The effect of temperature on symptoms of neurological disorders is complicated and has not been fully explored, neither in patients, nor in physiological research. A well-known effect is the worsening of symptoms in multiple sclerosis after a hot bath [²⁰, ²⁵]. In peripheral neuropathies, similar effects were described. In a patient with chronic inflammatory demyelinating polyneuropathy (CIDP), symptoms considerably increased during fever and in seven patients with various demyelinating neuropathies, electrophysiological signs of conduction block increased after warming and decreased after cooling [⁴, ⁸, ²⁰]. This so-called heat paresis may occur in demyelinating disorders and is caused by an unfavorable combination of factors that block saltatory conduction between an active node of Ranvier and the node that is next to be activated. First, paranodal demyelination results in current leakage at the node-to-be-activated, so that less current is available to depolarize this node [¹¹]. Secondly, raising temperature will decrease sodium-channel open time at the active node [⁶, ²¹]. This yields a further decrease in the current available to depolarize the node-to-be-activated. If there is just sufficient current available to sustain action potential propagation at a demyelinated internode, an additional decrease in current due to rising temperature may, therefore, induce conduction block as was shown in single demyelinated internodes [¹⁹].

In the authors’ experience, however, patients with peripheral nervous system disorders often state that weakness during cold increases. This cold paresis is not consistent with the above-described mechanism because that would lead to less weakness in cold. Whether or not cold paresis is really common and if it is related to specific disorders has not been investigated. Furthermore, the mechanisms of cold paresis are not clear. Cold paresis was only reported in single cases of multifocal motor neuropathy (MMN) and in patients with Hirayama disease [¹¹, ²³]. To explain cold paresis in MMN, Kaji hypothesized the existence of inflammatory nerve lesions with permanently depolarized axons that only just conduct at normal temperature [¹¹]. At lower temperatures, Na/K-pump activity decreases due to the thermal reduction of its ATP-ase activity; this may cause additional depolarization, yielding depolarizing conduction block and weakness [⁹–¹¹]. Recent findings from our group supported the hypothesis of Kaji by showing that cooling indeed induced depolarization of human motor axons, and that this depolarization was best explained by decreased pump-activity [⁷]. Since, in the future, such axons may possibly be saved from degeneration by neuroprotective treatment, cold paresis may become an important symptom [¹³, ²², ³⁰].

The present study was not conducted to investigate the mechanism of cold paresis but to assess if symptoms of cold paresis are common in MMN and if they are specific for MMN. If cold paresis were a common symptom, it may instigate further research into its mechanisms. If cold paresis were specific for MMN, it may form a clue for the pathogenesis of MMN, which is currently unknown. We investigated the frequency of symptoms of cold paresis and heat paresis in a large group of patients with MMN. Control groups included CIDP, sporadic progressive spinal muscular atrophy (PSMA), and chronic idiopathic axonal polyneuropathy (CIAP).

Exposure to cooling or warming did not differ significantly between MMN, CIDP, and PSMA, but CIAP patients were significantly less often exposed to cooling than MMN and PSMA patients (Tables 2, 3). Exposure to warming was more common than exposure to cooling and cold weather was the most common exposure to cooling. Despite this, patients reported cold paresis more often than heat paresis (Table 4).

Cold paresis in arms or legs was reported more frequently in MMN than in CIDP, PSMA, and CIAP (Table 4). Multivariate analysis showed that these differences were independent of sex, age, and disease duration, except for CIAP (Table 5). Patients with MMN had a fivefold higher risk of symptoms of cold paresis than patients with CIDP, and a sixfold higher risk than patients with PSMA. Multivariate analysis for symptoms of cold paresis in arms and legs separately, showed that these differences were still significant for the arms but not for the legs (Table 5).

Heat paresis in arms or legs was not significantly more or less frequently reported in MMN than in CIDP, PSMA and CIAP (Table 4). Neither univariate nor multivariate analysis showed any significant differences between MMN and the other patient groups (data not shown).

The present study demonstrated that cold paresis was experienced by a substantial proportion of patients with MMN, CIDP, PSMA, and CIAP. Cold paresis was more frequently reported than heat paresis. Cold paresis was more frequently reported in MMN than in CIDP, PSMA, or CIAP. Multivariate analysis indicated that MMN patients had a 4- to 6-fold greater risk of experiencing cold paresis than CIDP or PSMA patients. Because MMN predominantly affects arm nerves [²⁷], we analyzed symptoms of cold paresis in arms and legs separately; this showed that these differences were still significant for the arms but not for the legs.

