Characterizing idiopathic intracranial hypertension socioeconomic disparities and clinical risk factors: A retrospective case-control study

INTRODUCTION
Against the backdrop of the diverse minority-majority state of Hawaii, this study seeks to better characterize associations between idiopathic intracranial hypertension (IIH) with sociodemographic variables and medical comorbidities.


METHODS
A retrospective case-control study was conducted by utilizing 54 IIH patients and 216 age-, sex-, and race-matched controls, 216 unmatched controls, and 63 age-, sex-, and race-matched migraine patients.


RESULTS
Relative to controls, IIH were 25 years younger (p < 0.0001) and 10.18 kg/m2 heavier (p < 0.0001), as well as exhibited greater odds of the following variables (p < 0.05): female (odds ratio [OR]: 8.87), the lowest income quartile (OR: 2.33), Native Hawaiian or other Pacific Islander (NHPI; OR: 2.23), Native American or Alaskan Native (OR: 16.50), obesity class 2 (35.0-39.9 kg/m2; OR: 4.10), obesity class 3 (>40 kg/m2; OR: 6.10), recent weight gain (OR: 11.66), current smoker (OR: 2.48), hypertensive (OR: 3.08), and peripheral vascular disease (OR: 16.42). Odds of IIH were reduced (p < 0.05) for patients who were Asian (OR: 0.27) or students (OR: 0.30;). Unique from Whites, NHPI IIH patients exhibited greater odds (p < 0.05) for being from lower socioeconomic status and currently smoking, as well as potential association with seizures (p = 0.08). Compared to migraines, IIH headaches were at increased odds of occurring (p < 0.05) occipitally, for greater than 15 days per month, aggravated by postural changes, and comorbid with dizziness and tinnitus.


CONCLUSIONS
These results not only better characterize IIH, but also highlight socioeconomic and racial disparities in diagnosis.


Introduction
Without an identified etiology, idiopathic intracranial hypertension (IIH) is defined by an elevated intracranial pressure in the setting of normal ventricles [1]. While IIH presents with insidious headaches, in the absence of intervention, permanent vision loss can develop [2,3]. Hence, efficiently diagnosing IIH is paramount for averting morbidity. One avenue to improve accurate and timely IIH diagnosis is to better characterize the disorder's associated sociodemographic and medical comorbidities, to therefore increase a clinician suspicion for conducting an IIH diagnostic work-up.
To elucidate the potential socioeconomic, demographic, medical risk factors associated with IIH, we conducted a retrospective case-control study within the minority-majority population in the state of Hawai'i [4]. By utilizing Hawai'i, such enabled the investigation of IIH in relation to Native Hawaiians and other Pacific Islanders (NHPI), a population not only at greater predisposition to obesity-a predictor of IIH-, but also traditionally combined in demographic classifications with Abbreviations: 95% CI, 95% Confidence Interval; BMI, Body Mass Index; NHPI, Native Hawaiian or Other Pacific Islander; HPN, Hawai'i Pacific Neuroscience; IIH, Idiopathic Intracranial Hypertension; IIHTT, Idiopathic Intracranial Hypertension Treatment Trial; ICD-9, International Classification of Diseases 9th Edition; ICD-10, International Classification of Diseases 10th Edition; IQR, Interquartile Range; NAAN, Native American or Alaska Native; OR, Odds Ratio; ZIP, Zone Improvement Plan.
Asians, who themselves experience lower rates of obesity [5][6][7]. Furthermore, to better delineate IIH headaches, our investigation also compared headache characteristics between IIH and migraines.

Study design and setting
University of Hawai'i at Mānoa, Office of Research Compliance (protocol number: 2020-01010), provided institutional review board exception prior to study initiation. Electronic medical records at Hawai'i Pacific Neuroscience (HPN; Honolulu, Hawai'i) were retrospectively searched from January 1, 2009 to January 5, 2021. Patients were identified via the International Classification of Diseases 9th and 10th Editions, Clinical Modification, (ICD-9-CM or ICD-10) codes: 348.2 (ICD-9) and G93 (ICD-10). To meet inclusion, patients required diagnosis of IIH with the modified Dandy diagnostic criteria from the Idiopathic Intracranial Hypertension Treatment Trial (IIHTT) [8,9].

