Abstract

Although idiopathic normal pressure hydrocephalus (iNPH) is considered a treatable dementia, there is still some controversy regarding the cognitive improvement in these patients. The main aims of this study were to analyze baseline cognitive status and to study the neuropsychological changes after surgical treatment in a sample of 185 consecutive iNPH patients. An additional aim was to identify the variables that influenced the cognitive outcome. Specific tests assessing memory, attention, visual scanning, executive functions (EFs), and motor speed were used before and 6 months after shunting. The cognitive domains most affected at baseline were memory, EFs, attention, and psychomotor speed. After shunting, significant differences in the group as a whole were found in all tests except Digits Forward and Trail Making Part B. However, less than 50% of patients showed a significant improvement when analyzed individually. Previous global cognitive status assessed by Mini-Mental State Examination baseline scores was the best predictor for the cognitive outcome.

Introduction

Idiopathic normal pressure hydrocephalus (iNPH) is a treatable cause of dementia, characterized by gait abnormalities, urinary and/or fecal control disturbances, and progressive dementia combined with ventricular enlargement (Adams, Fisher, Hakim, Ojemann, & Sweet, 1965; Chang, Agarwal, Williams, Rigamonti, & Hillis, 2006; Chaudhry et al., 2007; Devito et al., 2005; Gallia, Rigamonti, & Williams, 2006; Iddon et al., 1999; Kahlon, Sundbarg, & Rehncrona, 2002; Poca, Sahuquillo, Barba, Añez, & Arikan, 2004; Pujari et al., 2008; Raftopoulos et al., 1994; Savolainen, Hurskainen, Paljarvi, Alafuzoff, & Vapalahti, 2002; Vanneste, 2000). The symptoms of iNPH can be treated by implanting a cerebrospinal fluid (CSF) shunt. However, in clinical practice, the cognitive impairment of iNPH patients must be distinguished from other causes of dementia, such as Alzheimer's disease, Parkinson's disease, and vascular dementia, among others. Consequently, the management of patients with suspected iNPH requires accurate diagnosis and the identification of the target population likely to benefit from surgical treatment (Devito et al., 2005; Gustafson & Hagberg, 1978; Pujari et al., 2008).

Several complementary tools that contribute to an accurate diagnosis and improving the success rate of shunting have been described by many authors. One of these tools is neuropsychological assessment, which may help clinicians to distinguish iNPH from other causes of dementia, which can predominantly affect cortical functions such as aphasia, apraxia, or agnosia (Benejam, Poca, Junque, Solana, & Sahuquillo, 2008), and also to objectively evaluate the outcome of surgical treatment (Devito et al., 2005; Farace & Shaffrey, 2005). Unfortunately, there is still no consensus about the best neuropsychological battery for testing these patients.

Although iNPH is considered reversible, controversy still exists on whether improvement can be achieved after surgery, especially in terms of the reversibility of cognitive impairment (Chaudhry et al., 2007; Duinkerke, Williams, Rigamonti, & Hillis, 2004; Iddon et al., 1999; Poca, Mataro, et al., (2004)). Changes in cognitive performance vary widely in different series and range from no improvement to a high rate of improvement (Gleichgerrcht et al., 2009; Hellstrom et al., 2008; Katzen et al. 2011; Thomas et al., 2005). Several factors hamper the evaluation of cognitive improvement after shunting in patients with iNPH: the lack of standard criteria for post-surgical improvement, the presence of patients with distinct etiologies in the same sample (idiopathic vs. secondary NPH), the small sample size of most series, the use of general scales with insufficient sensitivity to discriminate among different types of dementia, the ceiling effect (Duinkerke et al., 2004; Iddon et al., 1999; Stambrook et al., 1988; Vanneste, 2000), and the potential learning effect associated with the testing and retesting of patients. According to Iddon and colleagues, scales of general cognition, such as the Mini-Mental State Examination (MMSE), are successfully used to assess cognitive improvement in iNPH patients. However, complete neuropsychological examinations have shown cognitive dysfunction more clearly, especially in executive functions (EFs), in iNPH patients after treatment (Iddon et al., 1999). Nevertheless, these conclusions were based on a small series of patients. The present study is focused on a large series of homogeneous patients studied with a wide neuropsychological battery before and after treatment.

The aim of the present study was to analyze baseline cognitive status and to study the neuropsychological changes after surgical treatment in a large cohort of 185 iNPH patients consecutively treated in our department. A secondary goal was to investigate whether some predefined clinical variables (age, sex, level of education, illness duration, cognition disturbance as the first symptom, general cognitive status assessed using the MMSE, and the presence of vascular risk factors) were independent predictors of the cognitive outcome after treatment.

Patients and Methods

Patient Sample and General Management Protocol

From November 1997 to July 2007, 224 patients with suspected iNPH syndrome were evaluated at the Department of Neurosurgery of Vall d'Hebron University Hospital. All patients presented an increase in ventricular size (Evans' index ≥0.30: maximum bifrontal distance of the lateral ventricles measured in an axial slice scan where the Monro foramens were visible divided by the maximum inner diameter of the skull taken from the same slice; Evans, 1942) evaluated in the computed tomography scan or magnetic resonance imaging and at least two of the following clinical symptoms unexplained by other neurological or non-neurological conditions: (a) gait disturbances, (b) cognitive dysfunction, and (c) urinary and/or fecal disturbances. The decision to implant a shunt was always based on continuous intracranial pressure (ICP) monitoring and/or CSF dynamics studies (Poca, Mataro, et al., 2004; Poca et al., 2005). Written informed consent was obtained from all patients or from relatives when patients' cognitive condition precluded them from understanding or signing the informed consent.

Of the 224 iNPH patients initially studied, five patients did not undergo a neuropsychological assessment due to their clinical condition. In four patients, a different neuropsychological protocol was used. Thirty iNPH patients were lost in the follow-up (5 patients died, 1 patient had a shunt malfunction, 2 patients worsened, making cognitive assessment impossible, and 22 did not attend the follow-up visits).

The resulting final cohort was formed by 185 patients with confirmed iNPH (111 men and 74 women, aged 73.96 ± 6.3, range: 50–85) in whom a neuropsychological assessment was performed at both before and 6 months after surgery (Fig. 1).

Fig. 1.

Algorithm of patient selection in this study. The final sample included 185 patients with iNPH in whom a neuropsychological (NPS) assessment was performed before and 6 months after shunting.

Fig. 1.

Algorithm of patient selection in this study. The final sample included 185 patients with iNPH in whom a neuropsychological (NPS) assessment was performed before and 6 months after shunting.

During the assessment of cognition, some patients did not perform all neuropsychological tests due to low education level, physical disability, not understanding the task, their clinical condition, or other factors that interfered with the assessment.

Our protocol for the study and management of patients with iNPH has previously been described in detail (Benejam et al., 2008; Poca, Sahuquillo, et al., 2004, 2005). The three cardinal clinical symptoms of the Hakim and Adams triad, gait, urinary, and/or fecal control and cognitive dysfunction, were graded according to the NPH scale (Sahuquillo et al., 1991; Table 1), which is used to evaluate the severity of clinical status. Clinical evaluation also included the analysis of several additional motor tasks, such as getting up from a standard height chair, getting up and down a 23-cm high step, maintaining equilibrium during tandem walking, and walking 5 m two times (to measure walking speed and step length), among others (Solana et al., 2010).

Table 1.