In MMN, electrophysiological, pathological, and immunological studies suggest that several mechanisms may be involved in the unique process that results in asymmetric loss of strength and muscular atrophy. Motor nerve conduction studies may show conduction block and slowing consistent with demyelination in the same segment; however, conduction velocity may also be normal in segments with motor conduction block [²⁷, ²⁸]. Some pathological studies of nerve branches containing motor and sensory axons showed demyelination but others only showed loss of axons [², ⁵, ¹², ²⁴]. Needle electromyography studies showed extensive neurogenic abnormalities in non-atrophic as well as in atrophic muscles [²⁶]. Excitability studies showed focal depolarization or hyperpolarization of motor axons of which the significance is not clear [¹⁴, ¹⁷]. In many MMN patients, antibodies against epitopes presented on the axolemma are found, and a recent study showed that these anti-GM1 antibodies are associated with loss of motor axons [³]. Thus, whereas demyelination clearly occurs in MMN, mechanisms leading to primary axonal degeneration may possibly occur as well. It is not clear if the conduction block is due to unfavorable effects of paranodal demyelination, blocking of impulse generation at the node, or both. Because MMN may lead to demyelination as well as loss of motor axons, we have chosen CIDP, PSMA, and CIAP as control groups in order to make a comparison to peripheral nervous system disorders that are also characterized by weakness: CIDP is characterized mainly by demyelination of motor and sensory axons, PSMA by loss of motor axons, and CIAP by length-dependent loss of motor and sensory axons.

Cold paresis was reported by two-thirds of patients with distal upper limb muscular atrophy (Hirayama disease), a disorder affecting peripheral motor neurons in the anterior horn of the cervical cord, which results in denervation of hand muscles [²³]. Nerve conduction studies in 11 patients with Hirayama disease, and one patient with hypothenar atrophy due to ulnar neuropathy, showed that cold induced excessive conduction delay and waning of the compound muscle action potential during 20-Hz repetitive stimulation [¹⁵, ¹⁸]. These findings were attributed to an increased sensitivity of reinnervated muscle fibers to develop depolarizing conduction block in cold.

Cold paresis in MMN was suggested to be related to the existence of inflammatory nerve lesions where axons are depolarized but just able to conduct impulses at ambient temperature [¹¹]. At these lesions, thermal reduction of Na/K-pump activity due to cooling may induce further depolarization to a point where impulse conduction fails. Some findings may support this hypothesis. First, animal models of inflammatory spinal root lesions indicated that inflammation may induce nitric oxide-mediated mitochondrial dysfunction, energy depletion of the ATP-ase dependent Na/K-pump, and axonal depolarization [¹³]. Secondly, excitability studies in MMN indirectly suggested that axons at the site of block may possibly be depolarized [¹⁴, ¹⁷]. Thirdly, cooling of normal human motor axons results in their depolarization, probably due to decreased Na/K-pump activity [⁷]. Some of the steps in this hypothesis were corroborated by experiments, but it must be emphasized, however, that its entire sequence has not been proven. Moreover, some experiments were carried out in central nervous system neurons so that the results possibly cannot be extrapolated to peripheral motor neurons.

Alternatively, cold paresis in MMN may be caused by development of conduction block in reinnervated muscle fibers due to the lower temperature [¹⁵, ¹⁸]. The possibility that this mechanism, which may be similar to that described in Hirayama disease, occurs in MMN is supported by several studies. Extensive needle electromyography studies in MMN patients showed signs of reinnervation, indicating collateral sprouting, in non-atrophic as well as in atrophic muscles [²⁶]. Some pathological studies of motor nerves in MMN have shown prominent loss of axons and little demyelination; the loss of axons may lead to collateral sprouting and reinnervated muscle fibers [²⁴].

In conclusion, symptoms of cold paresis occur in several peripheral nervous system disorders, but are considerably more common in MMN. Because the present study only assessed subjective symptoms of weakness, it is worthwhile to investigate if cold also induces an objective increase of weakness in MMN.

J. T. H. van Asseldonk and D. C. G. Straver were supported by grants from the Prinses Beatrix Fonds.

Conflict of interest The authors declare that they have no conflict of interest.

Open Access This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.