J Am Board Fam Pract: first published as 10.3122/jabfm.17.2.127 on 13 April 2004. Downloaded from

BRIEF REPORTS Peripheral Neuropathy and “Borderline”

Neil R. Holland, MB, BS, and Calin I. Prodan, MD

There is a clear association between sensory poly- Table 1. Classification of Abnormal neuropathy and overt diabetes mellitus. The inci- Metabolism Based on the 2-Hour Oral Glucose dence of neuropathy increases with duration of Tolerance Test, Which Consists of Fasting Blood diabetes; ultimately, up to 50% of patients are Glucose Level and Blood Glucose Level Measured 2 affected.1,2 Improved glycemic control has been Hours after a 75-g Oral Glucose Load4 shown to prevent and delay progression of diabetic Fasting 2-Hour Post neuropathy,3 emphasizingthe importance of early Glucose Load Glucose diagnosis and aggressive management in these pa- Diagnosis (mg/dL) (mg/dL) tients. There is increasingevidence that patients Normal glucose Ͻ110 and Ͻ140 with milder degrees of abnormal glucose metabo- metabolism lism, includingimpaired glucosetolerance (IGT) Impaired fasting 110–125 and Ͻ140 glucose and impaired fastingglucose(formerly referred to Impaired glucose Ͻ126 and 140–199 as “borderline” diabetes), are also at risk for devel- tolerance opingsymptomatic polyneuropathy (Tables 1 and Diabetes mellitus Ն126 or Ն200 2).4,5 We have personally observed patients for sev- eral years with “idiopathic” painful sensory neurop- athy who subsequently develop overt diabetes; in normal. Blood work, includingfastingglucose,gly- retrospect, it seems likely that their symptoms were cohemoglobin, and vitamin B12 levels and serum related to longstanding impaired glucose metabo- protein electrophoresis, was normal. She did not lism. This interestingassociation calls for further undergo an oral (OGTT). investigation with prospective studies. She was diagnosed with idiopathic small-fiber neu-

ropathy. Her symptoms were treated first with nor- http://www.jabfm.org/ Case Reports triptyline and then with gabapentin. Her pain con- Case 1 tinued to progress, and over the course of the next This 67-year-old woman presented with a 3-year year, she was also takingtramadol, replaced within history of numbness and burningpain in her feet. another year by oxycodone. One year later, 5 years Initially, she noticed numbness in the toes associ- after the onset of her initial complaints, she had a ated with burningpain and . With vitreous hemorrhage and was found to have pro- time, these symptoms slowly intensified, and pro- liferative retinopathy. She underwent a 2-hour on 27 September 2021 by guest. Protected copyright. gressed to involve her legs to the knees and into the OGTT. The fasting glucose was normal at 99 mg/ fingertips. She also complained of nocturnal leg dL, but the 2-hour glucose was elevated at 228 pain, cramping, and restlessness. Her examination mg/dL, satisfying diagnostic criteria for diabetes, showed diminished sensitivity to pinprick and heat and she was started on metformin. She continued pain over both feet up to the knee and the distal to take low-dose oxycodone and gabapentin but her fingertips. Vibratory sensitivity was mildly reduced neuropathic symptoms stabilized. at the great toe but normal at the ankle. The ankle jerks were present. Electrodiagnostic testing was Table 2. Age-Adjusted Prevalence of Neuropathy in Patients with Altered Glucose Metabolism Compared 5 Submitted, revised, 30 July 2003. with That of Control Subjects From West LongBranch, New Jersey (NRH) and the Department of Neurology, University of Oklahoma Medical Normal Glucose Metabolism 3.9% Center, Oklahoma City (CIP). Address correspondence to Impaired Glucose Tolerance 11.2% Dr. Neil R. Holland, 107 Monmouth Road, West Long Diabetes Mellitus 25.8% Branch, NJ 07764 (e-mail: [email protected]).

