Does a Healthy Weight Body Mass Index at Onset of Idiopathic Intracranial Hypertension Change the Outcomes? A United Kingdom Prospective Cohort Study

Idiopathic intracranial hypertension (IIH) predominantly affects reproductive-aged females with obesity. However, the prevalence and impact of a healthy weight body mass index (BMI) at disease presentation is not known. This study aimed to evaluate the visual and headache outcomes stratified by the presenting BMI. This was a longitudinal prospective cohort study (IIH Life) based on a tertiary neuro-ophthalmology IIH service, University Hospitals Birmingham NHS Trust, United Kingdom, recruiting consecutive patients living with IIH between 2012 and 2021. Those with a presenting BMI were included. The outcome measures included visual outcomes of LogMAR visual acuity, Humphrey visual field perimetric mean deviation (PMD), optical coherence tomography (OCT) measurements


Introduction
Idiopathic intracranial hypertension (IIH) predominantly affects reproductive-aged females with obesity. 1,24][5][6][7][8] Atypical IIH is a term given to people who meet the diagnostic criteria for IIH but do not fit the typical phenotype, including males or females living without obesity. 3,9besity is defined by the World Health Organisation as a body mass index (BMI) ≥30 kg/m 2,10 which is reduced to 27.5 kg/m 2 in some ethnic groups due to increased cardiometabolic risk at a lower BMI. 11,12he adiposity associated with IIH is centripetal 13,14 and is correlated with lumbar puncture opening pressure. 13A two-fold increase in cardiovascular disease risk has been found in people living with IIH, in comparison to age-, sex-and BMI-matched controls. 2Other metabolic complications are associated with IIH, such as diabetes mellitus, gestational diabetes, pre-eclampsia, and arterial hypertension. 2,15,16n animal studies, both high-fat diet and IIH-associated inflammatory factor-exposed female rats showed increased CSF secretion. 17The adipocytes in IIH appear to be metabolically primed for lipogenesis 18 with an additional insulin and leptin resistant phenotype in excess of that driven by obesity alone. 18,19IIH is also associated with hypertestosteronism in females, independent of obesity, and testosterone increases CSF secretion through the Na + /K + -ATPase pump in the choroid plexus. 20There is also a debate in the literature about whether IIH is caused by problems with CSF secretion or drainage (including cerebral venous stenosis), or a combination of both, and multiple factors are likely to influence both these mechanisms. 21,226 In the IIH weight trial, those with a BMI ≥35 kg/m 2 disease remission, defined as normalisation of intracranial pressure (ICP), may require up to 24% weight loss. 27he presence of a healthy weight BMI on IIH and associated pathophysiology has previously been studied in a single retrospective study, which found better visual outcomes in those with a healthy weight BMI as compared to those with an increased BMI. 9 The lack of literature in this area may be because a healthy weight BMI at IIH diagnosis may be uncommon.The metabolic alterations related to obesity have been thought to be the main pathological drivers of IIH given the improvement associated with weight loss, what is not known is that, in the subset of IIH with a healthy weight BMI, there may be a different underlying pathogenesis.Therefore, in a rare condition, despite its rising prevalence, it would be difficult to adequately statistically power a trial to assess differing interventions in this subset.9,28,29 This study is the first to prospectively assess the prognosis in IIH in a cohort of people with a healthy weight BMI.The hypothesis is that IIH patients without obesity would have different visual and headache outcomes compared to typically obese IIH counterparts.

