Abstract
Background:
An altered balance of gonadal hormones in males with gender identity disorders (GIDs) may increase multiple sclerosis (MS) risk both inherently and secondary to treatment in undergoing male-to-female conversion.
Objective:
We investigated any association between GIDs and MS through analysis of record-linked hospital statistics.
Method:
Analysis of English Hospital Episode Statistics, 1999–2012.
Results:
The adjusted rate ratio (RR) of MS following GIDs in males was 6.63 (95% confidence interval (95% CI) = 1.81–17.01, p = 0.0002). The RR of MS following GIDs in females was 1.44 (95% CI = 0.47–3.37, p = 0.58).
Conclusion:
We report a strong association between GIDs and MS in male-to-females, supporting a potential role for low testosterone and/or feminising hormones on MS risk in males.
Introduction
Gender is an important factor influencing multiple sclerosis (MS) risk. Females have a three-fold higher prevalence of MS, earlier disease onset and slower disability progression.1,2 Gonadal hormones may be important in mediating gender differences in MS risk, but their influence remains largely undetermined. Transgenderism or gender identity disorders (GIDs) have been described as ‘an extreme form of gender dysphoria characterized by a strong and persistent identification with the opposite sex’. 3 The pathophysiology of GIDs is likely complex, involving hormonal derangements, genetic predisposition and sociocultural factors. Testosterone is thought to directly influence male foetal brain development, sexual orientation and gender identity, and individuals with male-to-female (MtF) GIDs (individuals born male but identifying as female) have been associated with genetic variants resulting in reduced androgen sensitivity. 4 Treatment of GIDs in accordance with the World Professional Association for Transgender Health recommends hormonal therapy, psychotherapy, changes in gender expression and potential surgical sex reassignment. The most studied treatment regimen for MtF individuals involves the use of feminising hormones (e.g. oestrogen) and anti-androgens (e.g. gonadotropin-releasing hormone analogues) which reduce testosterone secretion or neutralise testosterone activity. We hypothesised that an altered balance of gonadal hormones in males with GIDs will increase MS risk both inherently and secondary to treatment in undergoing MtF conversion.
Methods
Population and data
A retrospective cohort study was conducted through analysis of linked English national Hospital Episode Statistics (HES) and mortality data, January 1999–March 2012. The Oxford record-linkage group undertook record-linkage to construct a time-sequenced record of successive episodes of care (and death, if it occurred) for each person.
Cohorts of males and females with GIDs were constructed by identifying the first episode of day-case care or hospital admission in which a GID or sexual transformation procedure was coded in any diagnostic position (International Statistical Classification of Diseases and Related Health Problems, Revision 10 (ICD-10) codes F64.0, transsexualism; F64.1, dual-role transvestism; F64.2, GID of childhood; F64.8, other GIDs; F64.9, GIDs unspecified; and/or Office of Population Censuses and Surveys Classification of Interventions and Procedures codes X15.1, combined operations for transformation from male to female; X15.2, combined operations for transformation from female to male; X15.8, unspecified operations for sexual transformation).
A male and female reference cohort was constructed by identifying the first admission of individuals for various other, mainly minor conditions (Table 1 footnotes). We used standard epidemiological practice in selecting a wide range of ‘control’ conditions in the reference cohort (to avoid the possibility that any one condition might have an atypically high or low subsequent risk of MS). People were excluded if they had a MS admission either before or at the same time as the admission for GIDs or the reference condition. Building and use of the datasets was approved by the Central and South Bristol Research Ethics Committee (ref 04/Q2006/176).
GID: gender identity disorder.
We included all people eligible to be in the control cohort: for example, the risk of MS in the 195 men in the GID group aged 15–24 years was compared with the risk in the 577,785 (195 × 2963) men in the control cohort. We did so because in a stratified analysis of a dataset with all data already collected, there is no merit in discarding controls. All calculations of observed and expected numbers of ‘outcome’ conditions were done within each age stratum (in 5-year bands, rather than those showed for simplicity in the table) and within the other strata described in the ‘Methods’ section. The age-specific values were then summed to give all-ages summary statistics, as given in the text.
Conditions used in control cohort: squint, otitis externa/media, varicose veins, haemorrhoids, upper respiratory tract infections, deflected nasal septum, nasal polyp, teeth disorders, inguinal hernia, in-growing toenail and other diseases of nail, sebaceous cyst, internal derangement of knee, bunion, selected limb fractures, dislocations, sprains and strains, superficial injury and contusion, tonsillectomy, adenoidectomy, appendicectomy, hip replacement and knee replacement.