Normal Pressure Hydrocephalus Scale (Sahuquillo et al., 1991)

Gait Disturbances 
 5. Normal gait 
 4. Abnormal but stable gait 
 3. Independent walking is possible but unstable or the patient falls 
 2. Ambulation is possible with help 
 1. Patient is bedridden or unable to ambulate 
Cognitive Function 
 5. Cognitive disturbances are only found by specific assessment 
 4. Memory problems reported by patient or relatives 
 3. Severe memory problems with behavior disturbances 
 2. Severe dementia 
 1. Patient is vegetative 
Sphincter Disturbances 
 5. No objective or subjective sphincter dysfunction 
 4. Urinary urgency 
 3. Sporadic urinary incontinence 
 2. Continuous urinary incontinence 
 1. Urinary and fecal incontinence 
Gait Disturbances 
 5. Normal gait 
 4. Abnormal but stable gait 
 3. Independent walking is possible but unstable or the patient falls 
 2. Ambulation is possible with help 
 1. Patient is bedridden or unable to ambulate 
Cognitive Function 
 5. Cognitive disturbances are only found by specific assessment 
 4. Memory problems reported by patient or relatives 
 3. Severe memory problems with behavior disturbances 
 2. Severe dementia 
 1. Patient is vegetative 
Sphincter Disturbances 
 5. No objective or subjective sphincter dysfunction 
 4. Urinary urgency 
 3. Sporadic urinary incontinence 
 2. Continuous urinary incontinence 
 1. Urinary and fecal incontinence 

Note: The maximum score is 15 and the minimum score is 3.

The coexistence of cerebrovascular disorders was assessed using the cerebrovascular risk scale (CVR), which ranges from 0 to 5 according to whether patients had one or more of the following conditions: previous stroke (2 points), arterial hypertension (1 point), diabetes or other vasculopathy (1 point), and coronary disease (1 point).

The study protocol also included a battery of neuropsychological tests (Benejam et al., 2008; Poca, Mataro, et al., 2004, 2005), continuous ICP monitoring using an epidural sensor (Poca et al., 2004) and/or CSF dynamics studies (resistance to the outflow of CSF, Rout, calculated by the Katzman constant rate infusion test; Katzman & Hussey, 1970). Patients with B-waves (0.5–2 ICP waves/min, lasting for at least 10 min) in more than 10% of the total recording time and/or a Rout of >10 mmHg ml−1 min−1 were selected for shunting, regardless of mean ICP (Poca, Mataro, et al., (2004), 2005). All patients were treated with a differential low-pressure valve system combined with a gravitational device.

Cognitive Assessment Protocol

All patients included in the study underwent comprehensive neuropsychological assessment. The baseline evaluation was carried out during the stay in hospital and the follow-up evaluation at 6 months after surgery. The protocol of neuropsychological assessment for suspected iNPH has been described previously (Benejam et al., 2008). The battery of tests was selected for assessing aspects of cognition that are frequently affected in NPH patients, such as immediate and delayed memory, attention, EFs, motor speed, and manual coordination, and for screening and orientation tests (Table 2).

Table 2.

Clinical and neuropsychological protocol for the examination of patients with suspected NPH

Type of assessment Assessment tool 
Clinical assessment NPH scale (Sahuquillo et al., 1991
Motor Performance Test (MPT) 
Screening test MMSE (Folstein, Folstein, & McHugh, 1975
Information and Orientation (WMS-R) (Wechsler, 1945
Neuropsychological assessment Memory WMS-R Visual reproduction (Wechsler, 1945
Rey Auditory-Verbal Learning Tests (Lezak, 2004; Strauss, 2006) 
Executive function Verbal Fluency Test (Lezak, 2004; Strauss, 2006) 
Trail Making Test (TMT) B (Reitan, 1955
Digits Backwards (WMS-R) (Wechsler, 1945
Attention and psychomotor speed Digits Forwards (WMS-R) (Wechsler, 1945
Trail Making Test (TMT) A (Reitan, 1955
Purdue Pegboard Test (Lezak, 2004; Strauss, 2006) 
Type of assessment Assessment tool 
Clinical assessment NPH scale (Sahuquillo et al., 1991
Motor Performance Test (MPT) 
Screening test MMSE (Folstein, Folstein, & McHugh, 1975
Information and Orientation (WMS-R) (Wechsler, 1945
Neuropsychological assessment Memory WMS-R Visual reproduction (Wechsler, 1945
Rey Auditory-Verbal Learning Tests (Lezak, 2004; Strauss, 2006) 
Executive function Verbal Fluency Test (Lezak, 2004; Strauss, 2006) 
Trail Making Test (TMT) B (Reitan, 1955
Digits Backwards (WMS-R) (Wechsler, 1945
Attention and psychomotor speed Digits Forwards (WMS-R) (Wechsler, 1945
Trail Making Test (TMT) A (Reitan, 1955
Purdue Pegboard Test (Lezak, 2004; Strauss, 2006) 

Notes: MMSE = Mini-Mental State Examination; NPH scale = Normal Pressure Hydrocephalus Scale; WMS-R = Wechsler Memory Scale-Revised.

Due to the extended inclusion period, the neuropsychological protocol underwent some changes during the study period. Consequently, not all patients were evaluated using the same tests. Some tests, such as the MMSE and the use of the letter “s” instead of the three letters “f,” “a,” and “s” for the Phonetic Verbal Fluency tests, were included or modified during the inclusion period. A score of <24 for the MMSE was considered abnormal (Iddon et al., 1999).

Functional Status Assessment

The functional status of every patient was assessed using three rating scales that registered the patient's functioning in daily life activities:

  1. The “Rapid Disability Rating Scale-2” (Linn & Linn, 1982), which assesses the degree of disability. An overall score of 18 indicates that a person is totally independent and a score of 72 indicates total dependence.

  2. A modification of the “Stein and Langfitt Scale” (Borgesen, 1984), which includes six grades of dependence: 0 (no neurological deficit), the patient is able to work or perform the same duties as before the disease onset; 1 (minimal deficit), the patient is able to function independently at home; 2, the patient requires some supervision at home; 3, the patient requires custodial care despite considerable independent function; 4, the patient has no practical capacity for independent function; 5, the patient is bedridden or in a minimally conscious or vegetative state with no/minimal spontaneous activity or verbal contact.

  3. The “Everyday Activities scale” (Fillenbaum, 1985), which registers the patient's functional capacity. Patients were asked to rate on a three-point scale (0 = unable; 1 = with help; 2 = without help) how much help they needed to perform five daily life activities (mobility, shopping, cooking, household tasks, and money management). The minimum score was 0 point and the maximum 10.

Assessment of Improvement after Treatment

Outcome was independently assessed by the neurosurgeon in charge of the patient and by an independent neuropsychologist using the NPH scale both before and 6 months after shunting. If discrepancies were found between the neurosurgeon and the neuropsychologist, the patient was re-evaluated and the final score was reached by consensus. Neuropsychological tests, functional scales, and MPT were administered by the neuropsychologist, whereas patients were in hospital for presurgical evaluation and again 6 months after surgery.

For assessing changes in the clinical status, and because a small change in the NPH scale score represents a substantial change in the patient's functional status, we defined moderate improvement as a one point increase and marked improvement as an increase in at least two points in this scale (Poca, Mataro, et al., 2004, 2005).

When considering the group as a whole, cognitive changes before and after surgical treatment were analyzed using non-parametric tests. For individual analysis, a change in cognitive and behavioral features was considered to be clinically significant when cognitive performance increased by 1 SD, adjusted by the patient's age, sex, and educational level, when compared with the baseline score, or by an increase of at least 20 percentile points from the baseline score (Trial Making Test [TMT] Parts A and B and visual memory of the Wechsler Memory Scale-Revised, WMS-R; Duinkerke, Williams, Rigamonti, & Hillis, 2004; Thomas et al., 2005). All scores were corrected using normative data. We also defined overall cognitive improvement after shunting as a minimum increase of 4 points in the MMSE (Thomas et al., 2005).

Statistical Analysis

All descriptive statistics were analyzed by using the SPSS package for Windows (Version 15, SPSS, Chicago, IL, USA). The assumption that data were normally distributed was tested using the Kolmogorov–Smirnov test. In normally distributed data, the mean ± 1 SD was used to summarize the variables. In skewed samples, the median and the interquartile range (IQR), minimum (min), and maximum (max) values were used. Complete data sets, including outliers, were used for the analyses.