Methods
The methodology 5 of this prospective observational cohort study has been published and ethically approved by NHS National Research Ethics Committee (14/LO/1208), IIH LIFE study.All the patients included in this study attended a specialist IIH clinic (neuro-ophthalmology and neurology joint clinic) at a single tertiary neuroscience centre (University Hospitals Birmingham NHS Foundation Trust (UHB), United Kingdom).All the patients gave written informed consent to participate in this study.The consecutive data were collected between April 2012 and September 2021, for all visits for each patient.
Eligible patients included those who met either a definite or probable diagnosis for IIH based on the 2013 revised diagnostic criteria for IIH 30 and had a baseline BMI recorded.
The revised diagnostic criteria required for a definite diagnosis of IIH (adapted from Friedman et al. 31

include:
(A) the presence of papilloedema, (B) a normal neurological examination (except sixth nerve palsy), (C) normal brain parenchyma on neuroimaging (no hydrocephalus, mass, structural lesion, or meningeal enhancement), with exclusion of venous sinus thrombosis, (D) normal CSF constituents, and (E) an elevated lumbar puncture opening pressure ≥25 cm CSF. 30 A probable IIH diagnosis could be made in a typical patient, where criteria A to D are met.Patients with secondary causes for raised intracranial pressure or IIH without papilloedema were excluded. 3,31Secondary causes excluded were cerebral venous sinus thrombosis, mass lesions, infective causes, and medical contributors such as profound anaemia and regular/recent use of tetracyclines.Appropriate imaging was performed in all patients either after first visit or at their previous hospital.
This study stratified patients according to their BMI status at their first visit (baseline) in the specialist clinic.Patients were initially categorised as having (≥30 kg/m 2 ) or not having obesity (<30 kg/ m 2 ).Then, subsequently by World Health Organisation (WHO) BMI categories, 10 : healthy weight BMI (18.5-24.9kg/m 2 ), overweight (25.0-29.9kg/m 2 ), obesity class 1 (30.0-34.9kg/m 2 ) and obesity class 2, and above (≥35.0kg/m 2 ).As IIH has now been recognised as an obesity related disease by the National Institute of Health and Clinical Excellence in the UK, with regard to criteria for referral to a weight management speciality clinic obesity class 2 and above was used as the highest grouping. 12BMI groups were not adjusted by ethnicity.
Visual outcomes were: LogMAR visual acuity (measured using Logarithm of the Minimum Angle of Resolution), Humphrey visual field perimetric mean deviation (PMD) (24-2 SITA standard) and optical coherence tomography (OCT, Heidelberg Spectralis™) measures of average global peripapillary retinal nerve fibre layer thickness (RNFL), automatically measured total retinal thickness (TRT) (by the software on the RNFL circle and extracted as TRT) and macular ganglion cell layer (GCL) volume (1, 2.22, 3.45 mm ETDRS plot with the volume measurement as given by the software which would include all nine sectors, by the pre-set macular volume and/or posterior pole methods).To ensure accuracy of segmentation of the retinal layers in moderate-to-severe papilloedema, manual segmentation of RNFL and TRT was performed in peripapillary scans and in cross-sectional slices of optic disc scans for the basement membrane (BM) and inner limiting layer (ILM), where appropriate.During the course of the study, the scanning protocol for GCL changed; however, excellent correlation has previously been published between the macular volume and posterior pole measures (Pearson correlation 0.98). 5eadache outcomes were clinically assessed using the following indices: monthly headache days (days/month); monthly migraine-like headache days (days/month); headache severity (0-10 numerical rating scale, where 0 is no pain and 10 equates to the most severe); and headache disability using the Headache Impact Test-6 (HIT-6) score (score is between 36 and 78).
The methods for data collection and statistical analysis have previously been reported. 5The statistical analysis was performed using R v4.1.0. 32Continuous variables were reported as mean (standard deviation (SD)) and categorical variables as number (percentage).Lme4 33 used for regression modelling with continuous form of the dependent variables assumed and independent modelling.Statistical significance is indicated by 95% confidence intervals and is met when the intervals do not cross zero or overlap.Adjustments were made for multiple testing.LOESS (locally weighted scatterplot smoothing) graphs were created prior to regression analysis to ascertain the variables' relationships and any trends.
The study was guided by patient and public involvement of IIHUK, a national patient charity (Registered Charity in England and Wales no 1,143,522 & Scotland SCO43294) that supports carers and patients living with IIH.They provided advice and guidance for the development of the IIH Life questionnaire.
Cerebrospinal fluid diversion surgery rates (during the course of follow-up for this study) were lower for those with a healthy weight BMI at presentation (7%) compared to patients with IIH and higher BMI, with 3.5% of overweight, 19% of obesity class 1, and 15% of obesity class 2+ (Table 1).However, this was not statistically significantly different between any group.Similar proportions of patients were treated with ICP lowering medications in all groups: 57% of healthy weight BMI, 61% of overweight, 63% obesity class 1 and 54% obesity class 2 + .
The duration of follow-up in this cohort seen within our specialist service was a median of 17 months (range 1-86, interquartile range 6-37) for those with at least one follow-up visit.This was analogous to the BMI groups: healthy weight BMI
Visual outcomes over time, however, were affected differently (Figure 1).Visual acuity had a worse trajectory in those with a BMI <30 kg/m 2 compared to patients with obesity at baseline (Table 2, Figure 1).There was a small additional statistically significant worsening (increase) in acuity of 0.004 logunits/month (0.002, 0.0061) in the former group compared to the latter, although this would likely remain clinically imperceptible.The PMD trajectory did not differ between the groups with similar long-term outcomes (Figure 1).
Papilloedema OCT trajectories were similar between the groups (Table 2); however, in the initial six months following the baseline visit, patients with a BMI <30 kg/m 2 had a more rapid additional reduction in RNFL of 10.79 µm/month (−26.51,+4.92) and TRT of 5.90 µm/month (−36.29,+24.48) (Figure 1).This was not statistically significant.The macular ganglion cell layer volume declined slightly quicker in patients without obesity (Table 2), with a statistically significant greater decline of 0.000559 mm 3 /month (−0.001068, −0.00005).

