Serial Tap Test of patients with idiopathic normal pressure hydrocephalus: impact on cognitive function and its meaning

Study subjects were initially searched at the CSF Circulation Disturbance Research Program 2004–2017 files. This database contains information on well-established standardized procedures carried out at the Neurological Institute of Curitiba in suspected INPH subjects. Of the hundred and forty-eight subjects suspected of having INPH initially found, ten were not included for being aged 58 years or less, 27 for being classified as “unlikely”, three for having been previously submitted to a neurosurgical procedure and further twelve for refusing to consent with the study procedures and other reasons (e.g.: missing data). Of the remaining 96 individuals with "probable" or "possible" INPH (according to the 2005 Euro-American Consensus) [5] were selected. Of those 96 subjects, 30 individuals were excluded due to comorbidities that could potentially impact cognitive functioning, such as active alcoholism, Parkinson's disease or parkinsonism, Alzheimer's disease, epilepsy, central nervous system [CNS] tumor, other neurological diseases such as vascular dementia, Lewy body disease, and multiple sclerosis; psychiatric disorders such as major depressive disorder, bipolar affective disorder, and attention deficit and cerebrovascular disorders (Fig. 1 and Table 1).

Individuals with a history of depression under control for more than three years or a history of mild TBI without permanent sequelae (cognitive or motor) were recruited (Table 2).

A control group was recruited from the community (patient’s and or hospital staff’s families and from an aged Baptist Church group). They were submitted to an interview about medical, scholarship and functional history, besides current health. Individuals suspected of neurological or psychiatric diseases were not included. An initial group of 57 individuals was selected. Of these, three subjects were excluded because of depression and brain trauma in the preceding 3 years.

The remaining 66 INPH individuals and 54 control subjects underwent propensity score matching and both groups were homogenized according to age, gender, and educational level. This led to the exclusion of five individuals from the INPH group and 12 individuals from the control group. Data of 61 individuals from the INPH group and 42 from the control group were then compared (Fig. 2).

Procedures

Patients were admitted to the hospital for two to three days. Two lumbar punctures were performed at intervals of at least 24 h apart (minimum 24 h and maximum 28 h, to guarantee total LCR turnover) and 30 mL of CSF (M: 26 ± 1.2), was withdrawn in each, as described by Adams [1]. CSF opening pressure was measured and a CSF sample was sent for laboratory analysis. The first LP was preceded by cognitive and gait examinations performed by a team of three neuropsychologists and four neurologists, respectively. After being examined for excluding post-puncture pain or headache [9] and between two and six hours after each LP, INPH subjects were tested again for both gait and cognitive examinations. Gait and cognition were never assessed within an interval shorter than one hour after the LP (M = 4,4 ± 2,5). Neither persistent headache nor persistent pain after lumbar puncture were reported by individuals in this series, and further information regarding this issue can be found in a recent paper of our group [15].

To minimize interindividual bias, the neuropsychology team was constantly trained on study procedures. Furthermore, study data were discussed weekly and investigators remained blind to investigation subjects' diagnosis during study procedures. Gait protocol has been described elsewhere [16] and will not be discussed in this paper.

