Multivariate Normative Comparison Research Articles

Multivariate normative comparison, a novel method for improved use of the retinal nerve fiber layer thickness to detect early glaucoma

AbstractPurposeTo assess whether compensating the RNFL thickness for multiple demographic and anatomical factors improves the detection of glaucoma.MethodsThis case‐control study enrolled 2699 healthy participants to construct and test a multivariate compensation model, which was then applied in 387 healthy controls and 387 glaucoma cases (early, n = 219; moderate, n = 97; and, advanced, n = 71). Participants underwent Cirrus spectral‐domain OCT (Carl Zeiss Meditec) imaging of the optic disc and macular cubes. Compensated RNFL thickness was generated based on ethnicity, age, refractive error, optic disc (ratio, orientation, and area), fovea (distance and angle), and retinal vessel density. RNFL thickness measurements and their corresponding areas under the receiver operating characteristic curves (AUCs) were obtained.ResultsAfter applying the Asian‐specific compensation model, the standard deviation (SD) of RNFL thickness reduced, where the effect was greatest for Chinese (16.9%), followed by Malays (13.9%) and then Indians (12.1%). Multivariate normative comparison outperformed measured RNFL for discrimination of early glaucoma (AUC = 0.90 vs 0.85; p < 0.001), moderate glaucoma (AUC = 0.94 vs 0.91; p < 0.001) and advanced glaucoma (AUC = 0.98 vs 0.96; p < 0.001).ConclusionsThe multivariate normative database of RNFL showed better glaucoma discrimination capability than measured RNFL thickness, which suggests there may be utility in accounting for demographic and anatomical variance in RNFL thickness to improve glaucoma detection.

Multivariate Normative Comparison, a Novel Method for Improved Use of Retinal Nerve Fiber Layer Thickness to Detect Early Glaucoma

Detection of early glaucoma remains limited with the conventional analysis of the retinal nerve fiber layer (RNFL). This study assessed whether compensating the RNFL thickness for multiple demographic and anatomic factors improves the detection of glaucoma. Cross-sectional study. Three hundred eighty-seven patients with glaucoma and 2699 healthy participants. Two thousand six hundred ninety-nine healthy participants were enrolled to construct and test a multivariate compensation model, which then was applied in 387 healthy participants and 387 patients with glaucoma (early glaucoma, n= 219; moderate glaucoma, n= 97; and advanced glaucoma, n= 71). Participants underwent Cirrus spectral-domain OCT (Carl Zeiss Meditec) imaging of the optic disc and macular cubes. Compensated RNFL thickness was generated based on ethnicity, age, refractive error, optic disc (ratio, orientation, and area), fovea (distance and angle), and retinal vessel density. The RNFL thickness measurements and their corresponding areas under the receiver operating characteristic curve (AUCs) were obtained. Measured and compensated RNFL thickness measurements. After applying the Asian-specific compensation model, the standard deviation of RNFL thickness reduced, where the effect was greatest for Chinese participants (16.9%), followed by Malay participants (13.9%), and Indian participants (12.1%). Multivariate normative comparison outperformed measured RNFL for discrimination of early glaucoma (AUC, 0.90 vs. 0.85; P < 0.001), moderate glaucoma (AUC, 0.94 vs. 0.91; P < 0.001), and advanced glaucoma (AUC, 0.98 vs. 0.96; P < 0.001). The multivariate normative database of RNFL showed better glaucoma discrimination capability than conventional age-matched comparisons, suggesting that accounting for demographic and anatomic variance in RNFL thickness may have usefulness in improving glaucoma detection.

Brain White Matter Microstructure as a Risk Factor for Cognitive Decline After Chemotherapy for Breast Cancer.

