To our knowledge, this is the first systematic study on familial risks in nerve, nerve root and plexus disorders. The use of hospitalisations as an outcome may have advantages and disadvantages.14,17
Many of the diagnoses will not require hospitalisation and the rates would be lower than those seen in the general population. On the other hand, diagnostic accuracy was probably good overall because hospitalisation normally involved a diagnosis made by several physicians, including neurologists. An advantage of a nationwide, fully register‐based study is that selection biases are minimised and both the probands and the patients are medically diagnosed. Almost any alternative approach would require interviewing patients about similar diseases in their relatives, which has been shown to yield highly inaccurate data even for relatively well defined and life‐threatening diseases, such as cancer.18
A potential problem could be selective hospitalisation; when one sibling is hospitalised, other siblings may preferentially also seek care. Such a selection bias would be likely for conditions that do not invariably lead to hospitalisation, including many of the studied diseases in mild forms. To test for the likelihood of such bias, we calculated risks for spouses, who would be expected to show a higher level of environmental sharing than siblings for any adult‐onset diseases. We have previously estimated the potential for such a selection by comparing hospitalisation rates for migraine between spouses, and found no evidence supporting bias.14
Moreover, as this was a countrywide study, the diagnostic practices of single hospitals were unlikely to bias the results. However, for the studied diseases, correlations between spouses were observed, as discussed later. As the national Hospital Discharge Register has only been in operation since 1987, this study covered a time period of no longer than 15 years, partially contributing to the small numbers of cases for specific diagnostic groups. Overall, familial cases were rare, limiting the statistical power of the study. All family studies are prone to data on false paternity, which would bias results towards null.
The incidence of trigeminal neuralgia is given as 4.3 per 100
000 person‐years, with a female excess of 3:2.1
In our study, the hospitalisation rate in the 0–69‐year‐old population was 1.1 per 100
000, with a moderate female excess. Our lower incidence figures could reflect selection for hospitalisation or the truncated population, lacking elderly groups who have a high incidence (fig 1). The reported incidence for Bell's palsy has varied between studies from about 10 to 40 per 100
000 person‐years, equal in both sexes.3,17,19
Our study found a rate of 4 per 100
000 person‐years, equal for both sexes; however, again, part of this lower incidence would be explained by truncation of the population and lacking of the high‐risk population at ages
70 years (fig 1). In the UK study, only 19% of the patients with Bell's palsy were referred to a hospital during the year after the initial onset of the disease; the hospitalisation rate in this UK study was almost exactly as ours, 4 per 100
The incidence of carpal tunnel syndrome was found to be 88 per 100
000 for men and 193 per 100
000 for women in the UK General Practice Research Database.20
The crude rate in our study was much lower, 19 per 100
000 for men and 46 per 100
000 for women, but interestingly, the female excess was almost exactly that of the UK study. These data are consistent with the notion that our study data are lower than population morbidity, but hospitalisation rates may not bias sex‐specific rates. Thus, because of the various truncations, the hospitalisation rates should not be regarded as population incidence rates for all or any subtypes of these diseases. Hospitalisations are likely to be more preferential for severe symptoms and diseases, for which diagnostics require specialist consultation. Another explanation could be worse prognosis and preferential hospital referral in the familial cases. However, the truncations as such should not bias familial risks, the estimation of which was the novel aspect of this study.
What is already known
- Nerve, nerve root and plexus disorders are common disorders, including Bell's palsy and carpal tunnel syndrome, with many types of known environmental causes.
- Little is known about their possible familial clustering, which might indicate shared environmental causes or inheritable aetiology.
What this paper adds
- Familial clustering between siblings was observed for all nerve, nerve root and plexus disorders and for specific subtypes.
- Some concordance was also observed for spouses, probably indicating environmental sharing.
- However, as the risks between siblings were higher than those between spouses, an inheritable contribution is probable.
The overall familial risk was 2.59, equal for men and women. Facial nerve disorders and mononeuritis of the limbs showed familial risk when the co‐siblings presented with any nerve, nerve root and plexus disorder (table 2). Limb mononeuritis did not include radiculopathies because these were classified under nerve root and plexus disorders, which showed no evidence on familial clustering. For men, even disorders of other cranial roots were significant (SIR 10.41). When specific subtypes were compared in both siblings, disorders of the trigeminal nerve (6.34), Bell's palsy (3.94) and carpal tunnel syndrome (9.22) showed increased risk. Only mononeuritis of the upper limbs was associated with discordant subtypes, nerve root and plexus disorders (3.32) and mononeuritis of the lower limb (3.07); this observation may imply shared disease mechanisms or aetiology. Herpes simplex infection is considered the major cause of trigeminal neuralgia and Bell's palsy, and it is thus conceivable that siblings share infection and the viral source of the disease. The correlation between spouses for trigeminal neuralgia supported this possibility, at least as a partial explanation. For Bell's palsy, no large correlation between spouses was observed, and the familial risks may reflect inheritable susceptibility to viral infection or other genetic causes. For carpal tunnel syndrome, the familial risk (9.22) exceeded correlation between spouses (4.03), and environmental factors may not fully explain the sibling risk. Known familial clusters often involve bilateral disease, and it has been suggested that carpal tunnel diameters may be inheritable and thus contribute to disease proneness.13
The results of this study should encourage an eetiological search for the causes of the familial aggregation. Although our data cannot distinguish between the contributions of inheritability or shared environmental factors, or their interactions, to the observed familial aggregation, they clearly show that familial clustering is present. Molecular genetic techniques have been successful in dissecting many monogenic neurological disease,21
but nerve, nerve root and plexus disorders seem to be multifactorial, which may require a careful selection of the population with disease for study.