Cognitive changes produced after surgical treatment were evaluated by paired samples t-test for normally distributed variables and the Wilcoxon Signed-Rank Test for non-parametric data. Logistic regression models were performed in order to identify demographic data and clinical variables that better predicted cognitive improvement after treatment, including age, sex, level of education, illness duration, baseline cognitive status, cognitive disturbance as a first symptom, and CVR factors. These procedures were analyzed from raw scores. A p-value of <0.05 was considered statistically significant.

Results

Continuous ICP monitoring was performed in 178 patients. The median ICP was 6 mmHg (IQR: 8.1 [3.94–12], min: −5, max: 47). The mean percentage of B-waves was 45.7 ± 23.8%. A Rout of 11.5 mmHg ml−1min−1 was observed in the only patient with B-waves registering at less than 10%. Demographic data, clinical characteristics, ICP values, and brain ventricular size of patients included in the study are summarized in Table 3.

Table 3.

Demographic, clinical, and radiological data of iNPH patients

Demographic data 
 Age (mean ± SD [range]) 73.96 ± 6.3 (50–85) 
 Sex (number [%]) 
  Men 111 (60) 
  Women 74 (40) 
 Education (number [%]) 
  Illiterate 8 (4.3) 
  Not finished basic education 80 (43.2) 
  Finished basic education 62 (33.5) 
  Secondary school education 17 (9.2) 
  University education 16 (8.6) 
  Unknown 2 (1.1) 
 Cerebrovascular risk (CVR) scale 1 (0–2) min: 0, max: 5a 
Clinical and radiological data 
 First symptom (number [%]) 
  Gait disturbance 104 (56.2) 
  Cognitive impairment 38 (21.1) 
  Sphincter incontinence 9 (4.9) 
  Other 28 (15.1) 
  Unknown 5 (2.7) 
 Duration of symptoms (months) 24 (12–36) min: 0, max: 24a 
 ICP (mmHg) 6 (3.9–12) min: 0, max: 24a 
 Percentage of B-waves (mean ± SD [range]) 45.74 ± 23.88 (0–100) 
Routb (mean ± SD [range]) 16.49 ± 6.57 (3.03–41.20) 
 Ventricular Index (mean ± SD [range]) 0.35 ± 0.04 (0.21–0.44)c 
Demographic data 
 Age (mean ± SD [range]) 73.96 ± 6.3 (50–85) 
 Sex (number [%]) 
  Men 111 (60) 
  Women 74 (40) 
 Education (number [%]) 
  Illiterate 8 (4.3) 
  Not finished basic education 80 (43.2) 
  Finished basic education 62 (33.5) 
  Secondary school education 17 (9.2) 
  University education 16 (8.6) 
  Unknown 2 (1.1) 
 Cerebrovascular risk (CVR) scale 1 (0–2) min: 0, max: 5a 
Clinical and radiological data 
 First symptom (number [%]) 
  Gait disturbance 104 (56.2) 
  Cognitive impairment 38 (21.1) 
  Sphincter incontinence 9 (4.9) 
  Other 28 (15.1) 
  Unknown 5 (2.7) 
 Duration of symptoms (months) 24 (12–36) min: 0, max: 24a 
 ICP (mmHg) 6 (3.9–12) min: 0, max: 24a 
 Percentage of B-waves (mean ± SD [range]) 45.74 ± 23.88 (0–100) 
Routb (mean ± SD [range]) 16.49 ± 6.57 (3.03–41.20) 
 Ventricular Index (mean ± SD [range]) 0.35 ± 0.04 (0.21–0.44)c 

Note: ICP = intracranial pressure.

aMedian and interquartile range (IQR) and the minimum and maximum.

bRout resistance to absorption LCR.

cEvans' index calculated by magnetic resonance. The value is underrated due to the gantry position.

Clinical Status before Treatment

At baseline, the median score for the NPH scale was 9 (IQR: 4 [7–11], min: 4, max: 14). The complete clinical triad (patients with a score of 4 or less in any of the NPH scale components) was observed in 155 of the 185 patients studied (83.8%). The most frequent combination of symptoms in patients with an incomplete triad was gait dysfunction and cognitive impairment in 17 (56.7%) patients, followed by gait dysfunction and urinary incontinence in 5 (16.6%) patients.

The most frequent initial clinical symptom was gait impairment, found in 104 (56.2%) patients, followed by loss of memory in 39 (21.1%) patients. Additional motor disturbances were found in practically all patients, with a mean MPT score of 6.04 ± 4.74 (range: 0–14), a mean step length of 24.8 ± 18 cm (range: 0–62.5), and a mean walking speed of 0.41 ± 0.33 m/s (range: 0–1.25; Table 4).

Table 4.

Cognitive and motor results at baseline and 6 months after shunting

 Before treatment
 
6 months after treatment
 
  
Test Number of cases Median (IQR) Range Number of cases Median (IQR) Range Z p-value 
Functional scales 
 RDRS-2 182 33 (26–42) 19–57 181 26 (22–34) 18–54 −8.212a <.001 
 Everyday Activities scale 171 4 (1–8) 0–10 171 7 (3–10) 0–10 −7.254a <.001 
 Stein and Langfitt-modified 182 3 (2–4) 0–4 181 2 (1–3) 0–4 −7.404a <.001 
Screening test 
 MMSE 144 24 (9–26) 6–30 149 25 (19–27) 6–30 −5.010a <.001 
 Information and Orientation 181 12 (9–13) 0–14 185 13 (11–14) 0–14 −5.723a <.001 
Neuropsychological test 
 Visual Reproduction (immediate recall) 157 14.01 ± 9.41b 0–38 159 12.72 ± 9.05b 0–36 −6.205b <.001 
 Visual Reproduction (delayed recall) 157 1 (0–9) 0–34 158 4 (0–11) 0–29 −4.003a <.001 
 RAVLT (total words) 173 19.18 ± 8.63b 0–42 178 22.25 ± 9.99b 0–51 −5.507c <.001 
 RAVLT (delayed recall) 173 1 (0–3) 0–14 178 1 (0–4) 0–12 −4.919a <.001 
 Digits Forwards 180 4 (4–5) 0–6 182 4 (4–5) 0–7 −0.820a .412 
 Digits Backwards 180 3 (2–3) 0–5 181 3 (2–3) 0–5 −3.076a .002 
 Phonetic Verbal Fluency 162 2 (0–5.25) 0–17 163 4 (1–6) 0–17 −3.823a <.001 
 Semantic Verbal Fluency 183 8.34 ± 4.86b 0–24 185 9.83 ± 4.59b 0–24 −5.320a <.001 
 Trail Making Test Part A (s)d 142 156 (85–262.25) 32–840 148 112 (74–186.75) 36–732 −5.364a <.001 
 Trail Making Test Part B (s)d 70 333.44 ± 149.51b 107–853 83 329.28 ± 153.16b 96–983 0.51c .960 
 Purdue Pegboard Test 
  Dominant hand 164 6.75 ± 3.24b 0–15 168 7.95 ± 2.96b 0–15 −6.370c <.001 
  Non-dominant hand 169 6 ± 3.24b 0–13 163 8 (5–9) 0–15 −7.494a <.001 
Motor 
 MPT 146 6.04 ± 4.74b 0–14 150 8.44 ± 4.34b 0–14 −8.897c <.001 
  Steps length (cm) 166 24.8 ± 18b 0–62.5 165 34.74 ± 16.73b 0–76.92 −9.352c <.001 
  Speed velocity (m/s) 166 0.41 ± 0.33b 0–1.25 165 0.62 ± 0.33b 0–1.4 −9.939c <.001 
 Before treatment
 