Presence of obesity and headache outcomes
The headache and migraine-like headache frequencies were high in the IIH cohort at baseline (Table 2, Figure 2).2).Headache severity was moderate in both groups (Table 2) and did not significantly differ at baseline or in trajectory (Figure 2).Headaches had a substantial impact on the quality of life as measured by the HIT-6, with very similar scores and confidence intervals (baseline scores 61.57(58.04, 65.09) and 61.27 (59.74, 62.80) for BMI <30 kg/m 2 and ≥30 kg/m 2 , respectively) (Table 2, Figure 2).The trajectories were not significantly different between <30 kg/m 2 and ≥30 kg/m 2 groups for any of the headache outcomes.
Higher headache frequency was reported on average in those with higher baseline BMIs (Table 3, Figure 4).There was no group significantly different from healthy weight BMI at baseline (Figure 4) due to wide confidence intervals.Similarly, migraine-like headache frequency, headache severity, and HIT6 did not differ by BMI group at baseline (Figure 4b-d).
Trajectories for headache outcomes were similar for all outcomes (Table 3, Figure 4) apart from headache severity in the overweight group which had a worse trajectory by 0.15 units/ month (0.02, 0.29) compared to healthy weight BMI patients.

Discussion
In this prospective longitudinal cohort of people living with IIH, a healthy weight BMI at presentation is uncommon.The visual outcomes are similar between BMI groups; however, the improvement in papilloedema was observed to be more rapid in patients without obesity.The headache frequency was higher in IIH patients with obesity as compared to those without, and over time, the headache outcomes were similar between BMI groups.The prevalence of a healthy weight BMI in this study is similar to a previous large study of IIH. 9 In this cohort, those with a healthy weight BMI, including those without obesity, had worse visual field function at baseline, despite similar OCT papilloedema measures.This could potentially reflect a delay in diagnosis due to a lower suspicion of IIH in a person with a healthy weight BMI.The OCT imaging measures have been shown to be a more acute marker of disease activity, 5 as well as a better correlation to intracranial pressure than clinical symptoms. 34,35Thus, the comparable OCT measures, with worse PMD, may indicate a delayed presentation of the disease. 5The trend seen in CSF diversion surgery rates may reflect that patients with a higher BMI are more likely to present in fulminant IIH and thus require surgery, although this requires further evaluation in other cohorts as was not statistically significant in this cohort.A statistically significant difference was seen in the visual acuity and macular ganglion cell layer volume trajectories between those with obesity and those without (Table 2).However, the clinical significance is uncertain due to the small magnitude of this variation which is likely not be perceivable by the patient or of clinical significance.
The trajectories of the headache outcomes were not significantly different between <30 kg/m 2 and ≥30 kg/m 2 .This is consistent in that baseline BMI potentially and has a greater influence on visual outcomes than headache outcomes which may be influenced additionally by more significant factors as previous demonstrated in the literature. 5lthough IIH predominantly affects reproductive-aged women with obesity, 3,5,36,37 there are atypical presentations. 9,38Defining patients by whether obesity was present at baseline visit is clinically useful for potential management of their metabolic syndrome, although this involves an arbitrary cut off at 30 kg/m 2 which could miss some patients who could benefit.It is important for clinicians to note that baseline BMI did not impact the trajectory of visual or headache outcomes, while for some this could be reassuring; it is important to actively manage those who do not have obesity.
