OBJECTIVE—Magnetic resonance spectroscopy (MRS) (specifically, 1H-MRS) has been used to show changes in the brain following peripheral nerve injury in subjects without diabetes. This study used 1H-MRS to examine the brain in subjects with or without painful diabetic neuropathy.
RESEARCH DESIGN AND METHODS—Twenty-six diabetic subjects (12 with and 14 without chronic neuropathic pain) were compared, with 18 subjects without diabetes and pain. The left thalamus, anterior cingulate cortex (ACC), and dorsolateral prefrontal cortex (DLPFC) were assessed using 1H-MRS.
RESULTS—In the DLPFC, diabetic subjects had a decrease in N-acetyl aspartate (NAA) and creatine relative to the control group. In the thalamus, there was a reduction of NAA in the diabetic group with pain compared with that in patients with diabetes and no pain.
CONCLUSION—Subjects with diabetes have metabolite differences in the brain compared with control subjects. Subjects with painful neuropathy showed reduced NAA in the thalamus, which may explain the genesis of pain in some cases.
Pain due to diabetic neuropathy can be severe and disabling. Despite numerous studies of the peripheral nervous system, the genesis of this type of neuropathic pain remains poorly understood. Studies of the central nervous system after peripheral nerve injury have demonstrated biochemical and structural changes, including abnormal firing of thalamic neurons (1) and changes in metabolite concentrations in the thalamus (2). Thus, the central nervous system may be important in the genesis of pain in diabetic neuropathy.
Magnetic resonance spectroscopy (MRS) provides an excellent tool to study the central nervous system. However, MRS studies in diabetes are limited (3,4), and none have focused directly on painful neuropathy. It is therefore the purpose of the current study to use MRS to provide further information on diabetes, particularly in the context of painful neuropathy.
RESEARCH DESIGN AND METHODS—
Twelve diabetic subjects with chronic neuropathic pain of more than 6 months’ duration were matched for age, sex, type of diabetes, and glycemic control with 14 patients without pain. They were compared with 18 subjects without diabetes and pain. Subjects were excluded from the study if they had acute or chronic pain not associated with diabetic neuropathy. All subjects had a blood glucose level ≥4.0 mmol/l before scanning. Vibration perception threshold (VPT) was measured using a biothesiometer (Bio-Medical Instrument, Newbury, OH). Pain was measured on a 10-cm visual analog scale.
The left thalamus, anterior cingulate cortex (ACC), and dorsolateral prefrontal cortex (DLPFC) were assessed using 1H-MRS. Magnetic resonance examinations were performed using a 1.5-T magnetic resonance image scanner and standard quadrature head coil. Following image-guided voxel placement, water-suppressed and -unsuppressed 1H-MRS data were collected using a stimulated echo acquisition mode sequence (5) with chemical shift–selected water suppression. Acquisition parameters were TE/TR 25/1,500 ms, voxel size 8 cm3, 256 signal acquisitions, spectral width 2,500 Hz, and 2,000 data points.
Interpretation of spectra was performed using the java-based magnetic resonance user interface (jMRUI, version 2.0) (6) and a nonlinear least-squares algorithm (AMARES). Metabolites are expressed relative to the water resonance of the water-unsuppressed spectra. Differences between groups were calculated using ANOVA with a Levenes test for equality of variance with post hoc independent Student's t test. Data are presented as median (interquartile range) or mean ± SD.
RESULTS—
Clinical profiles of the 18 control and 26 diabetic subjects and their pain scores are shown in Table 1. VPT scores were higher in the pain group, and this was the only difference between diabetes groups (t = −2.2; P = 0.04). In the DLPFC, diabetes independent of pain was associated with a significant decrease in both NAA and creatine relative to the control group. In the thalamus, there was a significant reduction of N-acetyl aspartate (NAA) in the diabetic group with pain compared with that in the group with diabetes and no pain. The diabetes and no-pain group also showed an elevation in creatine in the thalamus relative to that in the control group. There were no differences found in the anterior cingulate cortex.
CONCLUSIONS—
Our study demonstrates that there are significant changes in brain metabolites in pain-processing regions of the brain associated with diabetes and with neuropathic pain. Indiviuals with diabetes demonstrated a marked reduction in NAA and creatine in the DLPFC, independent of pain. This observation demonstrates that diabetes is associated with altered cerebral metabolites, although by themselves they could not explain the presence of pain. Within the diabetic cohort, subjects with pain had significantly reduced NAA in the thalamus compared with the no-pain group. This novel finding of a reduction in thalamic NAA in subjects with pain due to diabetes is consistent with findings from studies on other types of neuropathic pain using MRS in individuals without diabetes (2,7). The reduction in thalamic NAA is likely to represent abnormal neuronal activity and oxidative energy transmission. With relevance to neuropathic pain, a reduction in NAA in the thalamus may be responsible for altered amplification and perception of the pain signal in people with diabetes.