6 months after treatment
 
  
Test Number of cases Median (IQR) Range Number of cases Median (IQR) Range Z p-value 
Functional scales 
 RDRS-2 182 33 (26–42) 19–57 181 26 (22–34) 18–54 −8.212a <.001 
 Everyday Activities scale 171 4 (1–8) 0–10 171 7 (3–10) 0–10 −7.254a <.001 
 Stein and Langfitt-modified 182 3 (2–4) 0–4 181 2 (1–3) 0–4 −7.404a <.001 
Screening test 
 MMSE 144 24 (9–26) 6–30 149 25 (19–27) 6–30 −5.010a <.001 
 Information and Orientation 181 12 (9–13) 0–14 185 13 (11–14) 0–14 −5.723a <.001 
Neuropsychological test 
 Visual Reproduction (immediate recall) 157 14.01 ± 9.41b 0–38 159 12.72 ± 9.05b 0–36 −6.205b <.001 
 Visual Reproduction (delayed recall) 157 1 (0–9) 0–34 158 4 (0–11) 0–29 −4.003a <.001 
 RAVLT (total words) 173 19.18 ± 8.63b 0–42 178 22.25 ± 9.99b 0–51 −5.507c <.001 
 RAVLT (delayed recall) 173 1 (0–3) 0–14 178 1 (0–4) 0–12 −4.919a <.001 
 Digits Forwards 180 4 (4–5) 0–6 182 4 (4–5) 0–7 −0.820a .412 
 Digits Backwards 180 3 (2–3) 0–5 181 3 (2–3) 0–5 −3.076a .002 
 Phonetic Verbal Fluency 162 2 (0–5.25) 0–17 163 4 (1–6) 0–17 −3.823a <.001 
 Semantic Verbal Fluency 183 8.34 ± 4.86b 0–24 185 9.83 ± 4.59b 0–24 −5.320a <.001 
 Trail Making Test Part A (s)d 142 156 (85–262.25) 32–840 148 112 (74–186.75) 36–732 −5.364a <.001 
 Trail Making Test Part B (s)d 70 333.44 ± 149.51b 107–853 83 329.28 ± 153.16b 96–983 0.51c .960 
 Purdue Pegboard Test 
  Dominant hand 164 6.75 ± 3.24b 0–15 168 7.95 ± 2.96b 0–15 −6.370c <.001 
  Non-dominant hand 169 6 ± 3.24b 0–13 163 8 (5–9) 0–15 −7.494a <.001 
Motor 
 MPT 146 6.04 ± 4.74b 0–14 150 8.44 ± 4.34b 0–14 −8.897c <.001 
  Steps length (cm) 166 24.8 ± 18b 0–62.5 165 34.74 ± 16.73b 0–76.92 −9.352c <.001 
  Speed velocity (m/s) 166 0.41 ± 0.33b 0–1.25 165 0.62 ± 0.33b 0–1.4 −9.939c <.001 

Notes: RDRS-2 = Rapid Disability Rating Scale-2; MMSE = Mini-Mental State Examination; RAVLT = Rey Auditory-Verbal Learning Test; MPT = Motor Performance Test; IQR = interquartile range.

Statistical analysis was performed with raw scores.

az-values by the Wilcoxon Signed-Rank Test.

bMean ± SD.

ct-Student.

dAfter treatment, only 7 of the 142 patients in TMT-A and in 2 of 83 patients in TMT-B completed the test in the required time.

Regarding dependency, 66 (36.9%) patients were considered independent for the activities of daily living (Grades 0, 1, and 2), 37 (20.7%) patients required some help (Grade 3), and 76 (42.5%) patients were dependent (Grades 4 and 5) in the modified Stein and Langfitt scale. Patients also presented low scores in the level of autonomy in daily life activities analyzed by the Rapid Disability Rating Scale-2 (RDRS-2) scale (median score: 33, IQR: 16 [26–42], min: 19, max: 57) and by the Everyday Activities scale (median score: 4, IQR: 7 [1–8], min: 0, max; 10) (Table 4).

When we divided the whole group into two subgroups based on whether or not the first symptom was cognitive impairment, we found that patients whose clinical symptoms began with cognitive impairment had a significantly lower performance at baseline in memory (RAVLT total words, RAVLT and Visual Memory delayed recall), general cognition (MMSE score and Information and Orientation test), and the Semantic Verbal Fluency test when they were compared with the remaining patients (Table 5).

Table 5.

Cognitive differences at baseline between iNPH group with first symptom was cognitive impairment and iNPH group with other symptom as a first symptom

 iNPH patients with first symptom were cognitive impairment
 
iNPH patients with other first symptom
 
  
Test Number of cases Median (IQR) Range Number of cases Median (IQR) Range z p-value 
Screening Test 
 MMSE 32 18 (14–24.5) 6–28 108 24.5 (18–27) 8–30 −3.348a .001 
 Information and Orientation 38 9 (7–12.25) 1–14 139 12 (10–14) 0–14 −3.512a <.001 
Neuropsychological Test 
 Visual Reproduction (Immediate recall) 32 12.69 + 9.59b 0–38 121 14.58 + 9.36 0–36 1.011c .314 
 Visual Reproduction (delayed recall) 32 0 (0–4.75) 0–34 121 2 (0–9) 0–23 −2.029a .042 
 RAVLT (total words) 37 15.68 + 8.3b 0–36 132 20.25 + 8.6b 0–42 2.886c .004 
 RAVLT (delayed recall) 37 0 (0–2) 0–4 132 1 (0–3.75) 0–14 −2.418a .016 
 Digits Forwards 38 4 (3–4) 0–6 137 4 (4–5) 0–6 −1.846a .065 
 Digits Backwards 38 2 (2–3) 0–4 137 3 (2–3) 0–5 −1.309a .191 
 Phonetic Verbal Fluency 33 3 (0–5) 0–11 124 2 (0–6) 0–17 −0.406a .685 
 Semantic Verbal Fluency 38 6.61 + 4.79b 0–22 140 8.88 + 4.84b 0–24 2.576c .010 
 Trail Making Test Part A (s) 27 197 (80–320) 32–459 112 144.5 (85–255) 37–840 −0.996a .319 
 Trail Making Test Part B (s) 249.75 + 116.8b 110–420 61 343 + 151.53b 107–853 1.672c .099 
 Purdue Pegboard Test         
  Dominant hand 33 6.39 + 3.5b 0–13 127 6.88 + 3.22b 0–15 0.762c .447 
  Non-dominant hand 34 5.44 + 3.73b 0–13 121 6.18 + 3.11b 0–12 1.172c .243 
 iNPH patients with first symptom were cognitive impairment
 
iNPH patients with other first symptom
 
  
Test Number of cases Median (IQR) Range Number of cases Median (IQR) Range z p-value 
Screening Test 
 MMSE 32 18 (14–24.5) 6–28 108 24.5 (18–27) 8–30 −3.348a .001 
 Information and Orientation 38 9 (7–12.25) 1–14 139 12 (10–14) 0–14 −3.512a <.001 
Neuropsychological Test 
 Visual Reproduction (Immediate recall) 32 12.69 + 9.59b 0–38 121 14.58 + 9.36 0–36 1.011c .314 
 Visual Reproduction (delayed recall) 32 0 (0–4.75) 0–34 121 2 (0–9) 0–23 −2.029a .042 
 RAVLT (total words) 37 15.68 + 8.3b 0–36 132 20.25 + 8.6b 0–42 2.886c .004 
 RAVLT (delayed recall) 37 0 (0–2) 0–4 132 1 (0–3.75) 0–14 −2.418a .016 
 Digits Forwards 38 4 (3–4) 0–6 137 4 (4–5) 0–6 −1.846a .065 
 Digits Backwards 38 2 (2–3) 0–4 137 3 (2–3) 0–5 −1.309a .191 
 Phonetic Verbal Fluency 33 3 (0–5) 0–11 124 2 (0–6) 0–17 −0.406a .685 
 Semantic Verbal Fluency 38 6.61 + 4.79b 0–22 140 8.88 + 4.84b 0–24 2.576c .010 
 Trail Making Test Part A (s) 27 197 (80–320) 32–459 112 144.5 (85–255) 37–840 −0.996a .319 
 Trail Making Test Part B (s) 249.75 + 116.8b 110–420 61 343 + 151.53b 107–853 1.672c .099 
 Purdue Pegboard Test         
  Dominant hand 33 6.39 + 3.5b 0–13 127 6.88 + 3.22b 0–15 0.762c .447 
  Non-dominant hand 34 5.44 + 3.73b 0–13 121 6.18 + 3.11b 0–12 1.172c .243 

Notes: MMSE = Mini-Mental State Examination; RAVLT = Rey Auditory-Verbal Learning Test; IQR = interquartile range. Statistical analysis was performed with raw scores.

az-values by the Wilcoxon Signed-Rank Test.

bMean ± SD.ct-Student.