Neuropathy and Diabetes 127 J Am Board Fam Pract: first published as 10.3122/jabfm.17.2.127 on 13 April 2004. Downloaded from

Case 2 from the right ankle, which was disproportionately This 65-year-old obese woman had a history of weak. He had patchy areas of sensory loss, includ- chronic low back pain. Magentic resonance imag- ing the right shoulder, left forearm, and right lat- ingof the lumbar spine had shown multilevel de- eral calf. Reflexes were absent. Electrodiagnostic generative changes, and she had been treated with testingshowed asymmetric motor and sensory ax- intermittent facet and epidural steroid injections. onal loss. Magnetic resonance imaging and angiog- She presented with a 6-year history of burningpain raphy of the brain, performed for the suspicion of and numbness affectingthe soles of both feet that vasculitis, were normal. Repeat lumbar puncture was unresponsive to these treatments. Her exami- showed only the same albuminocytologic dissocia- nation showed reduced sensitivity to pinprick, tem- tion. He ultimately underwent a nerve and muscle perature and pain affectingboth feet up to the level biopsy, which showed reduced densities of large of the ankles. Vibratory sensitivity, strength and and small nerve fibers and denervation atrophy. reflexes were normal. Electrodiagnostic testing was There was no histologic evidence of inflammation normal. Skin biopsy showed epidermal denerva- or vasculitis. He was started on gabapentin and tion. Blood work, includingfastingglucose,glyco- sildenafil, and tapered off the prednisone. He was hemoglobin, and vitamin B12 levels, syphilis serol- reevaluated for diabetes duringthe discontinuation ogy, and serum protein electrophoresis, was of the prednisone, and glycohemoglobin was found normal. She subsequently underwent a 2-hour to be elevated at 6.4% (reference value, Յ5.7%). OGTT, which showed a normal fastingglucoseof He was believed to have developed steroid-induced 101 mg/dL but elevated 2-hour glucose at 188 diabetes. Three months after discontinuingthe mg/dL, indicating impaired glucose tolerance. She prednisone, his glycohemoglobin was back within was started on a diabetic diet, exercise program, the normal range (5.6%), but a 2-hour OGTT and gabapentin. When re-evaluated 3 months later, showed a borderline normal fastingglucoseat 108 her symptoms were improved, and her OGTT was mg/dL, and a 2-hour glucose that was elevated at now normal (2-hour glucose, 108 mg/dL). 180 mg/dL, consistent with impaired glucose tol- erance. Three months after startinga diabetic diet, Case 3 his OGTT was within normal limits (fastingglu- This 54 year-old man presented with an 18-month cose, 102 mg/dL; 2-hour glucose, 123 mg/dL), and history of multiple mononeuropathies. He first de- his neurologic deficits had stabilized. The initial veloped right-sided foot drop, followed by right presentation was consistent with mononeuropathy http://www.jabfm.org/ hand numbness and weakness, and then weakness multiplex, includingpartial third nerve palsy. He at the left ankle. An extensive battery of blood work went on to develop a painful polyneuropathy. He was performed at the referringfacility, including was ultimately found to have impaired glucose tol- fastingglucose,vitamin B 12 level, syphilis serology erance, perhaps exacerbated by steroids. Both and serum protein electrophoresis and glycohemo- mononeuropathy multiplex and albuminocytologic globin; results of all tests were normal. He did not dissociation are seen with diabetes, and in retro- on 27 September 2021 by guest. Protected copyright. undergo an OGTT at that time. Lumbar puncture spect, impaired glucose metabolism may have been showed elevated protein at 122 mg/dL (reference the cause of his symptoms all along. range, 15 to 45 mg/dL) and white blood cell count of 3/mm3 (reference range, 0–5/mm3). These re- sults were interpreted as albuminocytologic disso- Discussion ciation supportive of a diagnosis of chronic inflam- The cause of polyneuropathy remains unknown in matory demyelinatingneuropathy, and he was more than 20% of cases despite extensive labora- placed on 60 mg/day oral prednisone for 1 month. tory testing.6 The incidence of idiopathic cases is Nevertheless, his condition continued to worsen, believed to be even higher among patients present- and he noted increasingburningpain and pares- ingwith painful sensory neuropathy. 7,8 However, thesias as well as erectile dysfunction. On examina- some patients with “idiopathic” painful sensory tion, he had mild right-sided ptosis with mild weak- neuropathy have ultimately developed overt diabe- ness of right medial rectus and an enlarged right tes mellitus after many years of follow-up, suggest- pupil. There was mild wasting of the right thigh ingthat occult impairment of glucosemetabolism and calf, with mild symmetric distal weakness, aside may have been the cause of their symptoms all

128 JABFP March–April 2004 Vol. 17 No. 2 J Am Board Fam Pract: first published as 10.3122/jabfm.17.2.127 on 13 April 2004. Downloaded from

Table 3. The Incidence of Occult Diabetes and diagnosed diabetic patients already have clinical Impaired Glucose Tolerance in Patients Referred to and electrophysiologic evidence of neuropathy,22,23 Neuromuscular Clinics with a Diagnosis of ؅Idiopathic؅ suggesting that lesser degrees of Sensory Neuropathy may be associated with the onset of nerve fiber