Statistical methods
Date of entry into each cohort was the date of the first admission for GID, or reference condition, and date of exit was the date of the first record of MS, death or the end of data collection, whichever was first. We calculated rates for MS, stratified and then standardised by age (in 5-year groups), calendar year of first recorded admission, region of residence and quintile of patients’ Index of Deprivation score. Methods used in the calculation of observed and expected numbers have been published in detail elsewhere. 5 Rate ratios (RRs) were calculated using the formula (OGID/EGID)/(OREF/EREF), where O and E are the observed and expected numbers of MS cases in the GID and reference cohorts, respectively. The confidence interval (CI) for the RR of MS and χ2 statistics for its significance were calculated as described elsewhere. 6
Results
There were 1157 males and 2390 females in the GID cohorts, and 4.6 million males and 3.4 million females in the respective reference cohorts. Their age distribution is given in Table 1. The adjusted RR of MS following GID in males was 6.63 (95% CI = 1.81–17.01, significant at p = 0.0002), based on 4 observed cases and 0.6 expected. The age of first known record of MS in MtF individuals varied but was within the 35- to 74-year age group. The adjusted RR of MS following GID in females was 1.44 (95% CI = 0.47–3.37, not significant, p = 0.58), based on 5 observed cases and 3.5 expected.
Discussion
We report a positive association between GIDs and subsequent MS in males. Our study design cannot give insights into mechanisms that might explain the association. However, there is evidence suggesting an association between gonadal hormones and MS, including animal models suggesting anti-inflammatory and/or neuroprotective actions of testosterone, 7 studies indicating a high prevalence of hypogonadism in male MS patients and improved cognitive function and slowed brain atrophy in a small pilot trial of testosterone.8,9 Furthermore, using a similar methodology and drawing upon the same population, we have previously reported a strong association between testicular hypofunction and MS risk. 10
Strengths of this study include the person-based cohort design and the use of a very large dataset to study an extremely challenging logistical undertaking: even drawing upon a 50 million population, numbers of cases were small. We have considered, as best possible using routine data, potential confounding factors such as socioeconomic deprivation. However, we cannot rule out unknown shared risk factors between GIDs and MS.
Limitations also include the use of prevalent cases of GIDs based on the first recorded day-case care or inpatient admission, rather than follow-up from the point of initial diagnosis, and the fact that there is a general increased healthcare usage by individuals with MS. The nature of GIDs, which are thought to onset early in life with a long period prior to medical attention being sought (if ever), in addition to the coding requirement that the GIDs always preceded MS, suggests that any true direction of effect is almost certainly one in which GIDs precede MS. However, we do not have detailed clinical information available to determine the temporal relationship between GIDs and any prior undiagnosed presentations of MS symptoms, or blood samples to elucidate the relationship between sex hormone levels and MS more directly. The people with GIDs would be a complex non-homogenous group of individuals. Available information about them (other than basic demographic data) is lacking. This lack of detail, and the small numbers of cases, means that we have had very little scope to take account of potential confounders (other than the specified demographic adjustments). Moreover, we assume that the majority of people with MS will have been hospitalised at least once, but cannot confirm this. The fact we used hospital data means that the GID cohort probably reflects the severe end of the condition’s spectrum.
In conclusion, our findings suggest that low testosterone levels and/or feminising gonadal hormones might influence MS risk in some men and highlight a need for further work to explore any potential role for gonadal hormones in management and/or prevention strategies. The demographic profile of MS is one of its most intriguing features and one which causal hypotheses must take into account.
Footnotes
Acknowledgements
Hospital Episode Statistics (HES) data were supplied by the Health and Social Care Information Centre, and mortality data were supplied by the Office for National Statistics.
Author contributions
J.P. proposed the study and wrote the first draft of the manuscript. C.J.W. undertook the analyses. M.J.G. is the guarantor of, and designed, the study. All authors contributed to the interpretation of the data and revision of the manuscript for important intellectual content.
Conflict of interest
All authors declare no conflicts of interest. Dr Pakpoor reports no disclosures. Ms Wotton reports no disclosures. Professor Goldacre reports no disclosures. Dr Schmierer has received speaking honoraria from and/or served on advisory boards for Novartis, Biogen, Teva, Merck Serono and Merck Inc. and has received support to attend the AAN 2014 from Genzyme. Dr Schmierer is a Principal Investigator (PI) on studies sponsored by Novartis, Roche and Teva and has received research support from Novartis. Professor Giovannoni serves on scientific advisory boards for Merck Serono and Biogen Idec and Vertex Pharmaceuticals; served on the editorial board of Multiple Sclerosis; has received speaker honoraria from Bayer Schering Pharma, Merck Serono, Biogen Idec, Pfizer Inc, Teva Pharmaceutical Industries Ltd, Vertex Pharmaceuticals, Genzyme Corporation, Ironwood and Novartis; has served as a consultant for Bayer Schering Pharma, Biogen Idec, GlaxoSmithKline, Merck Serono, Protein Discovery Laboratories, Teva Pharmaceutical Industries Ltd., UCB, Vertex Pharmaceuticals, GW Pharma, Novartis and FivePrime; serves on the speakers bureau for Merck Serono; and has received research support from Bayer Schering Pharma, Biogen Idec, Merck Serono, Novartis, UCB, Merz Pharmaceuticals, LLC, Teva Pharmaceutical Industries Ltd, GW Pharma and Ironwood.
Funding
The building of the linked datasets, and the development of the analytical software used to study disease associations, was funded by the English National Institute for Health Research. This study had no specific funding. The funder had no role in study design, data collection, data analysis, data interpretation, writing of the report or for the decision to submit for publication. The views expressed in the paper do not necessarily reflect those of the funding body. K.S. has been supported by a Higher Education Funding Council for England Clinical Senior Lectureship Award.