Cognitive Status before Treatment

For general cognition, 74 of 144 (51.4%) patients showed a normal score in MMSE (scoring ≥24; mean: 16.6 ± 4 [range: 6–23]) and 70 of 144 (48.6%) patients showed an impaired score in MMSE (scoring <24; mean: 26.4 ± 1.7 [range: 24–30]). For orientation in time and place, 106 of 183 (57.9%) patients evaluated by the Information and Orientation subtests showed scores below the normal range.

In the cognitive analysis at baseline (Table 6), patients showed low performance, defined as a score of 1.5 SD or less below the cutoff score, in nearly all tests administered, especially in those sensitive to EFs, memory, and psychomotor speed (Fig. 2). This low performance was more evident in Phonetic Verbal Fluency tests and in both parts of the Trail Making Tests. Only one patient was able to complete Part B (EF) within the time required, and only five patients correctly performed Part A (psychomotor speed). In the remaining tests, patients showed better performances, especially in the Semantic Verbal Fluency test and the Digits Backwards test (Table 6). For memory tests, the patients in our cohort showed lower performance in visual memory compared with verbal memory, especially in delayed reproduction. The only test that was not affected before shunting was the Digits Forward (Table 6).

Table 6.

Individual analysis of cognition at baseline and its change after shunting

Neuropsychological tests N Number of patients with altered scores at baseline (−1.5 SDNumber of patients with altered scores at follow-up (−1.5 SDSignificant improvement* 
Screening Tests 
 Information and Orientation 183 106 (57.9%) 79 (43.2%) 62 (33.9%) 
Memory Tests 
 Immediate Visual memory 177 149 (84.2%) 140 (79.1%) 7 (4.6%) 
 Delayed Visual Memory 177 164 (92.7%) 157 (88.7%) 2 (1.4%) 
 RAVLT (total words) 181 136 (75.1%) 108 (59.7%) 46 (26.1%) 
 RAVLT (delayed recall) 181 127 (70.2%) 107 (59.1%) 42 (23.9%) 
Executive Functions 
 Digits Backwards 184 92 (50%) 67 (36.4%) 21 (11.7%) 
 Phonetic Verbal Fluency 165 137 (83%) 131 (79.4%) 15 (9.2%) 
 Semantic Verbal Fluency 185 97 (52.4%) 65 (35.1%) 50 (27%) 
 Trail Making Part B 152 151 (99.3%) 151 (99.3%) 37 (43.8%) 
Attention and Psychomotor Speed 
 Digits Forwards 184 37 (20.1%) 36 (19.6%) 13 (7.2%) 
 Trail Making Part A 173 168 (97.1%) 164 (94.8%) 77 (52.7%) 
 Purdue Pegboard (dominant hand) 174 126 (72.4%) 113 (64.9%) 72 (43.9%) 
 Purdue Pegboard (non-dominant hand) 170 134 (78.8%) 115 (67.7%) 78 (49.1%) 
Neuropsychological tests N Number of patients with altered scores at baseline (−1.5 SDNumber of patients with altered scores at follow-up (−1.5 SDSignificant improvement* 
Screening Tests 
 Information and Orientation 183 106 (57.9%) 79 (43.2%) 62 (33.9%) 
Memory Tests 
 Immediate Visual memory 177 149 (84.2%) 140 (79.1%) 7 (4.6%) 
 Delayed Visual Memory 177 164 (92.7%) 157 (88.7%) 2 (1.4%) 
 RAVLT (total words) 181 136 (75.1%) 108 (59.7%) 46 (26.1%) 
 RAVLT (delayed recall) 181 127 (70.2%) 107 (59.1%) 42 (23.9%) 
Executive Functions 
 Digits Backwards 184 92 (50%) 67 (36.4%) 21 (11.7%) 
 Phonetic Verbal Fluency 165 137 (83%) 131 (79.4%) 15 (9.2%) 
 Semantic Verbal Fluency 185 97 (52.4%) 65 (35.1%) 50 (27%) 
 Trail Making Part B 152 151 (99.3%) 151 (99.3%) 37 (43.8%) 
Attention and Psychomotor Speed 
 Digits Forwards 184 37 (20.1%) 36 (19.6%) 13 (7.2%) 
 Trail Making Part A 173 168 (97.1%) 164 (94.8%) 77 (52.7%) 
 Purdue Pegboard (dominant hand) 174 126 (72.4%) 113 (64.9%) 72 (43.9%) 
 Purdue Pegboard (non-dominant hand) 170 134 (78.8%) 115 (67.7%) 78 (49.1%) 

Notes: RAVLT = Rey Auditory-Verbal Learning Test; N = number; SD = standard deviation.

*Significant improvement defined as ≥1 SD.

Fig. 2.

Bar chart with the percentage of patients with below normal scores at baseline for the different tests used to evaluate the four cognitive functions assessed in these patients (general screening tests for dementia, memory tests, EFs, and attention and psychomotor speed tests). NDH = non-dominant hand; DH = dominant hand.

Fig. 2.

Bar chart with the percentage of patients with below normal scores at baseline for the different tests used to evaluate the four cognitive functions assessed in these patients (general screening tests for dementia, memory tests, EFs, and attention and psychomotor speed tests). NDH = non-dominant hand; DH = dominant hand.

Clinical Improvement after Treatment

After shunting, 170 patients (91.9%) showed an improvement according to the NPH scale. Moderate improvement was found in 24 patients (12.9%) and marked improvement in 146 patients (79.9%; median change score: 4, IQR: 4 [2–6], min: 1; max: 11). No change was observed in 11 patients (5.9%) and neurological worsening was seen in four patients (2.2%). After excluding patients with normal function in each domain in the presurgical evaluation who were thus not susceptible to improvement according to the NPH scale, we found that urinary sphincter disturbances improved in 84.4% of patients, followed by gait (83.9%) and cognition (64.4%).

Six months after shunting, 112 of the 179 patients (62.6%) were independent for the activities of daily living (Grades 0, 1, and 2 of the modified Stein and Langfitt Scale), 25 (13.9%) were partially dependent, and 42 (23.5%) remained dependent (Fig. 3). When evaluating the scores obtained with the Everyday Activities scale (n = 169), improvement was found in 102 patients (60%). Of these, an improvement >2 points was found in 57 patients. The level of autonomy of patients also improved significantly when analyzed by the RDRS-2 (Table 4).

Fig. 3.

Percentage of patients in each category of the Stein and Langfitt Modified Scale at baseline and at follow-up.

Fig. 3.

Percentage of patients in each category of the Stein and Langfitt Modified Scale at baseline and at follow-up.

Cognitive Changes after Treatment

Before treatment, 50% of patients showed a normal MMSE score (≥24 points). After shunting, significant improvement (≥4 points) was found in 37 patients, which increased the percentage of patients with a normal score in this test to 60.4%. Table 4 summarizes the results of the neuropsychological tests performed at baseline and 6 months after surgery for the entire cohort. Statistically significant differences were found between the two assessments in almost all tests, except for the Digits Forward and Part B of the Trail Making Test. In the latter, the time necessary to complete the task was shorter after shunting, but the difference was not statistically significant.