Outcome variables
Data collected for each IIH case included sociodemographic variables, Zone Improvement Plan (ZIP) code, clinical presentation, and medical comorbidities. Patient ZIP code served as a proxy measure for several socioeconomic variables, as described in a prior investigation [10]. IIH clinical variables included presenting symptoms to clinic (i.e., headache, visual disturbances, dizziness/syncope), symptoms at the time of diagnosis, and headache characteristics (i.e., duration, location, severity, laterality).

Controls
Four controls were selected for each case (n = 54) to maximize statistical power. Two sets of 216 randomly selected controls were collected from the HPN patient population (n = 29,049). The first set of controls (n = 216) was matched by age, sex, and race. To study differences with regards to age, sex, and race, the second set of controls was unmatched. For comparing headache characteristics of migraines against IIH, matched controls of migraine (n = 63) patients were compared to IIH.

Statistical analysis
Continuous variables were assessed by the independent Wilcoxon rank sum test, while categorical variables by either the Pearson's chisquared test or the Fisher's exact test of independence, with Haldane-Anscombe correction [11,12]. Univariate and multivariable logistic regression with Firth's correction, were conducted to identify variables independently predictive of IIH diagnosis [13]. All tests were two-tailed and used an alpha level of < 0.05 for statistical significance.

Overall prevalence and clinical characteristics of IIH cases
After identifying 63 IIH cases via ICD codes, nine were excluded for not meeting IIHTT modified Dandy diagnostic criteria, leaving 54 cases for analysis ( Table 1). The prevalence of IIH amongst the institute's population was 186 per 100,000 patients. The median incidence between 2010 and 2020 was 105 (Interquartile Range [IQR]: 73. 6,152) and annually static (Kendall's τ = − 0.31, p = 0.21).
The most common initial presentation of IIH patients to clinic (

Socioeconomic variables
Patients with IIH had a median household income of $4610 lower (p = 0.066) than controls. Stratified by income quartiles, IIH patients were at 2.33 (95% CI: 1.18, 4.58; p = 0.01) fold greater odds of being from the first income quartile (lowest income strata).
When examining the poverty level in patient's municipality of origin, IIH patients were found to reside in areas with a greater percentage of residents living below the poverty level (0.0050, 95% CI: − 0.000081, 0.011; p = 0.059). Stratified by age, IIH patients were similarly found to live in areas where a greater percentage of the residents 65 years and older living below the poverty level (0.0040, 95% CI: 0.000032, 0.016; p = 0.04).

Multivariable analysis of IIH entire cohort
After conducting the multivariable logistic regression (Table 6) with papilledema excluded, the strongest predictors increasing odds of diagnosis included, BMI (p = 0.01), other visual disturbances (p = 0.001), and poverty level for those 65 year and older (p = 0.03).

Native Hawaiian and other Pacific Islander IIH patients
Univariate logistic regression analysis was also conducted to identify unique association with IIH for NHPI patients. NHPI IIH patients were younger than controls (p < 0.0001) and at reduced odds of being male (0.03, 95% CI: 0.0014, 0.84; p = 0.04). NHPI IIH patients were at increased odds of living in municipalities with higher poverty rates across all age strata (p < 0.05) and having lower median household income (p = 0.03). Relative to the fourth income quartile, odds of IIH   Table 6 Multivariable logistic regression of idiopathic intracranial hypertension relative to controls. Race stratified analysis also included for Whites and NHPI.

Age at diagnosis
Patients with IIH had a median age of diagnosis at 32.00 years (IQR: 24.00, 40.75), which corroborated reports of individuals 18-44 years old most likely to have IIH [6,14,[20][21][22][23][24]. While differences in age of diagnosis by sex are yet to be reported, upon stratification by sex, median age of diagnosis was older amongst male IIH patients (37 years) than females (32 years); albeit, a low male number (n = 4) precluded identification of statistical significance (p = 0.28) [25].
When stratifying age of IIH diagnosis by race/ethnicity, Hispanic patients were found to have the youngest median age of diagnosis at 20.50 years (IQR: 20.25, 21.75), followed by NHPI at 29.00 years (IQR: 22.00, 37.25), Asians at 35.00 years (IQR: 31.25, 41.00), Whites at 35.00 years (IQR: 28.00, 41.50), NAAN at 47 years (IQR: 40.50, 53.50), and Blacks at 48 years (IQR: 35.50, 50.00); yet, insufficient sample sizes prevented meaningful statistical analysis. Notwithstanding, the absence of comparable studies examining age of diagnosis by race, our observed trends may be explained by the strong association between IIH and obesity [26]. The earlier age of diagnosis amongst Hispanics and NHPI, may arise secondary to higher rates of youth (2-19 years old) obesity amongst Hispanics and NHPI [27][28][29]. However, given Blacks and NAAN have greater prevalence of childhood obesity than Whites, yet older age of IIH diagnosis, there are potentially other environmental factors at play (i.e., disparities in healthcare access for timely diagnosis) contributing to the different ages of diagnosis per each racial strata [29].