Incidence of Impaired degeneration. Study Glucose Metabolism Although there is increasing evidence that im- paired glucose tolerance may cause a painful sen- Palade et al, 200010 10/75 (13%) sory neuropathy indistinguishable from mild dia- Singleton et al, 200111 48/107 (45%)* 13 DM and 35 IGT betic neuropathy, there have not yet been any Novella et al, 200112 24/48 (50%)† 11 DM and 13 IGT Sumner et al, 200313 41/73 (56%) 15 DM and 26 IGT prospective studies investigating the value of improved glycemic control in these patients. How- DM, diabetes mellitus; IGT, impaired glucose tolerance; GTT, ever, it is known that patients with impaired oral glucose tolerance test. glucose tolerance are at increased risk for the de- * Includes 13 of 33 patients with abnormal GTT and normal fastingglucose,and 10 of 21 patients with abnormal GTT and velopment of overt diabetes and that this risk can normal glycosylated hemoglobin level. be reduced by weight loss, exercise, and initiation † Of the 28 patients with painful neuropathy, 18 (65%) had of oral hypoglycemic agents.24,25 Furthermore, it abnormal glucose metabolism (8 of 28 DM and 10 of 28 IGT). also known that improved glycemic control can prevent and delay progression of diabetic neurop- along.9 There have been a number of recent series athy.3 suggesting than more than 50% of the patients referred to neuromuscular clinics with a diagnosis Conclusion of “idiopathic” painful sensory neuropathy have Based on the data available in the current medical abnormal glucose metabolism when they are eval- literature, we suggest that patients presenting with uated usingthe 2-hour glucosetolerance test (Ta- unexplained painful sensory neuropathy should be 10–13 ble 3). Of these, more than 50% have impaired evaluated for impaired glucose metabolism with a glucose tolerance, so the fasting glucose and glyco- 2-hour OGTT; if the results of the test are abnor- hemoglobin are frequently normal despite an ab- mal, they should be referred for lifestyle interven- 11 normal glucose tolerance test. These incidences tions and/or initiation of oral hypoglycemic agents of occult diabetes and impaired glucose tolerance in addition to management of neuropathy symp- are significantly higher than those reported for the toms. Improved glycemic control can prevent the http://www.jabfm.org/ general population, suggesting a causal relation- development of overt diabetes mellitus and may 11,14 ship. have some impact on progression of neuropathy in Diabetic polyneuropathy is typically a distal these cases. The association between “idiopathic” symmetric sensorimotor process that involves both painful sensory neuropathy and impaired glucose small- and large-caliber fibers. Clinically, sensory metabolism argues very strongly for prospective symptoms are very common, and in many cases, the studies in larger populations looking at the cost- on 27 September 2021 by guest. Protected copyright. neuropathy is painful because of prominent small effectiveness of this approach. fiber involvement.15 Skin biopsy for epidermal nerve fiber quantification has been useful for eval- References uatingpainful and shows epi- 1. Dyck PJ, Kratz KM, Karnes JL, et al. The prevalence dermal denervation in these cases.16–19 A number by staged severity of various types of diabetic neu- of recent studies have also shown epidermal dener- ropathy, retinopathy, and nephropathy in a popula- vation in patients with impaired glucose tolerance tion-based cohort: the Rochester Diabetic Neurop- and painful sensory neuropathy.20,21 These patients athy Study [published erratum appears in Neurology have lesser degrees of epidermal denervation and 1993;43:2345]. Neurology 1993;43:817–24. are more likely to have normal nerve conduction 2. Cohen JA, Jeffers BW, Faldut D, Marcoux M, studies than those with diabetic neuropathy,11,13,20 Schrier RW. Risks for sensorimotor peripheral neu- ropathy and autonomic neuropathy in non-- consistent with the hypothesis that neuropathy dependent diabetes mellitus (NIDDM). Muscle from impaired glucose tolerance may be a precur- Nerve 1998;21:72–80. sor to diabetic neuropathy with more selective 3. Anonymous. The effect of intensive treatment of small fiber involvement. Furthermore, many newly diabetes on the development and progression of

