = 96)	Diabetic (n = 96)	OR [95% CI]	P Value
Hospitalization (Days)	46.5 ± 51.1	65.4 ± 94.1	-	0.86**
Complications (%)
Neurological deterioration (motor)	2 (2.1)	1 (1.0)	0.5 [0.01-9.6]	0.62″
Neurological deterioration (sensory)	1 (1.0)	1 (1.0)	1.0 [0.01-78.4]	1″
Cerebral infarction	0 (0)	1 (1.0)	Inf [0.03-Inf]	1″
Delirium	3 (3.2)	5 (5.2)	1.7 [0.3-11.1]	0.72″
Dysphagia	2 (2.1)	3 (3.1)	1.5 [0.2-18.3]	1″
Respiratory failure	0 (0)	1 (1.0)	Inf [0.03-Inf]	1″
Pulmonary embolism	0 (0)	1 (1.0)	Inf [0.03-Inf]	1″
Pneumonia	5 (5.2)	2 (2.1)	0.4 [0.04-2.4]	0.28″
Renal infection	9 (9.5)	6 (6.2)	0.6 [0.2-2.1]	0.43′
Time point of discharge
Total AMS	81.9 ± 18.9	78.5 ± 23.7	-	0.51**
6 months after the onset
AIS grade (%)
A	0 (0)	1 (1.0)
B	0 (0)	2 (2.1)
C	11 (11.5)	14 (14.6)
D	75 (78.1)	70 (72.9)
E	10 (10.4)	9 (9.4)	-	0.56**
Total AMS	86.4 ± 17.2	82.0 ± 22.5	-	0.28**
Improvement of total AMS	20.6 ± 21.7	14.9 ± 16.2	-	0.14**
Upper extremity AMS	41.9 ± 9.0	40.3 ± 10.8	-	0.39**
Improvement of upper extremity AMS	11.6 ± 10.9	9.0 ± 8.7	-	0.19**
Lower extremity AMS	44.5 ± 9.1	41.8 ± 12.7	-	0.17**
Improvement of lower extremity AMS	9.1 ± 14.1	5.9 ± 10.1	-	0.29**

Means and standard deviations.

*unpaired t-test, **Wilcoxon rank-sum test, ***Welch’s t-test.

‘chi-squared test, “Fisher’s exact test.

AIS American spinal injury association impairment scale, AMS American spinal injury association motor score, OR odds ratio, CI confidence interval.

Comparison of Baseline Characteristics, Medical Comorbidities, and Post-Injury Complications Between Nondiabetic And Moderate-Severe Diabetic Groups

Of the 119 patients with diabetes, 55 (46.2%) were categorized into the moderate-severe diabetic group. The prevalence rates of rheumatoid arthritis, dementia, cervical OPLL, and post-injury dysphagia in the moderate-severe diabetic group were significantly higher than those in the nondiabetic group (Table 5). The prevalence of steroid therapy for CSCI upon admission in the moderate-severe diabetic group was significantly lower than that in the nondiabetic group.

Table 5.

Comparison of Baseline Characteristics, Medical Comorbidities and Post-injury Complications Between Non-diabetic and Moderate-Severe Diabetic Groups.