In the individual analysis of cognitive improvement, the patient's competence after shunting increased in nearly all-cognitive tests, except for the Digits Forward (Fig. 4). The cognitive abilities that showed greater improvement were psychomotor speed and verbal memory, with a 26.1% change at follow-up for immediate memory and 23.9% for delayed memory when compared with baseline values. Figure 5 shows an example of visual memory assessed using the Immediate Recall subtest from the WMS-R in a patient with iNPH before and 6 months after treatment.

Fig. 4.

Bar chart with the percentages of the distribution of cognitive performance after shunting for the different tests used to evaluate the four cognitive functions assessed in these patients (general screening tests for dementia, memory tests, EFs, and attention and psychomotor speed tests). NDH = non-dominant hand; DH = dominant hand. Improvement was defined as an increase in baseline scores with respect to baseline. The bars also show the percentages of patients with no changes after treatment as well as those who could not be tested.

Fig. 4.

Bar chart with the percentages of the distribution of cognitive performance after shunting for the different tests used to evaluate the four cognitive functions assessed in these patients (general screening tests for dementia, memory tests, EFs, and attention and psychomotor speed tests). NDH = non-dominant hand; DH = dominant hand. Improvement was defined as an increase in baseline scores with respect to baseline. The bars also show the percentages of patients with no changes after treatment as well as those who could not be tested.

Predictors of Cognition after Surgery

In the logistic regression analysis, we found that global cognitive status defined by the baseline MMSE score was the best predictor of cognitive function after surgery. Higher values of baseline MMSE were related to improvement in RAVLT (delayed recall) and TMT Part B, and lower values of baseline MMSE were associated with an improvement in Information and Orientation, Digits Backwards, Semantic Verbal Fluency, and Purdue Pegboard (dominant hand). Shorter illness duration was independently associated with an improvement in MMSE and Purdue Pegboard (non-dominant hand) and women showed a better improvement in RAVLT (total words). Table 7 summarizes the variables that showed statistically significant associations with cognitive results after shunting. In some tests, no variables predicting performance at follow-up were found.

Table 7.

Significant influences between demographic and clinical variables and cognitive improvement after shunting

Variable OR Confidence interval (95%) p-value 
MMSE 
 Illness duration 0.948 0.913–0.984 .005 
 MMSE baseline 0.710 0.629–0.801 <.001 
Information and Orientation (WMS-R) 
 MMSE baseline 0.833 0.772–0.899 <.001 
Rey Auditory-Verbal Learning Test (total words) 
 Sex (men) 0.444 0.201 –0.979 .044 
Rey Auditory-Verbal Learning Test (delayed recall) 
 MMSE baseline 1.097 1.012–1.188 .024 
Digits Backwards 
 MMSE baseline 0.881 0.800–0.969 .009 
Semantic Verbal Fluency 
 Level of education 1.513 1.035–2.212 .033 
 MMSE baseline 0.920 0.860–0.985 .017 
Trail Making Test Part B 
 MMSE baseline 1.123 1.025–1.232 .013 
Purdue Pegboard (dominant hand) 
 MMSE baseline 0.924 0.869–0.983 .012 
Purdue Pegboard (non-dominant hand) 
 Illness duration 0.979 0.960–0.998 .034 
Variable OR Confidence interval (95%) p-value 
MMSE 
 Illness duration 0.948 0.913–0.984 .005 
 MMSE baseline 0.710 0.629–0.801 <.001 
Information and Orientation (WMS-R) 
 MMSE baseline 0.833 0.772–0.899 <.001 
Rey Auditory-Verbal Learning Test (total words) 
 Sex (men) 0.444 0.201 –0.979 .044 
Rey Auditory-Verbal Learning Test (delayed recall) 
 MMSE baseline 1.097 1.012–1.188 .024 
Digits Backwards 
 MMSE baseline 0.881 0.800–0.969 .009 
Semantic Verbal Fluency 
 Level of education 1.513 1.035–2.212 .033 
 MMSE baseline 0.920 0.860–0.985 .017 
Trail Making Test Part B 
 MMSE baseline 1.123 1.025–1.232 .013 
Purdue Pegboard (dominant hand) 
 MMSE baseline 0.924 0.869–0.983 .012 
Purdue Pegboard (non-dominant hand) 
 Illness duration 0.979 0.960–0.998 .034 

Notes: In some neuropsychological tests, we found that variable influence did not enter in the regression. MMSE = Mini-Mental State of Examination; WMS-R = Wechsler Memory Scale-Revised; CVR = cerebrovascular risk; OR = odds ratio.

Discussion

The present study constitutes the largest published cohort to be analyzed in terms of the cognitive status of patients with iNPH and the cognitive changes produced after surgical treatment. After studying 185 iNPH patients, we can conclude that more than 50% of patients presented with impairment in memory, EFs, attention, and psychomotor speed before treatment. In some tests used, such as TMT Parts A and B, the percentages of abnormal scores reached nearly 100% (Fig. 2). Although cognitive scores remained below normal values in many patients, cognitive improvement after treatment was observed in these patients, especially regarding psychomotor speed, verbal memory, and general cognition (Table 6). This improvement may increase the quality of life for these patients and their caregivers.

Our results are consistent with those found by other authors, who conclude that cognitive alterations in iNPH patients are manifested by a learning and memory dysfunction, attention and concentration deficit, slowing in psychomotor speed, and difficulties in complex information processing and EFs (Benejam et al., 2008; Devito et al., 2005; Duinkerke et al., 2004; Hellstrom et al., 2008; Savolainen et al., 2002; Thomas et al., 2005). They may also show alterations in visuospatial and visuoconstructional skills (Benejam, et al., 2008; Chaudhry, et al., 2007; Devito et al., 2005). Regarding EFs, we found that our patients did not show significant differences at follow-up, although they increased their efficiency in completing the tests. In the literature, we found that some authors give support to our findings whereas others found different conclusions. Iddon and colleagues (1999) concluded that patients showed impairment in cognitive domains after surgery, especially in EF, even in patients without dementia, defined by a MMSE <24. Thomas and colleagues corroborates these findings; they concluded that frontal lobe and frontostriatal pathways are affected early in iNPH patients, resulting in irreversible damage (Iddon et al., 1999; Thomas et al., 2005). In contrast, Gleichgerrcht and colleagues (2009) found an improvement in EF after shunting.

This cognitive alteration pattern has traditionally been classified as “subcortical dementia.” Because of the repercussion of ventriculomegaly in the blood flow (Mataro et al., 2003) and in the rich neural circuitry connecting subcortical structures and cortico-subcortical pathways of the frontal lobes that affect complex functions, some authors considered this condition to be “fronto-subcortical dementia” (Devito et al., 2005; Hellstrom et al., 2008; Iddon et al., 1999; Thomas et al., 2005). However, this cognitive profile may be present in other types of dementia, and therefore, it is not specific for iNPH.

How to Define Cognitive Improvement after Treatment and the Problem of the Learning Effect

One of the most important difficulties in determining cognitive changes in NPH patients is the lack of consensus for defining the improvement after treatment (Hellstrom et al., 2008). There are different methods used to define significant changes in cognitive domains. Some authors use the ratio between the post-operative result and the sum of the pre- and post-operative results [post/(pre + post)] (Raftopoulos et al., 1994). Other authors use the percentage of change [(control – basal)/basal × 100] (Poca, Mataro, et al., 2004). Additional criteria frequently used are an increase in four points in the MMSE (Duinkerke et al., 2004; Hellstrom et al., 2008; Thomas et al., 2005) or an improvement rate greater than 25% in more than 50% of applied tests (Chang et al., 2006; Duinkerke et al., 2004; Thomas et al., 2005; Wikkelso, Andersson, Blomstrand, Lindqvist, & Svendsen, 1986). One of the most frequent methods to determine improvement after treatment, however, is to find an increase equal to or more than 1SD respective to baseline (Duinkerke et al., 2004; Thomas et al., 2005). In the present study, we used this criterion, we considered differences in the various cognitive domains before and after treatment to be statistically significant when patients showed an increase of 1SD, adjusted by the patient's age, sex, and educational level when compared with baseline score, or, in some tests, an increase of at least 20 percentile points from baseline score (Duinkerke et al., 2004; Thomas et al., 2005). More sophisticated methods, such as the generalized least-squares regression method (GLS regression) with random effects was used by our group to analyze improvement discarding the influence of learning effects during four consecutive days of assessment (Solana et al., 2010).