The neuropsychological assessment was performed during hospitalization. The average duration of the cognitive protocol was one hour and thirty minutes. Some data was not collected from patients with visual or motor impairment. The cognitive tests used were: Orientation (self-awareness, orientation in place, time and current context). Mental Control (usual and unusual activities: to speak the days of the week, the months of the year, to count (e.g.: 1–20) and after this, all the same activities are repeated in backward), FAR-COWA (The examinee must produce orally as many words as possible beginning with a specified letter during one minute (F-A-R) and thereafter he is asked to produce as many animal names as possible within a one-minute interval); Rey Auditory Verbal Learning Test (RAVLT): This is an episodic memory test: a 15-word list is read aloud (A list) for five consecutive trials (A1–A5). After trial A5, a distractor 15-word list is presented once and the examinee needs to learn it (B1); immediately after this, the examinee is asked to recall the A list (A6—immediate recall) and twenty minutes later again (A7—delayed recall); Rey Complex Figure Test (CFT): A visuospatial memory test: the examinee needs to learn a new complex figure by copying it. After a three-minute interval, he is asked to reproduce the figure as similar as possible without looking at it and thirty minutes later again); Digit Span Test (WMS-R): A two-step working memory test: 1—the examinee is asked to repeat series of numbers in the same order that they are presented by the examiner; the quantity of digits increases progressively (digit span forward); 2—the examinee is asked to repeat similar series but in backward (digit span backward). In our protocol both parts are scored separately; Stroop test (University of Victoria Version): This is an attention test: the examinee is presented to three cards. In the first card, there are 24 colored dots (blue, red, yellow, green). The subject is required to name the colors in which the stimuli are printed as fast as possible. In the second part, neutral words (when, where, what) are printed in the same colors of the dots and the subject needs to name the colors while ignoring the words. In the third part, the words blue, red, yellow and green are printed in conflicting colors (e.g.: the word “blue” printed in red color). Individuals are asked to name the colors printed again while avoid reading the words (name of colors). Scores are calculated based on the speed to perform the tasks and on the number of errors only in the third part; Mini Mental Status Examination (MMSE): This is a 30-point scoring test composed of “Introduction” section—items of orientation, memory and attention (21 points); “Materials” section—items of naming, follow verbal and written commands, writing a sentence and drawing a figure with intersections (9 points); Frontal Assessment Battery (FAB): This is a brief battery consisting of six subtests: similarities, lexical fluency, motor series, conflicting instructions, Go-No-Go test and prehension behavior). Each subtest can be scored from 0 to 3 points. The score is based on the sum of the six test scores.

To avoid a learning effect, some tests were applied in up to three different versions. The tests used in this study are among the most cited in INPH literature [17, 18].

Patients with IHPH and control subjects were matched using propensity scores (PS). Three variables were selected to estimate PS: age, sex and formal education. The propensity score was calculated using logistic regression and patients with IHPH were matched with controls using the nearest neighbor technique with a predefined caliper of 0.2 and considering the ratio 1:1. This matching procedure was performed using the MatchIt package for R.16.

A propensity score model was used to homogenize the INPH and the control group considering a logistic regression model and conditioning the variables of age, sex and formal education.

To analyze the effect of STT on cognitive test results of the INPH group in comparison with the control group, Z scores were calculated for each INPH patient regarding mean (M) and standard deviation (SD) of the control group. The Z scores of variables whose improvements were equivalent to smaller values, such as run time number of errors, variables dots, words and colors, and errors of the Stroop test, were multiplied by − 1 (inverted). Thus, in all variables, the highest values of the Z score corresponded to the participants’ best results.

Results of quantitative variables were described by means and standard deviations. Categorical variables were presented in frequencies and percentages. Comparison between the INPH and control groups, concerning categorical variables, was conducted using Fisher's Exact Test. Comparison of either two disease-defined groups or clinical factors, relative to quantitative variables, was conducted using either Student's t-test for independent samples or the non-parametric Mann–Whitney test. More than two groups were compared considering the nonparametric Kruskal–Wallis test. Normality condition was evaluated through the Kolmogorov–Smirnov test. Values of p < 0.05 indicate statistical significance. Data were analyzed using the IBM Corp. Released 2011. IBM SPSS Statistics for Windows, Version 20.0. Armonk, NY: IBM Corp.

After propensity score matching, from the 66 subjects with INPH and the 54 control subjects initially recruited, 61 INPH subjects and 42 control subjects remained for the statistical analysis. These groups were balanced regarding age (p = 0.056), sex (p = 0.231) and educational level (p = 0.549).

Statistical analysis was supervised by a professional statistician.