Cognitive decline is frequently observed after chemotherapy. As chemotherapy is associated with changes in brain white matter microstructure, we investigated whether white matter microstructure before chemotherapy is a risk factor for cognitive decline after chemotherapy. Neuropsychologic tests were administered before and 6 months (n = 49), 2 years (n = 32), and 3 years (n = 32) after chemotherapy in patients with breast cancer receiving anthracycline-based chemotherapy (BC + CT group), at matched intervals to patients with BC who did not receive systemic therapy (BC - CT group: n = 39, 23, and 19, respectively) and to no-cancer controls (NC group: n = 37, 29, and 28, respectively). Using multivariate normative comparison, we evaluated to what extent the cognitive profiles of patients deviated from those of controls. Fractional anisotropy (FA), derived from magnetic resonance diffusion tensor imaging, was used to measure white matter microstructure before treatment. FA was evaluated as a risk factor for cognitive decline, in addition to baseline age, fatigue, cognitive complaints, and premorbid intelligence quotient. We subsequently ran voxel-wise diffusion tensor imaging analyses to investigate white matter microstructure in specific nerve tracts. Low FA independently predicted cognitive decline early (6 months, P = .013) and late (3 years, P < .001) after chemotherapy. FA did not predict cognitive decline in the BC - CT and NC groups. Voxel-wise analysis indicated involvement of white matter tracts essential for cognitive functioning. Low FA may reflect low white matter reserve. This may be a risk factor for cognitive decline after chemotherapy for BC. If validated in future trials, identification of patients with low white matter reserve could improve patient care, for example, by facilitating targeted, early interventions or even by influencing choices of patients and doctors for receiving chemotherapy.

Correlation between computerised and standard cognitive testing in people with HIV and HIV-negative individuals

ABSTRACT We investigated the correlations and agreement between cognitive assessments made using a computerised (CogState™, six domains) and a standard pen-and-paper battery (five domains) in PWH and lifestyle-similar HIV-negative individuals. Demographically adjusted domain and global T-scores were obtained and used to define cognitive impairment according to the multivariate normative comparison (MNC) criteria. Correlations between T-scores and the agreement between the classifications of cognitive impairment obtained from the two batteries were assessed using the Spearman’s rank correlation and Cohen’s κ, respectively. The correlation between global T-scores from the two batteries was 0.52 (95% CI 0.44–0.60) in PWH and 0.45 (0.29–0.59) in controls (p = 0.38 for their difference). Correlations were generally stronger between domains within the same battery than between those from different batteries. The agreement between the two batteries in classifying individuals as cognitively impaired or not impaired was fair in PWH (κ = 0.24) and poor in HIV-negative individuals (κ = −0.02). The moderate correlation between overall cognitive function and the modest agreement between binary classifications of cognitive impairment obtained from two different batteries indicate the two batteries may assess slightly different components of cognition.

Lower IQ and poorer cognitive profiles in treated perinatally HIV-infected children is irrespective of having a background of international adoption.

HIV-associated cognitive deficiency in perinatally HIV-infected (PHIV) children has been studied in Western countries in a population of which an increasing proportion has been internationally adopted. Studies often lack an appropriate internationally adopted HIV-uninfected control group, potentially confounding the relationship between HIV and cognitive functioning. This study aims to further elucidate the association between treated HIV infection and cognitive development by addressing the background of international adoption. We cross-sectionally studied the impact of HIV on cognition by comparing PHIV children and HIV- uninfected controls, matched for age-, sex-, ethnicity-, socioeconomic status (SES)- and adoption status. We used a standardized neuropsychological test battery to measure intelligence (IQ), and the cognitive domains of processing speed, working memory, executive function, learning ability and visual-motor function and compared outcomes using lineair regression models, adjusted for IQ. We determined cognitive profiles and cognitive impairment by using multivariate normative comparison (MNC) and explored associations with HIV disease- and treatment-related factors. We enrolled fourteen PHIV children (mean age 10.45 years [1.73 SD], 93% adopted from sub-Saharan Africa at a median age of 3.3 years [IQR 2.1-4.2]) and fifteen HIV- uninfected controls. Groups did not clinically nor statistically differ in age, sex, ethnicity, SES, region of birth, adoption status and age at adoption. PHIV scored consistently lower on all cognitive domains and MNC outcomes. Compared to controls, PHIV children had a significant lower IQ (mean 81 [SD 11] versus mean 97 [SD 15], p = 0.005), and a poorer cognitive profile by MNC (Hotelling's T2 mean -4.36 [SD 5.6] versus mean 0.16 [SD 4.5], p = 0.021), not associated with HIV disease- and treatment-related factors. Two PHIV (14%) and one control (7%) were classified as cognitively impaired (p = 0.598). Findings indicate treated HIV-infection to be independently associated with lower IQ and poorer cognitive profiles in PHIV children, irrespective of a background of international adoption.