The only current disease-modifying therapy for the systemic manifestations of IIH is weight loss. 3,5,23,26,27Its role in patients with a normal baseline BMI is not yet understood.This would be a useful future analysis as a recent analysis of IIH Life database found that a change in BMI was a prognostic marker for visual outcomes. 5In other scenarios such as pregnancy, the role of weight loss is not clear. 16,39lthough the outcomes based on BMI are similar for people living with IIH, it should not be concluded that the underlying pathogenesis is the same.IIH has been shown to be associated with hypertestosteronism independent of obesity, 20 with metabolic priming towards lipogenesis. 18,19In people with obesity, significant weight loss of 24% body weight is often required for disease remission. 25,27,40otential pathogenesis in lean IIH could be inferred from the phenotypically similar condition of polycystic ovarian syndrome (PCOS).In PCOS, one study reported that 42% of the patients had a healthy weight BMI (defined as BMI 18.5-23 kg/ m 2 ) with a significantly lower prevalence in lean PCOS of deranged lipid profiles, impaired glucose tolerance test, and insulin resistance. 41The treatment for insulin resistance with metformin has been shown to lessen hyperandrogenism in PCOS 42 and has also been shown to potentially be of benefit in IIH. 43A limitation of the IIH life database is that metformin use is not routinely documented, and therefore, no further analysis could be done.In IIH with a normal baseline BMI, there may be a role for dietary modifications and pharmacotherapy to avoid future weight gain as this has been shown to be of benefit in lean PCOS. 44necdotally, clinicians have recommended weight loss in those with a normal BMI where there was a direct history of weight gain prior to IIH diagnosis, as long as not deemed detrimental to the patient's health. 3There is a clear need for further research into the metabolic profile of IIH in those living without obesity, as potentially, there could be a similar metabolic complication risk profile despite the lower body weight, and this could warrant an alternative management approach.
IIH with a healthy weight BMI, or without obesity, may more closely reflect another condition, spaceflight associated neuro-ocular syndrome (SANS) given the healthy weight BMIs observed in astronauts.SANS occurs during long duration spaceflight and risks include hyperopic refractive error shifts, acute visual loss, and cognitive changes.The OCT changes have been noted in SANS with some similarities to IIH, and an increase in the total retinal thickness of 20 µm is used as the earliest indication of optic disc oedema. 45,46Translational benefits for both conditions could be established if the underlying pathophysiology was determined.
The major limitation of this study was the relatively small sample size for healthy weight BMI patients.This was similar to a previous study 9 and likely reflective of the rarity of atypical IIH in an already rare condition.The baseline BMI was used to categorise patients; however, lead-time bias may have been induced with some initial visits being elsewhere prior to tertiary neuro-ophthalmology referral.To minimise this, we created a surrogate marker of disease duration which was defined as the time from the diagnostic lumbar puncture to the first encounter at baseline visit, adjusted for it in the multivariate regression modelling.It is acknowledged that patient reported symptom onset is not recorded within the IIH Life, and this may predate diagnosis by an uncertain and variable duration.Visual function assessments are challenging for patients to perform with both intra-visit and inter-visit factors that are known to influence reliability. 4,47,48Given this was a realworld clinical practice study, follow-up intervals were based on the decisions of the attending clinician at each visit and would be influenced by the person's clinical factors such as disease state, papilloedema severity, and symptomatology.As a result, individuals whose disease was in remission would be more rapidly discharged or their care transferred back to local hospitals.Hence, missing follow-up data increased over the course of time.In addition, some people were lost to follow-up.Therefore, caution in the over interpretation of the long-term outcomes as seen in the LOESS smoother curve should be applied due to the reducing sample size over the course of the study.The diagnosis of IIH is based on the revised diagnostic criteria. 30however, due to referral patterns, the cohort contains those with definite and those with probable diagnoses of IIH.It is often challenging to retrieve the initial diagnostic lumbar puncture opening pressure measurement.This analysis prompted and included many further inquiries to external hospital units for this piece of essential clinical data for the clinical record.The lumbar punctures are invasive and can be challenging. 49,50It is therefore important that guideline recommendations are adopted to ensure the accurate recording of the opening pressure and help reduce diagnostic uncertainty. 3,48,51