Regarding the learning effect, in a previous study, we demonstrated that in contrast with healthy volunteers in whom most of the neuropsychological tests analyzed showed the presence of learning effect, it was absent in NPH patients regardless of MMSE scores (Solana et al., 2010). In this study, we used a battery of tests that evaluated attention and visual scanning (Toulouse-Pieron Test, TMT A), motor speed and manual coordination (Grooved Pegboard Test and TMT A), EFs (Word Fluency Test), and immediate recall visual memory (Bingley's Memory Test), which was applied over four consecutive days (Solana et al., 2010; Fig. 5).

Fig. 5.

Example of visual memory in a patient with iNPH before and 6 months after treatment who was evaluated using the Immediate Recall subtest from the WMS-R: (A) template, (B) immediate figure reproduction before shunting, and (C) 6 months after shunting.

Fig. 5.

Example of visual memory in a patient with iNPH before and 6 months after treatment who was evaluated using the Immediate Recall subtest from the WMS-R: (A) template, (B) immediate figure reproduction before shunting, and (C) 6 months after shunting.

Does Cognitive Status Improve after Shunting?

Clinical changes were recorded 6 months after shunting to avoid the influence of comorbidity on outcome due to the age of these patients. This time period is also selected by other authors. Kahlon, Sjunnesson, and Rehncrona (2007) evaluated the outcome of patients with suspected NPH at 6 months and 5 years after shunt surgery and showed that these patients benefited from shunt surgery for at least 5 years. Furthermore, other authors have reported that iNPH patients may benefit from shunting in the long term (Mirzayan, Luetjens, Borremans, Regel, & Krauss, 2010). In patients with initial improvement, subsequent worsening requires testing of the implanted shunt to detect any malfunction (Poca, Sahuquillo, & Mataró, 2001).

In some studies, significant improvements in cognitive function after shunting, particularly in memory tests, have been reported. However, practically all studies had small sample sizes. Duinkerke and colleagues (2004) found that 6 of 10 patients improved in more than 50% of neuropsychological tests, especially memory and psychomotor speed tests. Thomas and colleagues (2005) found a significant improvement in 52.3% of patients (22 of 42 patients) in all tests performed 3 months after shunting. Of the 27 patients with a neuropsychological assessment at follow-up, 51.8% (14 of 27 patients) improved significantly in 50% or more of the tests administered, and 8 of the 15 patients assessed at follow-up using MMSE improved significantly after shunting (Thomas et al., 2005). The authors found significant improvement in memory function and in psychomotor speed performance, but they did not find any improvement in EFs.

Other authors, however, reported different conclusions. Savolainen and colleagues (2002) did not find any differences in neuropsychological test results, even though patients felt subjectively better after treatment. Stambrook and colleagues (1988) found that although measures of overall cognitive function, such as MMSE, semantic and figural memory composition, improve after shunting, some cognitive impairment may still be detected in the post-operative assessment, especially in attention and concentration, arithmetic skills, verbal and non-verbal memory, language and communication, and spatial and constructional skills. Hellstrom and colleagues (2008) concluded that neuropsychological function is not fully restored and the performance of iNPH patients after treatment is still below normal values.

After shunting, significant differences in the group as a whole were found in all tests except Digits Forward and Trail Making Part B. However, less than 50% of patients showed a significant improvement (with a change of 1 SD from baseline scores) when analyzed individually. At the 6-month follow-up, we found a significant improvement in psychomotor speed and in verbal memory test results. In overall cognition, 37 of our patients showed an increase of ≥4 points at follow-up in MMSE. Patients and relatives also observed this improvement. When we analyzed cognitive changes using the NPH scale, we found that 64.4% of patients improved their cognitive status. Although a considerable percentage of patients could not restore normal cognitive functioning, the clinical and neuropsychological improvement implies an important increase in the quality of life both for patients and families. This is clearly reflected in satisfaction scales completed by the family (unpublished results).

Considering our group of iNPH patients as a whole, significant improvements in nearly all neuropsychological domains assessed have been found after shunting. Only two tests (Digits Forwards and Trail Making Test Part B) of our cognitive battery did not show significant differences between assessments. Maybe one reason for the lack of improvement in Digits Forwards is that this test is not affected in iNPH patients (79.5% and 78.9% of patients scored in normal range at baseline and follow-up assessment consecutively) and, for Trail Making Test Part B, on the contrary, is affected in nearly all patients before and after treatment (81.6% of patients scored below of their reference group in both assessments).

Factors Influencing Cognitive Recovery and the Possibility of Mixed Cases

We analyzed the influence of demographic variables (age, sex, level of education, illness duration), cognitive variables (baseline cognitive status, cognitive disturbance as the first symptom), and vascular risk factors on cognitive results using the logistic regression method. Using this technique, we found that lower preoperative scores in MMSE are associated with improvement in MMSE, Information and Orientation, Digits Backwards, Semantic Verbal Fluency, and Purdue Pegboard (dominant hand), whereas higher scores in MMSE at baseline are associated with improvement in RAVLT (delayed recall) and TMT Part B. Shorter disease duration predicted improvement in MMSE and Purdue Pegboard (non-dominant hand) after surgery. A higher level of education is associated with improvement in Semantic Verbal Fluency and female gender was seen to predict the improvement in immediate verbal memory (RAVLT total words). These results imply the importance of detecting and treating patients before they are greatly affected by this syndrome to ensure the best response to treatment.

Another important factor to consider is that the basal degree of dementia in iNPH patients, or the lack of cognitive improvement after treatment, could be due to the coexistence of other pathologies that also affect cognitive domains, like Alzheimer's disease and vascular dementia (Duinkerke et al., 2004; Hellstrom et al., 2008; Savolainen et al., 2002; Thomas et al., 2005). The presence of concomitant causes of dementia should be considered, especially in those patients with gait and urinary control improvement but without cognitive changes after treatment.

The coexistence of Alzheimer's disease and iNPH could occur in 30%–75% of patients (Thomas et al., 2005). When these two pathologies coexist, it is more difficult for the clinician to decide whether the patient may be a candidate for shunt implantation. In these patients, different complementary tests, especially the neuropsychological assessment, may help in this decision (Benejam et al., 2008; Devito et al., 2005; Farace & Shaffrey, 2005; Savolainen et al., 2002). However, it is important to remember that even in these mixed patients with limited possibilities of cognitive improvement, improvement in gait and urinary, and/or fecal control may increase the degree of independence for daily living activities, increasing the quality of life for many patients and their families (Poca et al., 2005).

Study Limitations

A theoretical limitation of this study could be the lack of a control group to make definitive conclusions about the effect of shunt surgery on cognition in iNPH. However, in a previous study (Solana et al., 2010), we specifically addressed this issue using a GLS regression method with random effects and we were able to show that iNPH patients do not have any relevant practice effects in the repetition of the cognitive tests. Therefore, in the present study, we considered that practice effects were not relevant in iNPH patients, and traditional statistical instruments can be used without correction or control group.

A clear limitation is the modification of the battery of neuropsychological tools used due to the long inclusion period. Additionally, some patients could not be tested using some of the neuropsychological tests due to the low education level, physical disability, poor understanding of the task, and clinical condition, among others, interfering with the assessment and the absence of a follow-up of more than 6 months after treatment are additional limitations.

