Sensitivity of a Dutch Wrist Circumference–Dependent Upper Limit of Normal of the Median Nerve in an Italian Population of Patients with Carpal Tunnel Syndrome: An Observational Study

1. Introduction

Carpal tunnel syndrome (CTS) is the most frequently encountered entrapment neuropathy [1]. It is caused by compression of the median nerve (MN) as it passes through the carpal tunnel. The clinical diagnosis of CTS lacks a widely accepted gold standard however, with diagnoses often confirmed using electrodiagnostic testing (EDX) or ultrasonography (US) in clinical practice and research settings [1,2].

In the Netherlands, US is the recommended as the first-choice diagnostic test because it is easily accessible and painless [3]. Various US parameters have been suggested in the literature for confirming CTS, and an increase of the cross-sectional area (CSA) of the MN at wrist level is the most commonly used [4,5]; however, there is no consensus in the literature about the upper limit of normal (ULN) of this CSA. A broad range of values for the ULN have been proposed, for example 9–15 mm2 in a post-hoc analysis [6].

A strong correlation between the dimensions of the MN, particularly the CSA, and the wrist circumference (WC) was seen in an earlier study. Other anthropometric features were not independently correlated to CSA [7]. Another study showed a positive correlation between WC and height and weight [8]. The average height of the Dutch population (self-reported, 20 years or older) is the tallest in the world, at 169.3 cm for women and 182.9 cm for men born in 2001 [9]. In the general healthy Italian population, the average height is shorter: 162.6 cm and 176.5 cm for women and men, respectively [10]. If height indeed influences WC and the CSA of the MN at the wrist level, this could lead to differences in CTS diagnosis between the Italian and Dutch population.

In this study, we investigated the sensitivity of US in patients with CTS using a Dutch WC-dependent (WCD) equation for the ULN of the MN CSA at wrist level in samples of the Italian and Dutch population. We hypothesise that the WC in the shorter Italian population would be lower and, therefore, would lead to a lower calculated ULN for the MN CSA at wrist level. As a consequence, we expected that our equation would be similarly sensitive for patients with CTS from Italy and the Netherlands.

2. Materials and Methods

In this observational study, we included patients with CTS from the Siena University Hospital (Italy) and retrospectively enrolled patients with CTS from the Netherlands (Canisius-Wilhelmina Hospital, Zuyderland Hospital and Radboud university medical centre) [11]. We obtained written informed consent from all patients and approval was waived by the local Medical Ethics Committee.

Patients with CTS aged 18 years and older were enrolled. They were eligible for inclusion if they complained of pain and/or paraesthesia in the region innervated by the MN. Furthermore, two of the following criteria had to be present: (1) paraesthesia relieved by shaking the hand (positive Flick sign), (2) aggravation of paraesthesia by certain activities (e.g. driving a car, biking, using a phone, holding a book) and (3) nocturnal paraesthesia. We excluded patients with a history or clinical signs of polyneuropathy, hereditary neuropathy with liability to pressure palsies, previous trauma or surgery to the wrist, history of rheumatoid arthritis, arthrosis of the wrist, diabetes mellitus, thyroid disease, alcoholism and bifid MNs. The weight and height of all patients were measured. The WC were measured at the level of the distal wrist crease using a plastic measuring tape.

All participants underwent US and EDX studies according to previously described protocols [12,13]. In Italy, patients were enrolled in the study if the symptoms compatible with the clinical diagnosis of CTS were confirmed by EDX evidence of delay of the distal conduction velocity of the MN. Data was collected between February 2022 and December 2022. In the Netherlands, enrolment was based on a clinical diagnosis of CTS only, so patients with normal and abnormal nerve conduction studies were included. In the Netherlands, US was performed by experienced neurophysiology technicians, while in Siena, Italy, it was conducted by an experienced rheumatologist sonographer. The CSA of the MN at the carpal tunnel was outlined using the inner margin of the hyperechoic rim. The CSA was calculated by the area measurement software of the US system. In the Netherlands, a Hitachi Aloka Arietta 850 ultrasound system was used (5–17 MHz linear array transducer), while in Siena, a MyLab X8 eXP Esaote ultrasound system (4–15 MHz and 8–24 MHz linear probes) was used.

We compared the number of wrists with CTS confirmation using fixed values for the ULN of the CSA and a WCD ULN for the CSA. The fixed ULN was set at > 9 mm2 (ULN9) in the Italian laboratory [14] and > 11 mm2 in the Dutch laboratories (ULN11) [15]. The WCD ULN for the CSA was calculated as y=0.88*x-4 (y = ULN of the CSA in mm2 and x = WC in cm), as described in a previous study [7]. These results were compared to the EDX studies.

A statistical analysis was performed using SPSS Statistics (version 26.0). We determined the distribution of our data via Q–Q plots, histograms and the Kolmogorov–Smirnov test. For categorical variables, we used the chi-square test, while continuous variables with a non-normal distribution were analysed using the Mann–Whitney test. The threshold for statistical significance was set at p < 0.05.

3. Results

Table 1 displays the biometric characteristics of the patients. In total, 260 wrists of 230 patients with CTS were included. Patients included in Italy were significantly older, while patients in the Netherlands were taller and heavier. There were no differences in gender, median WC or median CSA of the MN between the groups. In Italy, the median duration of symptoms was significantly longer.