Variables	Non-diabetic (n = 270)	Moderate-Severe Diabetic (n = 55)	OR [95% CI]	P Value
Demographic variables
Age (years)	74.4 ± 6.4	74.9 ± 6.2	-	0.46**
Sex (male, %)	188 (69.6)	45 (81.8)	2.0 [0.9-4.6]	0.10′
BMI	22.3 ± 3.3	22.1 ± 3.6	-	0.75*
Blood gluicose level (mg/dl)	125.1 ± 36.3	206.6 ± 84.9	-	5.9 × 10⁻¹⁵**
HbA1c (%)	-	7.8 ± 1.1	-	-
Medical comorbidities (%)
Hypertension	128 (47.4)	31 (56.4)	1.4 [0.8-2.7]	0.3′
Cardiovascular disease	37 (13.7)	8 (14.5)	1.1 [0.4-2.5]	1′
Cerebrovascular disease	20 (7.4)	4 (7.3)	1.0 [0.2-3.1]	1″
Rheumatoid arthritis	1 (0.4)	3 (5.5)	15.3 [1.2-810.3]	0.02″
Osteoporosis	15 (5.6)	2 (3.6)	0.6 [0.1-2.9]	0.75″
Respiratory disease	12 (4.4)	1 (1.8)	0.4 [0.01-2.8]	0.7″
Renal disease	12 (4.4)	2 (3.6)	0.8 [0.1-3.8]	1″
Parkinson Disease	5 (1.9)	0 (0)	0 [0-5.4]	0.59″
Dementia	8 (3.0)	7 (12.7)	4.7 [1.4-15.7]	5.4 × 10⁻³′
Surgical history for musculoskeletal disorders	34 (12.6)	7 (12.7)	1.0 [0.4-2.5]	1′
Cervical OPLL (%)	84 (31.1)	26 (47.3)	2.0 [1.1-3.7]	0.03′
Cervical DISH (%)	23 (8.5)	5 (9.1)	1.1 [0.3-3.1]	0.8″
SI Changes on MRI (%)	219 (81.1)	48 (87.3)	1.5 [0.6-4.2]	0.46′
Steroid therapy for CSCI (%)	60 (22.2)	4 (7.3)	0.3 [0.1-0.8]	9.0 × 10⁻³″
Surgical intervention (%)	136 (50.4)	31 (56.4)	1.3 [0.7-2.4]	0.46′
Baseline AIS grade (%)
A	Omit	Omit
B	7 (2.6)	2 (3.6)
C	91 (33.7)	20 (36.4)
D	172 (63.7)	33 (60.0)	-	0.73″
Total AMS	70.9 ± 26.2	67.9 ± 26.2	-	0.35**
Upper extremity AMS	32.2 ± 13.6	31.7 ± 13.0	-	0.71**
Lower extremity AMS	38.9 ± 15.1	36.3 ± 16.1	-	0.24**
Complications (%)
Neurological deterioration (motor)	5 (1.9)	1 (1.8)	1.0 [0.02-9.0]	1″
Neurological deterioration (sensory)	3 (1.1)	1 (1.8)	1.6 [0.03-20.8]	0.53″
Cerebral infarction	1 (0.4)	1 (1.8)	4.9 [0.1-388.2]	0.31″
Delirium	9 (3.3)	5 (9.1)	2.9 [0.7-10.0]	0.07″
Dysphagia	2 (0.7)	3 (5.5)	7.6 [0.9-93.2]	0.04″
Respiratory failure	0 (0)	0 (0)	-	-
Pulmonary embolism	0 (0)	0 (0)	-	-
Pneumonia	8 (3.0)	1 (1.8)	0.6 [0.01-4.7]	1″
Renal infection	23 (8.5)	2 (3.6)	0.4 [0.04-1.7]	0.28″

Means and standard deviations.

*unpaired t-test, **Wilcoxon rank-sum test, ***Welch’s t-test.

‘chi-squared test, “Fisher’s exact test.

Influence of Moderate-Severe Diabetes on Neurological Recovery in CSCI Without Bone Injury

To identify the influence of moderate-severe diabetes on the degree of improvement in the mean total AMS at 6 months post-injury, we conducted a multiple linear regression analysis. Adjusted R² was 0.52. Age, dementia, and baseline total AMS negatively influence neurological recovery at 6 months post-injury (Table 6). Blood glucose levels at admission or moderate-severe diabetes did not affect neurological recovery 6 months post-injury.

Table 6.

Multiple Linear Regression Analysis of the Degree of Improvement in Total AMS at 6 months After Injury.

	B	95% CI		VIF	P Value
	B	Lower	Upper	VIF	P Value
Demographic variables
Age	−0.34	−0.59	−0.08	1.12	0.01*
Sex	−1.25	−4.70	2.20	1.04	0.48
BMI	0.34	−0.13	0.82	1.04	0.16
Blood glucose level	0.01	−0.03	0.05	1.55	0.56
Medical comorbidities
Diabetes mellitus	−2.48	−7.56	2.60	1.64	0.34
Dementia	−16.50	−24.99	−8.01	1.17	1.7 × 10⁻⁴*
Rheumatoid arthritis	2.73	−15.28	20.75	1.09	0.77
Cervical OPLL	−1.10	−4.69	2.49	1.21	0.55
Cervical DISH	0.30	−5.43	6.03	1.11	0.92
SI Changes on MRI	−0.74	−5.17	3.69	1.14	0.74
Steroid therapy for SCI	2.94	−1.24	7.11	1.11	0.17
Surgical intervention	−1.02	−4.33	2.28	1.14	0.54
Baseline AIS grade [C]	7.39	−3.14	17.92	3.01	0.17
Baseline AIS grade [D]	8.04	−4.26	20.34	3.01	0.20
Baseline total AMS	−0.62	−0.72	−0.51	3.05	4.3 × 10⁻²⁵*
Complications
Dysphagia	−9.08	−20.94	2.78	1.17	0.13