Cognitive impairments in patients with persistent symptoms attributed to Lyme disease

BackgroundPersistent symptoms attributed to Lyme borreliosis often include self-reported cognitive impairment. However, it remains unclear whether these symptoms can be substantiated by objective cognitive testing.MethodsFor this observational study, cognitive performance was assessed in 280 adults with persistent symptoms attributed to Lyme borreliosis (as part of baseline data collected for the Dutch PLEASE study). Cognitive testing covered the five major domains: episodic memory, working memory / attention, verbal fluency, information-processing speed and executive function. Patients’ profiles of test scores were compared to a large age-, education- and sex-adjusted normative sample using multivariate normative comparison. Performance validity was assessed to detect suboptimal effort, and questionnaires were administered to measure self-reported cognitive complaints, fatigue, anxiety, depressive symptoms and several other psychological factors.ResultsOf 280 patients, one was excluded as the test battery could not be completed. Of the remaining 279 patients, 239 (85.4%) displayed sufficient performance validity. Patients with insufficient performance validity felt significantly more helpless and physically fatigued, and less orientated. Furthermore, they had a lower education level and less often paid work. Of the total study cohort 5.7% (n = 16) performed in the impaired range. Among the 239 patients who displayed sufficient performance validity, 2.9% (n = 7) were classified as cognitively impaired. No association between subjective cognitive symptoms and objective impairment was found.ConclusionsOnly a small percentage of patients with borreliosis-attributed persistent symptoms have objective cognitive impairment. Performance validity should be taken into account in neuropsychological examinations of these patients. Self-report questionnaires are insufficiently valid to diagnose cognitive impairment.Trial registrationClinicalTrials.gov NCT01207739. Registered 23 September 2010.

Increased Prevalence of Neurocognitive Impairment in Aging People Living With Human Immunodeficiency Virus: The ANRS EP58 HAND 55-70 Study.

There are limited data on the comparative prevalence of neurocognitive impairment (NCI) in aging people living with human immunodeficiency virus (PLHIV) and people not living with HIV. This was a cross-sectional study of PLHIV randomly matched by age (±4 years), gender, and education with 5 HIV-uninfected individuals from the CONSTANCES cohort. PLHIV were fluent in French and sequentially included during routine outpatient visits if aged 55-70 years, with HIV viral load <50 copies/mL, and lymphocyte T-CD4 level ≥200 cells/µL in the past 24 and 12 months, respectively. The primary outcome was NCI as defined by the Frascati criteria. Multivariate normative comparison (MNC) and -1.5 standard deviations in ≥2 neurocognitive domains were secondary outcomes of NCI. Two hundred PLHIV were matched with 1000 controls. Median age was 62 years, and 85% were men. In PLHIV, the median T-CD4 lymphocyte level was 650 cells/µL, and median nadir T-CD4 lymphocyte level was 176 cells/µL. NCI was found in 71 (35.5%) PLHIV and in 242 (24.2%) controls (odds ratio [OR], 1.74; 95% confidence interval [CI], 1.25, 2.41). After adjusting for confounders, HIV remained significantly associated with NCI (OR, 1.50; 95% CI, 1.04, 2.16). Adjusted results were similar with NCI defined by MNC (ORMNC, 2.95; 95% CI, 1.13, 3.50) or -1.5 SD (OR-1.5, 2.24; 95% CI, 1.39, 3.62). In this matched study of aging individuals, HIV was significantly associated with an increased risk of NCI after adjusting for major confounders. Results were confirmed with more stringent NCI classifications. NCT02592174.

Clinical and neuroimaging correlates of cognition in HIV

This study investigated whether HIV-positive participants, stable on combined antiretroviral therapy (cART), showed cognitive impairments relative to HIV-negative controls; and whether clinical and neuroimaging factors correlated with cognitive function in the HIV-positive participants. One hundred and twenty-six white men who have sex with men, of whom 78 were HIV-positive and stable on cART and 48 were HIV negative, were recruited to this cross-sectional study. The median age of HIV-positive participants in this study was 47. They underwent clinical and neuropsychological evaluation and magnetic resonance imaging of the brain, including diffusion tensor imaging (DTI). Cognitive scores for both groups were compared, and regression models were run to explore the influence of clinical, psychiatric, lifestyle, and neuroimaging variables on cognition. The prevalence of neurocognitive impairment, using the multivariate normative comparison criteria, was 28% in HIV-positive participants and 5% in HIV-negative participants. After covarying for age, years of education, and non-English speaking background, there were significant differences between the HIV group and the controls across four cognitive domains. The HIV group showed significantly higher mean diffusivity (MD) and lower fractional anisotropy (FA) than the control group on DTI. Although anxiety levels were clinically low, anxiety and DTI measures were the only variables to show significant correlations with cognitive function. In the HIV group, poorer cognitive performance was associated with higher MD and lower FA on DTI and higher (albeit clinically mild) levels of anxiety. Our findings suggest that white matter changes and subtle anxiety levels contribute independently to cognitive impairment in HIV.