Table 2 presents the findings derived from the US and EDX studies. The relative number of abnormal wrists as determined by US (three previously defined ULN values for the MN CSA) was similar between the Italian and Dutch participants, including in the WCD results. Using the smallest ULN, the fixed ULN used in Italy, most US results were considered abnormal. A comparison of the number of abnormal US in both populations based on the different ULN values are presented in Figure 1.

Table 3 shows the US results of a small selection of the Italian participants (n = 12) with WCs between 14.6 cm (–2 standard deviation) and 15.8 cm (–1 standard deviation).

4. Discussion

This study found that, in clinically defined patients with CTS, a Dutch WCD ULN for the MN CSA at the carpal tunnel could identify CTS with similar accuracies in both the Italian and Dutch population. The sensitivity of this method was poor in both populations, however. The application of the ULN developed by the Italian laboratory showed the highest sensitivity because of the relatively low cut-off value of 9 mm2.

It was hypothesised that the WCD formula would have a similar sensitivity in the Italian and Dutch participants despite their expected differences in WC and MN CSA. While the sensitivity was comparable, we found no differences in these anthropometric features. An earlier study showed a positive correlation between WC, height and weight [8]. This would mean that taller people have a larger WC and consequently a larger MN CSA. While the Dutch population is the tallest in the world [16], in this study the difference in height between the Italian and Dutch population was small. This might be partly explained by the fact that our data come from the southern provinces of the Netherlands, where the population is shorter than in the north [9]. On the other hand, the patients enrolled in Italy came from Tuscany, where people are on average taller than those in several other Italian regions [17]. The influence of past gene flows from southern Europe in the southern provinces of the Netherlands might also explain the similar WC and MN CSA measurements in the Dutch and Italian populations in this study. Migrational boundaries of this gene flow, especially the major Dutch rivers, could explain the differences between provinces in the Netherlands [18]. The Dutch data from this study was obtained from people living south of the major Dutch rivers.

While many studies have gathered data for normative value of the MN CSA [4], few studies primarily looked at population differences. In healthy participants, one study showed that, after correction for age, height and weight, the MN CSA difference between Indian and Dutch participants was still significant [19]. A systematic review concerning normative reference values for the MN CSA identified several differences between ethnicities, but it was noted that the majority of included studies did not report sub-population ethnicity [20]. One study compared data for patients with CTS collected by an American (USA) laboratory and an Italian laboratory, and found strong similarities between the MN CSA at forearm level. While the wrist-to-forearm ratio (WFR) was higher in the USA (because of a significantly larger mean MN CSA at wrist level), both ratios were considered elevated compared with the controls [21]. Here, we found no difference in the MN CSA at wrist level between the Dutch and Italian participants.

One unexpected finding was the low sensitivity of the WCD equation in both the Dutch and Italian population. We do not have a good explanation for this. An earlier study showed a higher sensitivity for this equation in a different Dutch population [7]. The age of our study population was higher than in this previous study and could play a part in the lower sensitivity. There are several approaches to improve diagnostic accuracy. A disease-specific cut-off value can be used and, as is common in nerve conduction studies as well, patients can serve as their own controls [22]. In patients with CTS, the WFR is commonly used and, more recently, the nerve/tendon ratio showed promise as an anthropometric-independent ULN [14,23].

We also looked at a small selection of our Italian population (n = 12) with particularly small WCs (between 14.6 cm (–2 standard deviation) and 15.8 cm (–1 standard deviation)); these WCs were less common in the Dutch population. The Dutch WCD ULN considered the US readings of eight of these 12 patients (66.7%) to be abnormal, while the ULN9 calculation considered nine readings to be abnormal (75.0%). Conversely, using ULN11 in this subpopulation, only 25% of the scanned MNs would be classified as abnormal. As reported before, we believe that a WCD cut-off value can add value, especially in patients with CTS who have small wrists [11].

There are several limitations to this study. First, the median duration of symptoms was significantly longer in the Italian population than in the Dutch patients. The literature has conflicting results regarding the effect of symptom duration and MN CSA at the wrist level [24,25,26,27]. As mentioned before, patients in Italy were significantly older, and it is known that the sensitivity of the MN CSA as a cut-off value is lower in older patients with CTS than in younger patients [28,29]. A recent expert panel advised that EDX studies should be performed in individuals aged over 70 [30]. Furthermore, all patients in Italy had abnormal EDX results, while some of the Dutch patients were included on the basis of a clinical CTS diagnosis despite normal EDX findings. We cannot exclude the possibility of misdiagnosis in some patients in the Netherlands, which could influence the diagnostic accuracy in this study.

5. Conclusion

In summary, this study demonstrates that a Dutch-derived formula for determining a WCD cut-off value for the MN CSA at the carpal tunnel, among individuals diagnosed with carpal tunnel syndrome, yielded comparable outcomes among Italian and Dutch cohorts. This formula may offer heightened utility among individuals with diminutive wrist dimensions. We found no differences in WC or MN CSA at wrist level between these two groups. Furthermore, this observational study shows the importance of establishing local normative values for the ULN of the MN area at the wrist used to confirm the clinical diagnosis of CTS.