*Statistically significant.

BMI body mass index, OPLL ossification of posterior longitudinal ligament, DISH diffuse idiopathic skeletal hyperostosis, MRI magnetic resonance imaging, SI signal intensity, SCI spinal cord injury, AIS American Spinal Injury Association impairment scale, AMS American spinal injury association motor score, VIF variance inflation factor.

Discussion

To the best of our knowledge, this is the first multicenter large-cohort study using propensity score matching to demonstrate the influence of diabetes on neurological recovery in patients with CSCI without bone injury. Propensity score matching revealed no significant differences in the degree of improvement in AMS at 6 months post-injury between the nondiabetic and diabetic groups. Furthermore, multiple linear regression analysis indicated that moderate-severe diabetes did not affect the degree of improvement in AMS. The aforementioned findings provide valuable insights for decision-making regarding the medical treatment of patients with CSCI without bone injury.

Influence of Diabetes on Neurological Outcomes in CSCI Without Bone Injury

Previous reports have emphasized that diabetes leads to poor neurological outcomes following surgical intervention in patients with lumbar degenerative disc disease.^11,12 However, the influence of diabetes on neurological recovery in patients with cervical spine disorders remains controversial. Kim et al reported that although diabetic patients with CSM could benefit from cervical laminoplasty, their rate of recovery was expected to be lower than those without diabetes.²⁸ Machino et al conducted a prospective cohort study of more than 500 patients with CSM and concluded that both diabetic and nondiabetic patients with CSM experienced similar benefits from cervical laminoplasty.¹⁴ Dokai et al and Nori et al noted that CSM patients with diabetes experienced improvements in neurological function as a result of posterior decompression surgery to the same extent as those without diabetes.^13,15 Although numerous studies have evaluated the influence of diabetes on neurological recovery in patients with CSM, few comparative studies have assessed this in patients with CSCI. Kobayakawa et al conducted a human cohort study of 206 patients with SCI, focusing on the relationship between blood glucose concentration on admission and functional outcomes, and supplemented by mouse model experiments.¹⁸ They reported that hyperglycemia on admission exacerbated secondary injury, resulting in poor functional outcomes after SCI, regardless of whether the patient had diabetes. In the present study, a propensity score matching demonstrated that AIS grade, AMS, and degree of improvement in AMS were not significantly different between two groups. Several possible factors influence the effects of diabetes on neurological recovery in the studies. First, the previous study included patients with CSCI as well as those with thoracic and lumbar SCI. Second, the sample size of the previous study was relatively small. Third, the previous study mainly focused on nondiabetic patients with SCI and hyperglycemia on admission. The study was strictly limited to patients with CSCI without bone injury and utilized a multicenter large-cohort analysis. We evaluated the differences in neurological recovery between the nondiabetic and diabetic groups using propensity score matching, which can accurately match baseline characteristics, including neurological status, on admission.