Neuropathy and Diabetes 129 J Am Board Fam Pract: first published as 10.3122/jabfm.17.2.127 on 13 April 2004. Downloaded from

long-term complications in insulin-dependent diabe- Health and Nutrition Survey, 1988–1994. Diabetes tes mellitus. The Diabetes Control and Complica- Care 1998;21:518–24. tions Trial Research Group. N Engl J Med 1993; 15. Said G, Slama G, Selva J. Progressive centripetal 329:977–86. degeneration of axons in small fibre diabetic neurop- 4. Anonymous. Report of the Expert Committee on the athy. Brain 1993;106:791–807. Diagnosis and Classification of Diabetes Mellitus. 16. Brown MJ, Martin JR, Asbury AK. Painful diabetic Diabetes Care 1997;20:1183–97. neuropathy: a morphometric study. Arch Neurol 5. Franklin GM, Kahn LB, Baxter J, Marshall JA, 1976;33:164–171. Hamman RF. Sensory neuropathy in non-insulin- 17. Levy DM, Karanth SS, Spingall DR, Polak JM. De- dependent diabetes mellitus. The San Luis Valley pletion of cutaneous nerves and neuropeptides in diabetes study. Am J Epidemiol 1990;131:633–43. diabetes mellitus: an immunocytochemical study. Diabetologica 1989;32:427–33. 6. Dyck PJ, Oviatt KF, Lambert EH. Intensive evalu- 18. Kennedy WR, Wendelschafer-Crabb G, Johnson T. ation of referred unclassified neuropathies yields im- Quantification of epidermal nerves in diabetic neu- proved diagnosis. Ann Neurol 1981;10:222–6. ropathy. Neurology 1996;47:1042–8. 7. Holland NR, Crawford TO, Hauer P, Cornblath 19. Holland NR, Stocks A, Hauer P, Cornblath DR, DR, Griffin JW, McArthur JC. Small-fiber sensory Griffin JW, McArthur JC. Intraepidermal nerve fi- neuropathies: clinical course and neuropathology of ber density in patients with painful sensory neurop- idiopathic cases. Ann Neurol 1998;44:47–59. athy. Neurology 1997;48:708–11. 8. Periquet MI, Novak V, Collins MP, et al. Painful 20. Smith AG, Ramachandran P, Tripp S, Singleton JR. sensory neuropathy: prospective evaluation using Epidermal nerve innervation in impaired glucose skin biopsy. Neurology 1999;53:1641–7. tolerance and diabetes-associated neuropathy. Neu- 9. Holland NR. Idiopathic painful sensory neuropathy. rology 2001;57:1701–4. J Clin Neuromusc Dis 2001;2:211–20. 21. Sheth S, Sumner C, Hauer P, et al. The spectrum of 10. Palade A, Small GA. Diabetes mellitus is underdiag- small fiber nerve involvement in IGT and diabetes nosed in patients with sensory peripheral neuropathy [abstract]. Neurology 2002;58(Suppl 3):A165. [abstract]. Muscle Nerve 2000;23:1632–3. 22. Ratzmann KP, Raschke M, Gander I, Schimke E. Prevalence of peripheral and autonomic neuropathy 11. Singleton JR, Smith AG, Bromberg MB. Painful in newly diagnosed type II (noninsulin-dependent) sensory polyneuropathy associated with impaired diabetes. J Diabet Complications 1991;5:1–5. glucose tolerance. Muscle Nerve 2001;24:1225–8. 23. Lehtinen JM, Uusitupa M, Siitonen O, Pyorala K. 12. Novella S, Inzucchi SE, Goldstein JM. The fre- Prevalence of neuropathy in newly diagnosed quency of undiagnosed diabetes and impaired glu- NIDDM and nondiabetic control subjects. Diabetes cose tolerance in patients with idiopathic sensory 1989;38:1307–13. neuropathy. Muscle Nerve 2001;24:1229–31. http://www.jabfm.org/ 24. Tuomilehto J, Lindstrom J, Eriksson JG et al. Pre- 13. Sumner CJ, Sheth S, Griffin JW, Cornblath DR, vention of mellitus by changes in Polydefkis M. The spectrum of neuropathy in dia- lifestyle amongsubjects with impaired glucosetol- betes and impaired glucose tolerance. Neurology erance. N Engl J Med 2001;344:1343–50. 2003;1460:108–11. 25. Knowler WC, Barrett-Connor E, Fowler SE, et al. 14. Harris MI, Flegal KM, Cowie CC et al. Prevalence Reduction in the incidence of type 2 diabetes with of diabetes, impaired fastingglucose,and impaired lifestyle intervention or metformin. N Engl J Med

glucose tolerance in US adults. The Third National 2002;346:393–403. on 27 September 2021 by guest. Protected copyright.

130 JABFP March–April 2004 Vol. 17 No. 2