Conclusions
People with a normal BMI make up a minority of people diagnosed with IIH.They appear to have similar visual and headache outcomes to people living with IIH and obesity.While at the baseline visit, people without obesity had worse visual field function but similar papilloedema measures, compared to people living with obesity, and their longterm outcomes were similar.The underlying pathophysiology of those without obesity and in whom IIH develops is yet to be discovered.

Figure 1 .
Figure 1.Longitudinal visual data from baseline visit for IIH patients categorised by whether BMI < 30 or ≥30 kg/m 2 , and LOESS smoothers added to show trends across the categories.(a) LogMAR visual acuity (logunits) (b) Perimetric mean deviation measured by Humphrey visual field 24-2 testing (dB) (c) Retinal nerve fibre layer thickness measured on optical coherence tomography (µm) (d) Total retinal thickness of optic nerve head measured on optical coherence tomography (µm) (e) Macular ganglion cell layer volume measured on optical coherence tomography (mm 3 ).

Figure 2 .
Figure 2. Longitudinal headache data from baseline visit for IIH patients categorised by whether BMI <30 or ≥30 kg/m 2 , and LOESS smoothers added to show trends across the categories.(a) Headache frequency (days per month) (b) Migraine-like headache frequency (days per month) (c) Headache mean severity of predominant headache (0-10 numerical rating scale) (d) Headache Impact Test 6 (HIT6) (quality of life measure score 36-78).

Figure 3 .
Figure 3. Longitudinal visual data from baseline visit for IIH patients categorised baseline BMI group, and LOESS smoothers added to show trends across the categories.(a) LogMAR visual acuity (logunits) (b) Perimetric mean deviation measured by Humphrey visual field 24-2 testing (dB) (c) Retinal nerve fibre layer thickness measured on optical coherence tomography (µm) (d) Total retinal thickness of optic nerve head measured on optical coherence tomography (µm) (e) Macular ganglion cell layer volume measured on optical coherence tomography (mm 3 ).

Figure 4 .
Figure 4. Longitudinal headache data from baseline visit for IIH patients categorised by baseline BMI group, and LOESS smoothers added to show trends across the categories.(a) Headache frequency (days per month) (b) Migraine-like headache frequency (days per month) (c) Headache mean severity of predominant headache (0-10 numerical rating scale) (d) Headache Impact Test 6 (HIT6) (quality of life measure score 36-78).
funded during this study by Medical Research Council grant number [MR/KO15184/1], National Institute of Health Research grant [NIHR-CS-011-028] (clinician scientist fellowship), Sir Jules Thorne Award for Biomedical Science.

Table 1 .
Baseline table by baseline BMI.

Table 2 .
Baseline estimates and trajectory from regression modelling for obesity BMI cut-off (30 kg/m 2 ).

Table 3 .
Baseline estimates and trajectory for visual and headache outcomes by baseline BMI from regression modelling.