Influence of Moderate-Severe Diabetes on Neurological Recovery

There has been only one study has investigated relationship between severity of diabetes and neurological recovery in patients with SCI.¹⁸ The study reported that the HbA1c was negatively associated with AMS and recovery rate of AMS.¹⁸ However, the severity of diabetes is defined by not only HbA1c but also by the requirement for insulin treatment. In line with previous studies, including the Clinical Practice Guideline for Diabetes, we classified patients with an HbA1c ≥ 7.0%, which is the treatment target for preventing diabetes-related complications, and those requiring insulin treatment as having moderate-severe diabetes.^26,27 Multiple linear regression analysis revealed that age on admission, preexisting dementia, and baseline total AMS negatively influenced neurological recovery, whereas blood glucose level and moderate-severe diabetes did not impact neurological recovery. These results indicated that tight glycemic control by diabetologists during the acute phase of CSCI might lead to better neurological outcomes even in moderate-severe diabetic patients who not only had conservative treatment but also underwent the surgery. Several reports have suggested that early surgery for CSCI is beneficial, and we believe that clinicians should not consider the sufficient justification to deny the spine surgery for CSCI with diabetes alone.^5,29 The present study did not evaluate the transition in blood glucose concentration during post-injury hospitalization. Further study is mandatory to reveal the impact of tight glycemic control during the acute phase of CSCI on neurological recovery in patients with CSCI. Age is widely recognized as a negative prognostic factor for neurological recovery in patients with SCI.³⁰ Jakob et al reported that elderly SCI patients have difficulties in translating an improvement of neurological deficit into function even after discharge from the rehabilitation center.³¹ Clinicians should consider the personalized rehabilitation approaches which focus on training of daily living activities and ensure that patients are motivated to apply the skills they have acquired. A systematic review of cognitive function after SCI demonstrated a strong correlation between cognitive impairment and SCI and identified cognitive impairment as a predictor of poor social participation, including rehabilitation post-discharge.^32,33 In patients with dementia, the quality of rehabilitation might decline, which could contribute to reduced neurological improvement. Systematic literature review based on prospective studies also revealed that diabetes mellitus is likely to increase the risk of cognitive impairment such as Alzheimer’s disease, which has frequently been attributed to cerebrovascular disease.³⁴ While diabetes can be managed with the tight glycemic control, interventions for dementia are often more challenging. Therefore, a multidisciplinary CSCI care approach should focus on preventing prolonged immobility and complications such as infections (pneumonia, renal infection and surgical site infection) that can exacerbate preexisting cognitive impairment. Nakajima et al reported similar findings, asserting that the AIS grade on admission and the presence of dementia/delirium were independent prognostic factors for walking recovery (for patients whose baseline AIS grades A-C converted to AIS grades D-E) among patients with CSCI without bone injury, using a comparable patient population from the same original dataset,⁷ and they concluded that diabetes did not influence neurological recovery in patients with CSCI without bone injury. Conversely, Nori et al, utilizing a similar patient population from the same original dataset, applied multiple linear regression analysis, which revealed that diabetes adversely impacted postoperative changes in total AMS among patients with CSCI without bone injury who had undergone surgery.²⁰ The discrepancy in results between the studies may be attributable to different patient selection criteria. Specifically, the present study and that of Nakajima et al included patients with CSCI without bone injury who underwent both surgery and conservative treatment, whereas the study of Nori et al focused solely on those who underwent surgery. The findings suggest that patients with CSCI without bone injury who have undergone surgery are more likely to develop moderate-severe diabetes.