It should be noted that usual pain medications were not withdrawn before scanning. However, not all of the subjects with pain used medication, and the reduction in NAA in the diabetes group with pain occurred irrespective of use (or lack thereof) of pain medication. The control group were also younger than the diabetic group; however, there was no significant age difference between the two diabetic cohorts to explain the association of pain with reduced NAA. Those with painful neuropathy had a higher VPT than those without pain. However, many patients among the no-pain group have both high VPT and NAA. Thus, it is more likely that the presence of pain rather than sensory loss accounts for the difference in NAA in the two diabetic groups.
In summary, MRS has identified metabolite differences in the brain in subjects with diabetes compared with control subjects. We also demonstrated reduced NAA in the thalamus of diabetic subjects with painful neuropathy. Further studies are required to determine whether this helps explain the genesis of pain in some, but not all, patients with diabetic neuropathy.
Clinical profile of subjects and concentrations of metabolites in the DLPFC, ACC, and thalamus brain regions
. | Control subjects . | Diabetes, no pain . | Diabetes, pain . |
---|---|---|---|
n | 18 | 14 | 12 |
Age (years) | 48 (33–60)* | 57 (53–63) | 61 (56–68) |
Diabetes duration (years) | N/A | 13.5 (8.8–27) | 15 (10.3–22.8) |
Sex (male) | 9 | 13 | 15 |
Type 2 diabetic subjects | N/A | 9 | 10 |
A1C (%) | N/A | 7.7 (6.7–8.9) | 7.5 (6.8–8.6) |
Vibration perception threshold (volts) | N/A | 26.6 (16–41) | 37.5 (26–50)† |
Pain score 10 cm VAS | N/A | 0 | 6 ± 2.45 |
DLPFC | |||
Choline | 218 ± 36 | 204 ± 31 | 207 ± 35 |
Creatine | 172 ± 22 | 152 ± 26‡ | 148 ± 16‡ |
NAA | 128 ± 10 | 108 ± 23‡ | 113 ± 8‡ |
ACC | |||
Choline | 247 ± 58 | 237 ± 41 | 249 ± 46 |
Creatine | 182 ± 30 | 189 ± 38 | 182 ± 17 |
NAA | 127 ± 28 | 127 ± 17 | 132 ± 15 |
Thalamus | |||
Choline | 195 ± 26 | 201 ± 36 | 194 ± 29 |
Creatine | 169 ± 13 | 184 ± 16‡ | 172 ± 15 |
NAA | 108 ± 12 | 117 ± 19‡ | 98 ± 13† |
. | Control subjects . | Diabetes, no pain . | Diabetes, pain . |
---|---|---|---|
n | 18 | 14 | 12 |
Age (years) | 48 (33–60)* | 57 (53–63) | 61 (56–68) |
Diabetes duration (years) | N/A | 13.5 (8.8–27) | 15 (10.3–22.8) |
Sex (male) | 9 | 13 | 15 |
Type 2 diabetic subjects | N/A | 9 | 10 |
A1C (%) | N/A | 7.7 (6.7–8.9) | 7.5 (6.8–8.6) |
Vibration perception threshold (volts) | N/A | 26.6 (16–41) | 37.5 (26–50)† |
Pain score 10 cm VAS | N/A | 0 | 6 ± 2.45 |
DLPFC | |||
Choline | 218 ± 36 | 204 ± 31 | 207 ± 35 |
Creatine | 172 ± 22 | 152 ± 26‡ | 148 ± 16‡ |
NAA | 128 ± 10 | 108 ± 23‡ | 113 ± 8‡ |
ACC | |||
Choline | 247 ± 58 | 237 ± 41 | 249 ± 46 |
Creatine | 182 ± 30 | 189 ± 38 | 182 ± 17 |
NAA | 127 ± 28 | 127 ± 17 | 132 ± 15 |
Thalamus | |||
Choline | 195 ± 26 | 201 ± 36 | 194 ± 29 |
Creatine | 169 ± 13 | 184 ± 16‡ | 172 ± 15 |
NAA | 108 ± 12 | 117 ± 19‡ | 98 ± 13† |
Results are n, median (interquartile range), or means ± SD.
Different to diabetic group, P < 0.001;
Different to control group, P = 0.04;
Diabetes, pain different to no pain group, P < 0.01. VAS, visual analog scale.
Article Information
This study was supported by Australian and New Zealand College of Anesthetists Grant 03/001.
We thank Dr. Anas Natfaji for assistance with data collection and recruitment of patients, Associate Prof. Carolyn Mountford for guidance with MRS, Peter Stanwell for performing the MRS, and Prof. Lindy Rae for review of the results and expert comments on neural biochemistry.
References
Published ahead of print at http://care.diabetesjournals.org on 25 February 2008. DOI: 10.2337/dc07-2088.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.