Neurocognitive Development in Perinatally Human Immunodeficiency Virus-infected Adolescents on Long-term Treatment, Compared to Healthy Matched Controls: A Longitudinal Study.

A cross-sectional analysis of the Neurological, cOgnitive and VIsual performance in hiv-infected Children cohort showed significant cognitive impairment in combination antiretroviral therapy (cART)-treated, perinatally human immunodeficiency virus (HIV)-infected adolescents (PHIV+) compared to age-, sex-, ethnicity- and socioeconomic status (SES)-matched HIV-negative controls (HIV-). In this longitudinal study, we compared cognitive development in the same adolescents over time. We repeated the standardized cognitive test battery after a mean of 4.6 years (standard deviation 0.3). In participants who completed both assessments, we compared cognitive trajectories between groups in the domains of intelligence quotient (IQ), processing speed, working memory, executive functioning, learning ability, and visual-motor function, using linear mixed models. We explored associations with disease- and treatment-related factors and used multivariate normative comparison (MNC) to determine the prevalence of cognitive impairment. There were 21 PHIV+ and 23 HIV- participants that completed 2 assessments and were similar concerning age, sex, ethnicity, and SES. Compared to HIV- participants, in PHIV+ participants the IQ score increased significantly more over time (group*time 6.01, 95% confidence interval [CI] 1.5-10.50; P = .012), whereas executive functioning decreased significantly more (group*time -1.43 z score, 95% CI -2.12 to -0.75; P < .001), resulting in the disappearance and appearance of significant differences. Processing speed, working memory, learning ability, and visual-motor function trajectories were not statistically different between groups. Univariately, those who had started cART at an older age deviated more in executive functioning (-0.13 z score, 95% CI -0.24 to -0.02; P = .043). The prevalence of cognitive impairments by MNC was similar in both groups, at both time points. The cART-treated PHIV+ adolescents appeared to have similar global cognitive development, compared to their healthy peers. Executive functioning trajectory appears to deviate, potentially explained by earlier brain damage.

Multivariate normative comparisons for neuropsychological assessment by a multilevel factor structure or multiple imputation approach.

Neuropsychologists administer neuropsychological tests to decide whether a patient is cognitively impaired. This clinical decision is made by comparing a patient's scores to those of healthy participants in a normative sample. In a multivariate normative comparison, a patient's entire profile of scores is compared to scores in a normative sample. Such a multivariate comparison has been shown to improve clinical decision making. However, it requires a multivariate normative data set, which often is unavailable. To obtain such a multivariate normative data set, the authors propose to aggregate healthy control group data from existing neuropsychological studies. As not all studies administered the same tests, this aggregated database will contain substantial amounts of missing data. The authors therefore propose two solutions: multiple imputation and factor modeling. Simulation studies show that factor modeling is preferred over multiple imputation, provided that the factor model is adequately specified. This factor modeling approach will therefore allow routine use of multivariate normative comparisons, enabling more accurate clinical decision making. (PsycINFO Database Record

Defining cognitive impairment in people-living-with-HIV: the POPPY study.

BackgroundThe reported prevalence of cognitive impairment (CI) varies widely in cohorts of people living with HIV (PLWH); this may partly be due to the use of different diagnostic criteria. Agreement between diagnostic criteria of CI, the optimal definition to use, and associations with patient-reported cognitive symptoms have not been fully investigated.MethodsTwo hundred ninety PLWH aged >50 years and 97 matched negative controls completed a detailed assessment of cognitive function and three questions regarding cognitive symptoms. Age- and education-adjusted test scores (T-scores) determined if subjects met the following definitions of CI: Frascati, global deficit score (GDS) and the multivariate normative comparison (MNC) method.ResultsPLWH were more likely than controls to meet each definition of CI (ORs were 2.17, 3.12 and 3.64 for Frascati, GDS and MNC, respectively). Agreement of MNC with Frascati and GDS was moderate (Cohen’s k = 0.42 and 0.48, respectively), whereas that between Frascati and GDS was good (k = 0.74). A significant association was found between all the three criteria and reporting of memory loss but not with attention and reasoning problems. The 41 (14 %) PLWH meeting all the three criteria had the lowest median global T-score (36.9) and highest rate of symptom reporting (42 %).ConclusionsDifferent CI criteria show fair diagnostic agreement, likely reflecting their ability to exclude CI in the same group of individuals. Given the lower overall cognitive performance and higher rates of symptom reporting in those meeting all three criteria of CI, further work assessing this as a definition of CI in PLWH is justified.Electronic supplementary materialThe online version of this article (doi:10.1186/s12879-016-1970-8) contains supplementary material, which is available to authorized users.