Limitations and Strength

The primary limitation of this study is its retrospective nature; thus, the study has some missing values. The present study had the retrospective nature and the analyses were conducted using a heterogeneous cohort without a control group, which was not a comprehensive survey. Since the clinical data were collected from 33 high-volume trauma centers, it was difficult to obtain the completely accurate medical records. The present study excluded patients with missing values in diabetes, baseline AIS grade, baseline AMS, AIS grade at 6 months after injury, or AMS at 6 months after injury, which can cause the selection bias. The present study excluded the patients with baseline AIS grade A as well even though that could cause selection bias. The previous study demonstrated the probability of the AIS grade converting to grades C–E is very low in SCI patients with AIS grade A.²³ We excluded 23 CSCI patients without bone injury who were classified into AIS grade A on admission, and 83% of them were AIS grade A or B at final follow-up. This result was comparable with the previous study and indicated that it was difficult to evaluate the neurological recovery especially motor function. Furthermore, it was difficult to completely compare treatment outcomes due to the biases that the surgical indications and procedures, and rehabilitation approaches differed across the facilities and there was no standardized treatment approach. Additionally, the survey was conducted in Japan, which has the unique background of being the most advanced aging society in the world. However, as the pace of population aging has been accelerating drastically worldwide, the results of the present study might be generalized beyond the specific population. Regarding the missing values, smoking and drinking histories, which could potentially influence the neurological outcomes, were excluded from the analysis due to the high proportion of missing data. Although the proportion of missing values in other demographic variables that evaluated in the present study were all less than 5% and pairwise deletion was used in each statistical analysis, that could potentially cause the bias. Preexisting conditions such as CSM and SI changes in the spinal cord on MRI have not evaluated because of the retrospective nature of the present study. Evaluation of cervical and global spinal alignments, which might have potentially impact on the neurological outcomes, was not performed. Follow-up period was relatively short at 6 months post-injury to evaluate cases that show delayed recovery following CSCI. It is common for CSCI patients to be transferred to the rehabilitation hospitals immediately after acute-phase treatment in Japan, resulting in a high proportion of missing AMS at 12 months post-injury (approximately 30%), which also had to be excluded from the analysis. Neurological outcomes at 6-month post-injury were chosen, since the previous clinical trials demonstrated that neurological recovery after traumatic SCI mainly occurs within the first 6-9 months.^24,25 Further studies are mandatory to determine the influence of diabetes on long-term neurological recovery in patients with CSCI without bone injury. The transition in blood glucose concentration during post-injury hospitalization was not observed in the present study. Kobayakawa et al reported that controlling the blood glucose concentration during the acute phase of SCI could prevent exacerbation of the pathophysiology and improve motor function in hyperglycemic mice post-SCI.¹⁸ Tight glycemic control by diabetologists during the acute phase of CSCI might lead to better neurological outcomes in diabetic patients with CSCI without bone injury in the present study. Despite these limitations, this is the first multicenter large-cohort study to demonstrate the influence of diabetes on neurological recovery in patients with CSCI without bone injury.

Conclusion

We investigated the influence of diabetes, particularly moderate-severe diabetes, on neurological recovery in patients with CSCI without bone injury. Diabetic patients with CSCI without bone injury experienced improvements in neurological function comparable to those of nondiabetic patients. Furthermore, moderate-severe diabetes did not affect neurological recovery.

Footnotes

Acknowledgments

The authors acknowledge the contributions of the members of the 33 participating institutions in the assistance with data collection.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical Statement

ORCID iDs

Kazuki Takeda

Kota Watanabe

Takeshi Sasagawa

Hiroaki Nakashima

Naoki Segi

Atsushi Yunde

Hideaki Nakajima

Tomohiro Yamada

Hidenori Suzuki

Yasuaki Imajo

Shota Ikegami

Ko Hashimoto

Hidetomi Terai

Gen Inoue

Yasushi Oshima

Toshitaka Yoshii

Masayuki Ishihara

Shiro Imagama

References

Miyakoshi

Suda

Kudo

, et al. A nationwide survey on the incidence and characteristics of traumatic spinal cord injury in Japan in 2018. Spinal Cord. 2021;59(6):626-634. doi:10.1038/s41393-020-00533-0

Ishida

Tominaga

. Predictors of neurologic recovery in acute central cervical cord injury with only upper extremity impairment. Spine. 2002;27(15):1652-1658. doi:10.1097/00007632-200208010-00011.

Kato

Kimura

Sasaki

, et al. Cervical spinal cord injury without bony injury: a multicenter retrospective study of emergency and critical care centers in Japan. J Trauma. 2008;65(2):373-379. doi:10.1097/TA.0b013e31817db11d

Oichi

Oshima

Okazaki

Azuma

. Preexisting severe cervical spinal cord compression is a significant risk factor for severe paralysis development in patients with traumatic cervical spinal cord injury without bone injury: a retrospective cohort study. Eur Spine J. 2016;25(1):96-102. doi:10.1007/s00586-015-4142-4

OSCIS investigators Chikuda

Koyama

, et al. Effect of early vs delayed surgical treatment on motor recovery in incomplete cervical spinal cord injury with preexisting cervical stenosis: a randomized clinical trial. JAMA Netw Open. 2021;4(11):e2133604. doi:10.1001/jamanetworkopen.2021.33604

Uribe

Green

Vanni

Moza

Guest

Levi

. Acute traumatic central cord syndrome--experience using surgical decompression with open-door expansile cervical laminoplasty. Surg Neurol. 2005;63(6):505-510. doi:10.1016/j.surneu.2004.09.037

Nakajima

Yokogawa

Sasagawa

, et al. Prognostic factors for cervical spinal cord injury without major bone injury in elderly patients. J Neurotrauma. 2022;39(9-10):658-666. doi:10.1089/neu.2021.0351

Browne

Cook

Pietrobon

Bethel

Richardson

. Diabetes and early postoperative outcomes following lumbar fusion. Spine. 2007;32(20):2214-2219. doi:10.1097/BRS.0b013e31814b1bc0

Golinvaux

Varthi

Bohl

Basques

Grauer

. Complication rates following elective lumbar fusion in patients with diabetes: insulin dependence makes the difference. Spine. 2014;39(21):1809-1816. doi:10.1097/BRS.0000000000000506

10.