Sex
The observation that females are at 8.87 greater odds of IIH diagnosis, with a female-to-male ratio of 12.5:1, corresponds with earlier studies, including one United States nationwide report finding females to have an incidence 5.47 times greater than males [14][15][16][17][18][19][20]24,30]. Given the female predisposition for IIH does not occur until after puberty, some hypothesize sex hormones may contribute to the pathogenesis and female predilection for IIH [14,31].

Race
Hawai'i's diverse minority-majority population enabled our investigation to examine disparities in IIH amongst NHPI, NAAN, Asians, and Blacks-historically marginalized populations which are often understudied [32,33]. The racial groups experiencing the greatest odds of IIH included NAAN (odds ratio [OR]: 16.50; p = 0.006), Blacks (OR: 6.23; p = 0.004), and NHPI (OR: 2.23; p = 0.02); while Asians (OR: 0.27; p = 0.004) experienced the lowest odds. Observed trends for Blacks and Asians corresponded with US nationwide data on IIH, yet for NAAN and NHPI comparable data was not available [14,21]. The greater odds of IIH amongst NAAN, Blacks, and NHPI, may be explained by the higher rates of obesity these populations experience relative to Whites [5][6][7]. Similarly, lower rates of IIH amongst Asians, could be attributed to Asians having one of the lowest rates of obesity in the US and internationally [5,6,14,30,34,35].

Socioeconomic variables
Upon examining several markers of socioeconomic status, patients in our IIH cohort were found to be at greater odds of having a lower median household income (in particular having greater odds of being from the lowest income quartile) and to reside in areas with a greater proportion of the residents living below the poverty line. The association with lower socioeconomic status is not confined to Hawai'i, as nationwide an increased incidence of IIH in the lowest income quartile was also identified [14]. Internationally, in England and Wales, areas with higher deprivation have also been noted to have greater rates of IIH, with the trend subsisting in females even after controlling for obesity-paralleling our finding of poverty level amongst those 65 years and older remaining an independent predictor of IIH diagnosis after multivariable analysis [36,37]. Hence, the link between lower socioeconomic status and increased rates of obesity may only explain part of the association between IIH and higher deprivation, with other variables tied to lower socioeconomic status playing a greater role (i.e., diet, employment, education, social support, and environmental hazards) [14,[36][37][38].
When examining marital and occupational status, despite matching to age, higher odds of IIH diagnosis were noted for patients who were married (OR: 1.82; p = 0.07), while lower odds for those single (OR: 0.53; p = 0.05) or students (OR: 0.30; p = 0.02). With marriage, women have been noted to gain weight, while with increasing education level obesity prevalence decreases, hence potentially accounting for the IIH diagnosis trends [5,39].

Cardiovascular risk factors
Overall, several cardiovascular comorbidities were found associated with IIH, including BMI, smoking status, hypertension, hyperlipidemia, and peripheral vascular disease. Regarding the established association of obesity, our IIH patients were 10.18 kg/m 2 heavier than controls, and at greater odds of being in obesity class 2 (35.0-39.9 kg/m 2 , OR: 4.10; p = 0.0002) and class 3 (> 40 kg/m 2 , OR: 6.10; p < 0.0001), as well as at greater odds of reporting recent weight gain (p < 0.0001) [15,16,20,24,36]. The association with obesity likely also accounts for the greater odds of hypertension (OR: 3.08; p = 0.006) in IIH patients, notwithstanding prior investigations having found that even after matching for age and BMI, IIH patients continue to have a 55% greater risk of hypertension [38,[40][41][42]. Other variables, including hyperlipidemia (OR: 2.09; p = 0.12), peripheral vascular disease (OR: 16.42; p = 0.007), and being a current smoker (OR: 2.48; p = 0.04) all associated with IIH diagnosis; previous studies have also reported increased rates of smoking in patients with IIH [42,43]. Provided the association with cardiovascular variables in addition to BMI, IIH patients may benefit from not only weight management, but also hypertensive control, lipid management, and smoking cessation [42,44].