Associations between cognitive impairment and patient-reported measures of physical/mental functioning in older people living with HIV.

While cognitive impairment is frequently reported in HIV-positive individuals and has historically been associated with poorer functional outcomes, the associations between cognitive impairment and patient-reported outcome measures (PROMs) in contemporary cohorts are unclear. We tested cognitive function using a computerized battery (CogState™ ) in 290 HIV-positive and 97 HIV-negative individuals aged ≥ 50 years participating in the Pharmacokinetic and Clinical Observations in People Over Fifty (POPPY) study. Participants completed questionnaires detailing physical and mental health [Short Form Health Survey (SF-36)], cognitive function [European AIDS Clinical Society (EACS) questions], activities of daily living [Lawton Instrumental Activities of Daily Living (IADL)], depression [Patient Depression Questionnaire (PHQ-9) and Centres for Epidemiologic Studies Depression scale (CES-D)], falls and sexual desire. Cognitive impairment was defined using the Frascati criteria, global deficit score (GDS) and multivariate normative comparison (MNC). In the HIV-positive group, the classification performances of the different definitions of cognitive impairment and dichotomized questionnaire results were calculated. The prevalence of cognitive impairment in the HIV-positive group was 34.5% (GDS), 30.0% (Frascati) and 22.1% (MNC), with only 2% diagnosed with HIV-associated dementia. In general, the associations between cognitive impairment and PROMs were weak regardless of the definition used: mean c-statistics were 0.543 (GDS), 0.530 (MNC) and 0.519 (Frascati). Associations were similar using the global T-score to define cognitive impairment. Summary health scores (SF-36) were lower, but only significantly so for those with cognitive impairment identified using MNC, for both mental health (61.4 vs. 75.8; P = 0.03) and physical health (60.9 vs. 75.0; P = 0.03). The associations between cognitive impairment and PROMs were weak, possibly because impairment was mild and therefore largely asymptomatic. Further work is needed to elucidate the clinical implications of cognitive impairment in HIV-disease.

Cognitive Impairment in a Subset of Breast Cancer Patients After Systemic Therapy—Results From a Longitudinal Study

ContextStudies indicate adverse effects of breast cancer (BC) and cancer treatment on cognitive function. ObjectivesTo investigate the effects of systemic treatment on cognitive performance in BC patients. MethodsParticipants were BC patients scheduled to receive systemic treatment (BC + SYST; n = 31), or no systemic treatment (BC; n = 24) and no-cancer (NC) controls (n = 33). Neuropsychological examinations were used to study cognitive performance on 18 tests grouped into eight cognitive domains, before adjuvant treatment (T1) and six months after chemotherapy (T2), or at similar intervals. We also assessed health-related quality of life, anxiety and depression, mood, stress, and cognitive problems. Analysis of variance was used to assess group differences of cognitive performance and multivariate normative comparison to classify impairment, comparing scores of each participant against the distribution of the scores of NC controls. ResultsOf BC + SYST, 16% were cognitively impaired at T2, compared to 4% in BC and 6% in NC. Although not significant, we observed moderate effect sizes for worse performance in the BC + SYST group compared to NC (Flanker congruent [effect size {ES} = 0.44] and stimulus incongruent [ES = 0.44]) and compared to BC (Controlled Oral Word Association Test [ES = 0.47], digit span [ES = 0.41], and Hopkins Verbal Learning Test immediate [ES = 0.71] and delayed recall [ES = 0.65]). Cognitively impaired patients had a significantly lower estimated premorbid intelligence, worse physical and social functioning, and more distress at T2 compared to unimpaired patients. ConclusionOur findings indicate that cognitive impairment after systemic treatment occurs in a subset of BC patients. The predictive value of demographic and psychosocial factors in cognitive impairment should be further investigated in a larger sample of impaired patients.