Tang

Zhu

Wang

Xie

Fei

. Risk factors for postoperative complication after spinal fusion and instrumentation in degenerative lumbar scoliosis patients. J Orthop Surg Res. 2014;9(1):15. doi:10.1186/1749-799X-9-15

11.

Airaksinen

Herno

Turunen

Saari

Suomlainen

. Surgical outcome of 438 patients treated surgically for lumbar spinal stenosis. Spine. 1997;22(19):2278-2282. doi:10.1097/00007632-199710010-00016

12.

Arinzon

Adunsky

Fidelman

Gepstein

. Outcomes of decompression surgery for lumbar spinal stenosis in elderly diabetic patients. Eur Spine J. 2004;13(1):32-37. doi:10.1007/s00586-003-0643-7

13.

Dokai

Nagashima

Nanjo

Tanida

Teshima

. Surgical outcomes and prognostic factors of cervical spondylotic myelopathy in diabetic patients. Arch Orthop Trauma Surg. 2012;132(5):577-582. doi:10.1007/s00402-011-1449-4

14.

Machino

Yukawa

Ito

, et al. Impact of diabetes on the outcomes of cervical laminoplasty: a prospective cohort study of more than 500 patients with cervical spondylotic myelopathy. Spine. 2014;39(3):220-227. doi:10.1097/BRS.0000000000000102

15.

Nori

Nagoshi

Yoshioka

, et al. Diabetes does not adversely affect neurological recovery and reduction of neck pain after posterior decompression surgery for cervical spondylotic myelopathy: results from a retrospective multicenter study of 675 patients. Spine. 2021;46(7):433-439. doi:10.1097/BRS.0000000000003817

16.

Kawaguchi

Matsui

Ishihara

Gejo

Yasuda

. Surgical outcome of cervical expansive laminoplasty in patients with diabetes mellitus. Spine. 2000;25(5):551-555. doi:10.1097/00007632-200003010-00004

17.

Machino

Yukawa

Ito

, et al. Risk factors for poor outcome of cervical laminoplasty for cervical spondylotic myelopathy in patients with diabetes. J Bone Joint Surg Am. 2014;96(24):2049-2055. doi:10.2106/JBJS.N.00064

18.

Kobayakawa

Kumamaru

Saiwai

, et al. Acute hyperglycemia impairs functional improvement after spinal cord injury in mice and humans. Sci Transl Med. 2014;6(256):256ra137. doi:10.1126/scitranslmed.3009430

19.

Chen

Guo

Sun

Jin

. Hyperglycemia aggravates spinal cord injury through endoplasmic reticulum stress mediated neuronal apoptosis, gliosis and activation. Biomed Pharmacother. 2019;112:108672. doi:10.1016/j.biopha.2019.108672

20.

Nori

Watanabe

Takeda

, et al. Influence of the timing of surgery for cervical spinal cord injury without bone injury in the elderly: a retrospective multicenter study. J Orthop Sci. 2023;29:480-485. doi:10.1016/j.jos.2023.01.004

21.

Nori

Watanabe

Takeda

, et al. Does surgery improve neurological outcomes in older individuals with cervical spinal cord injury without bone injury? A multicenter study. Spinal Cord. 2022;60(10):895-902. doi:10.1038/s41393-022-00818-6

22.

Maeda

Ueta

Mori

, et al. Soft-tissue damage and segmental instability in adult patients with cervical spinal cord injury without major bone injury. Spine. 2012;37(25):E1560-E1566. doi:10.1097/BRS.0b013e318272f345

23.

Kirshblum

Snider

Eren

Guest

. Characterizing natural recovery after traumatic spinal cord injury. J Neurotrauma. 2021;38(9):1267-1284. doi:10.1089/neu.2020.7473

24.