White and NHPI IIH patients
Subgroup analysis for White and NHPI patients was also conducted. For White patients, IIH was not associated with a lower socioeconomic status, but did retain correlations with greater BMI and hypertension. However, NHPI patients did present an association between lower socioeconomic status and IIH diagnosis. Moreover, for NHPI IIH patients, the strongest predictors of diagnosis included being a current smoker, in addition to higher BMI. NHPI IIH patients also exhibited increased odds of seizure history (OR: 8.78; p = 0.08), an association only rarely reported [45]. While seizure may be secondary to severe IIH disproportionally impacting NHPIs, seizures arising due to elevated intracranial pressure is uncommon and a controversial [46].

Comparison of migraines against IIH
Given the similarities between migraine and IIH headaches, as well as potential for permanent vision loss secondary to untreated IIH, there is impetus to efficiently distinguish IIH headaches to expedite correct treatment [1,17,37]. Our investigation identified that IIH, relative to migraine patients, were at greater odds of experiencing headaches in the occipital region (OR: 5.47; p = 0.04) and at a frequency of more than 15 days per month (OR: 5.31; p = 0.002). Similarly, IIH patients in the IIHTT reported constant/daily headache or intermittent headaches occurring for 12 days per month [6]. IIH headaches were also at greater odds of association with dizziness, tinnitus, loss of functionality, and to be exacerbated by postural changes, while migraine headaches were associated with nausea/vomiting, photophobia, and phonophobia. Prior investigations have also found these symptoms associated with IIH headaches, with increased intracranial pressure the suspected culprit [15,47,48]. Loss of functionality in particular has been noted to decreases quality of life for IIH patients [48]. Overall, by eliciting specific questions to better characterize headaches, a correct diagnosis of IIH may be expedited.

Limitations
The results should be considered in the context of several limitations. As a retrospective investigation there remains uncertainty in how consistently clinical history was elicited from patients. The relatively low sample size-including for the stratified race analysis of NPHI patients-significantly limits statistical power and the ability to appreciate many of the other associations which may exist within the IIH populations. Additionally, while this study utilized the modified Dandy criteria for IIH diagnosis, there is no international consensus for a goldstandard diagnostic protocol. Moreover, the accuracy of certain variables may be influenced by recall bias, such as recent weight gain, regular exercise, headache characterization, smoking history, drug use, and alcohol consumption. Finally, ICD codes themselves are susceptible to administrative errors in data input, which could yield in omitted patients.

Conclusion
The investigation identified several unique associations amongst IIH patients (Tables 7 and 8). Racial disparities were recognized, in that NHPI, NAAN, and Blacks were all disproportionately affected by IIH, with Asians at reduced odds. Additionally, the consistent correlation with lower socioeconomic status and cardiovascular risk factors (i.e., obesity, active smoking, hypertension, peripheral vascular disease) emphasize IIH patients are a vulnerable population, who may not have the means to follow-up with resource-intensive treatments. In particular, NHPI patients were found to have a much stronger association with lower socioeconomic status and smoking, relative to Whites, as well as exhibit a possible association with seizures. Meanwhile, when examining the clinical presentation, IIH headaches were found to be distinguished from migraines by occurring predominantly in the occipital region, greater than 15 days per month, exacerbated by postural changes, and linked with tinnitus, dizziness, and loss of functionality. In summary, by better characterizing IIH and recognizing the disparities within the disorder, these results not only may help expedite diagnosis, but also improve the quality of life for subsets of the IIH population.

Funding
None.

Conflicts of interest/Competing interests
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Table 7
Summary of variables associated with idiopathic intracranial hypertension compared to general controls and ischemic stroke. *Variables determined to be statistically significant after multivariable analysis. Variables with marginal significance (p < 0.1) also presented, as low sample size likely limited attainment of significance.

Code availability
Software application or custom code: Not applicable.

Ethics approval
Institutional review board exemption; University of Hawai'i at Mānoa, Office of Research Compliance (protocol number: 2020-01010).

Consent to participate
Not applicable.

Consent for publication
All authors approved the submitted manuscript version.

Availability of data and material (data transparency)
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.