Determinants of reduced cognitive performance in HIV-1-infected middle-aged men on combination antiretroviral therapy.

The spectrum of risk factors for HIV-associated cognitive impairment is likely very broad and includes not only HIV/antiretroviral therapy-specific factors but also other comorbid conditions. The purpose of this current study was to explore possible determinants for decreased cognitive performance. Neuropsychological assessment was performed on 103 HIV-1-infected men with suppressed viraemia on combination antiretroviral therapy for at least 12 months and 74 HIV-uninfected highly similar male controls, all aged at least 45 years. Cognitive impairment and cognitive performance were determined by multivariate normative comparison (MNC). Determinants of decreased cognitive performance and cognitive impairment were investigated by linear and logistic regression analysis, respectively. Cognitive impairment as diagnosed by MNC was found in 17% of HIV-1-infected men. Determinants for decreased cognitive performance by MNC as a continuous variable included cannabis use, history of prior cardiovascular disease, impaired renal function, diabetes mellitus type 2, having an above normal waist-to-hip ratio, presence of depressive symptoms, and lower nadir CD4⁺ cell count. Determinants for cognitive impairment, as dichotomized by MNC, included cannabis use, prior cardiovascular disease, impaired renal function, and diabetes mellitus type 2. Decreased cognitive performance probably results from a multifactorial process, including not only HIV-associated factors, such as having experienced more severe immune deficiency, but also cardiovascular/metabolic factors, cannabis use, and depressive symptoms.

Poorer cognitive performance in perinatally HIV-infected children versus healthy socioeconomically matched controls.

Despite the declining incidence of severe neurological complications such as HIV encephalopathy, human immunodeficiency virus (HIV) infection in children is still associated with a range of cognitive problems. Although most HIV-infected children in industrialized countries are immigrants with a relatively low socioeconomic status (SES), cognitive studies comparing HIV-infected children to SES-matched controls are lacking. This cross-sectional study included perinatally HIV-infected children and controls matched for age, sex, ethnicity, and SES, who completed a neuropsychological assessment evaluating intelligence, information processing speed, attention, memory, executive function, and visual-motor function. Multivariate normative comparison was used to assess the prevalence of cognitive impairment in the HIV-infected group. Multivariable regression analyses were performed to identify HIV- and combination antiretroviral therapy-related factors associated with cognitive performance. In total, 35 perinatally HIV-infected children (median age, 13.8 years; median CD4 count, 770 × 10(6) cells/L; 83% with undetectable HIV RNA) and 37 healthy children (median age, 12.1 years) were included. HIV-infected children scored lower than the healthy controls on all cognitive domains (eg, intelligence quotient [IQ], 76 [standard deviation {SD}, 15.7] vs 87.5 [SD, 13.6] for HIV-infected vs healthy children; P = .002). Cognitive impairment was found in 6 HIV-infected children (17%). The Centers for Disease Control and Prevention (CDC) clinical category at HIV diagnosis was inversely associated with verbal IQ (CDC clinical category C: coefficient -22.98; P = .010). Our results show that cognitive performance of HIV-infected children is poor compared with that of SES-matched healthy controls. Gaining insight into these cognitive deficits is essential, as subtle impairments may progress to more pronounced complications that will influence future intellectual performance, job opportunities, and community participation of HIV-infected children.

Departure from normality in multivariate normative comparison: The Cramér alternative for Hotelling's T2

Crawford and Howell (1998) have pointed out that the common practice of z-score inference on cognitive disability is inappropriate if a patient's performance on a task is compared with relatively few typical control individuals. Appropriate univariate and multivariate statistical tests have been proposed for these studies, but these are only valid if the data are Gaussian (normal) distributed. Previous studies have investigated the consequences for Type I error rates of using the univariate test when data are not Gaussian. In this paper we examine the effects of violation of the Gaussian assumption on nominal Type I error rates for the multivariate test. We also consider a new test that has been devised recently, called Cramér's test, as a viable alternative for the multivariate normative comparison. In simulations we show that the new test not only provides a distribution free alternative for existing methods, but also has the advantage that it is substantially more powerful in most common research settings. We demonstrate the use of the new test with an application to two individuals diagnosed with autism.