Bracken

Shepard

Holford

, et al. Administration of methylprednisolone for 24 or 48 hours or tirilazad mesylate for 48 hours in the treatment of acute spinal cord injury. Results of the third national acute spinal cord injury randomized controlled trial. National acute spinal cord injury study. JAMA. 1997;277(20):1597-1604.

25.

Geisler

Coleman

Grieco

Poonian

Sygen Study Group . The Sygen multicenter acute spinal cord injury study. Spine. 2001;26(24 Suppl):S87-S98. doi:10.1097/00007632-200112151-00015

26.

Araki

Goto

Kondo

, et al. Japanese clinical practice guideline for diabetes 2019. J Diabetes Investig. 2020;11(4):1020-1076. doi:10.1111/jdi.13306

27.

Mori

Nagata

Fujimoto

Fujino

Mori

. Diabetes severity measured by treatment control status and number of anti-diabetic drugs affects presenteeism among workers with type 2 diabetes. BMC Publ Health. 2021;21(1):1865. doi:10.1186/s12889-021-11913-3

28.

Kim

Moon

Kim

, et al. Diabetes and smoking as prognostic factors after cervical laminoplasty. J Bone Joint Surg Br. 2008;90(11):1468-1472. doi:10.1302/0301-620X.90B11.20632

29.

Inoue

Suzuki

Endo

Uenohara

Tominaga

. Efficacy of early surgery for neurological improvement in spinal cord injury without radiographic evidence of trauma in the elderly. World Neurosurg. 2017;105:790-795. doi:10.1016/j.wneu.2017.06.070

30.

Burns

Golding

Rolle

Jr. Graziani

Ditunno

. Recovery of ambulation in motor-incomplete tetraplegia. Arch Phys Med Rehabil. 1997;78(11):1169-1172. doi:10.1016/s0003-9993(97)90326-9

31.

Jakob

Wirz

van Hedel

Dietz

EM-SCI Study Group . Difficulty of elderly SCI subjects to translate motor recovery--“body function”--into daily living activities. J Neurotrauma. 2009;26(11):2037-2044. doi:10.1089/neu.2008.0824

32.

Craig

Nicholson Perry

Guest

Tran

Middleton

. Adjustment following chronic spinal cord injury: determining factors that contribute to social participation. Br J Health Psychol. 2015;20(4):807-823. doi:10.1111/bjhp.12143

33.

Sachdeva

Gao

Chan

CCH

Krassioukov

. Cognitive function after spinal cord injury: a systematic review. Neurology. 2018;91(13):611-621. doi:10.1212/WNL.0000000000006244

34.

Kopf

Frolich

. Risk of incident Alzheimer’s disease in diabetic patients: a systematic review of prospective trials. J Alzheimers Dis. 2009;16(4):677-685. doi:10.3233/JAD-2009-1011

Influence of Diabetes Mellitus on Neurological Recovery in Older Patients With Cervical Spinal Cord Injury Without Bone Injury: A Retrospective Multicenter Study

Abstract

Study Design

Objectives

Methods

Results

Conclusions

Keywords

Introduction

Materials and Methods

Participants

Data Collection

Statistical Analysis

Results

Comparison of Baseline Characteristics and Medical Comorbidities Between the Nondiabetic and Diabetic Groups

Comparison of Post-Injury Complications and Neurological Outcomes Between the Nondiabetic and Diabetic Groups

Propensity Score–Matched Comparison of Baseline Characteristics and Comorbidities Between the Nondiabetic and Diabetic Groups

Propensity Score–Matched Comparison of Post-Injury Complications and Neurological Outcomes Between the Nondiabetic and Diabetic Groups

Comparison of Baseline Characteristics, Medical Comorbidities, and Post-Injury Complications Between Nondiabetic And Moderate-Severe Diabetic Groups

Influence of Moderate-Severe Diabetes on Neurological Recovery in CSCI Without Bone Injury

Discussion

Influence of Diabetes on Neurological Outcomes in CSCI Without Bone Injury

Influence of Moderate-Severe Diabetes on Neurological Recovery

Limitations and Strength

Conclusion

Footnotes

Acknowledgments

Declaration of Conflicting Interests

Funding

Ethical Statement

ORCID iDs

References