Conclusions

Cognitive impairment occurs in practically all patients with idiopathic NPH before surgical treatment. The most affected domains are memory, EFs, attention, and psychomotor speed. However, these patients can improve after shunting, especially in psychomotor speed, verbal memory, and general cognition. The best predictive variable for cognitive improvement after shunting is global cognitive status at baseline. This illustrates the importance of diagnosing and treating patients in the early stages of the disease. Another notable aspect is that although cognitive status is not normalized in many iNPH patients after treatment, partial cognitive improvement may contribute to a better quality of life for these patients and their families.

Funding

This study was funded in part by the Vall d'Hebron Research Institute, Universitat Autònoma de Barcelona (ES), and the Fondo de Investigación Sanitaria (07/0681 awarded to M.A.P.).

Conflict of Interest

None declared.

Acknowledgements

The authors gratefully acknowledge Sabrina Voss for editorial assistance, our team of research neuropsychologists, who contributed to the assessment of patients during the study period, Maria Mataró, Mar Matarín, and Bessy Benejam, and finally, the collaboration of the neurosurgical nurses in this study, especially Maria Jesus Peñarrubia (N.R.) and Olga Mestres (N.R.)

References

Adams
R. D.
Fisher
C. M.
Hakim
S.
Ojemann
R. G.
Sweet
W. H.
Symptomatic occult hydrocephalus with “Normal” cerebrospinal-fluid pressure. A treatable syndrome
New England Journal of Medicine
 , 
1965
, vol. 
273
 (pg. 
117
-
126
)
Benejam
B.
Poca
M. A.
Junque
C.
Solana
E.
Sahuquillo
J.
Cognitive impairment in normal pressure hydrocephalus (NPH). A proposal for clinical evaluation protocol
Neurocirugia (Asturias)
 , 
2008
, vol. 
19
 
4
(pg. 
309
-
321
)
Borgesen
S. E.
Conductance to outflow of CSF in normal pressure hydrocephalus
Acta Neurochirurgica (Wien)
 , 
1984
, vol. 
71
 
1–2
(pg. 
1
-
45
)
Chang
S.
Agarwal
S.
Williams
M. A.
Rigamonti
D.
Hillis
A. E.
Demographic factors influence cognitive recovery after shunt for normal-pressure hydrocephalus
Neurologist
 , 
2006
, vol. 
12
 
1
(pg. 
39
-
42
)
Chaudhry
P.
Kharkar
S.
Heidler-Gary
J.
Hillis
A. E.
Newhart
M.
Kleinman
J. T.
, et al.  . 
Characteristics and reversibility of dementia in normal pressure hydrocephalus
Behavioral Neurology
 , 
2007
, vol. 
18
 
3
(pg. 
149
-
158
)
Devito
E. E.
Pickard
J. D.
Salmond
C. H.
Iddon
J. L.
Loveday
C.
Sahakian
B. J.
The neuropsychology of normal pressure hydrocephalus (NPH)
British Journal of Neurosurgery
 , 
2005
, vol. 
19
 
3
(pg. 
217
-
224
)
Duinkerke
A.
Williams
M. A.
Rigamonti
D.
Hillis
A. E.
Cognitive recovery in idiopathic normal pressure hydrocephalus after shunt
Cognitive and Behavioral Neurology
 , 
2004
, vol. 
17
 
3
(pg. 
179
-
184
)
Evans
W. A.
An encephalographic ratio for estimating ventricular enlargement and cerebral atrophy
Archives of Neurology and Psychiatry
 , 
1942
, vol. 
42
 (pg. 
931
-
937
)
Farace
E.
Shaffrey
M. E.
Value of neuropsychological information for improved understanding of the patient with normal-pressure hydrocephalus
Journal of Neurosurgery
 , 
2005
, vol. 
102
 
6
(pg. 
971
-
972
discussion 972–973
Fillenbaum
G. G.
Screening the elderly. A brief instrumental activities of daily living measure
Journal of the American Geriatrics Society
 , 
1985
, vol. 
33
 
10
(pg. 
698
-
706
)
Folstein
M. F.
Folstein
S. E.
McHugh
P. R.
“Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician
J Psychiatr Res
 , 
1975
, vol. 
12
 
3
(pg. 
189
-
198
)
Gallia
G. L.
Rigamonti
D.
Williams
M. A.
The diagnosis and treatment of idiopathic normal pressure hydrocephalus
Nature Clinical Practice Neurology
 , 
2006
, vol. 
2
 
7
(pg. 
375
-
381
)
Gleichgerrcht
E.
Cervio
A.
Salvat
J.
Loffredo
A. R.
Vita
L.
Roca
M.
, et al.  . 
Executive function improvement in normal pressure hydrocephalus following shunt surgery
Behavioral Neurology
 , 
2009
, vol. 
21
 
3
(pg. 
181
-
185
)
Gustafson
L.
Hagberg
B.
Recovery in hydrocephalic dementia after shunt operation
J Neurol Neurosurg Psychiatry
 , 
1978
, vol. 
41
 
10
(pg. 
940
-
947
)
Hellstrom
P.
Edsbagge
M.
Blomsterwall
E.
Archer
T.
Tisell
M.
Tullberg
M.
, et al.  . 
Neuropsychological effects of shunt treatment in idiopathic normal pressure hydrocephalus
Neurosurgery
 , 
2008
, vol. 
63
 
3
(pg. 
527
-
535
discussion 535–526
Iddon
J. L.
Pickard
J. D.
Cross
J. J.
Griffiths
P. D.
Czosnyka
M.
Sahakian
B. J.
Specific patterns of cognitive impairment in patients with idiopathic normal pressure hydrocephalus and Alzheimer's disease: A pilot study
Journal of Neurology, Neurosurgery and Psychiatry
 , 
1999
, vol. 
67
 
6
(pg. 
723
-
732
)
Kahlon
B.
Sjunnesson
J.
Rehncrona
S.
Long-term outcome in patients with suspected normal pressure hydrocephalus
Neurosurgery
 , 
2007
, vol. 
60
 
2
(pg. 
327
-
332
discussion 332
Kahlon
B.
Sundbarg
G.
Rehncrona
S.
Comparison between the lumbar infusion and CSF tap tests to predict outcome after shunt surgery in suspected normal pressure hydrocephalus
Journal of Neurology, Neurosurgery and Psychiatry
 , 
2002
, vol. 
73
 
6
(pg. 
721
-
726
)
Katzen
H.
Ravdin
L. D.
Assuras
S.
Heros
R.
Kaplitt
M.
Schwartz
T. H.
, et al.  . 
Postshunt cognitive and functional improvement in idiopathic normal pressure hydrocephalus
Neurosurgery
 , 
2011
, vol. 
68
 
2
(pg. 
416
-
419
)
Katzman
R.
Hussey
F.
A simple constant-infusion manometric test for measurement of CSF absorption. I. Rationale and method
Neurology
 , 
1970
, vol. 
20
 (pg. 
534
-
544
)
Lezak
M.
Howieson
D. B.
Loring
D. W.
Neuropsychological assessment
 , 
2004
4th ed.
New York
Oxford University Press
Linn
M. W.
Linn
B. S.
The rapid disability rating scale-2
Journal of the American Geriatric Society
 , 
1982
, vol. 
30
 
6
(pg. 
378
-
382
)
Mataro
M.
Poca
M. A.
Salgado-Pineda
P.
Castell-Conesa
J.
Sahuquillo
J.
Diez-Castro
M. J.
, et al.  . 
Postsurgical cerebral perfusion changes in idiopathic normal pressure hydrocephalus: A statistical parametric mapping study of SPECT images
Journal of Nuclear Medicine
 , 
2003
, vol. 
44
 
12
(pg. 
1884
-
1889
)
Mirzayan
M. J.
Luetjens
G.
Borremans
J. J.
Regel
J. P.
Krauss
J. K.
Extended long-term (>5 years) outcome of cerebrospinal fluid shunting in idiopathic normal pressure hydrocephalus
Neurosurgery
 , 
2010
, vol. 
67
 
2
(pg. 
295
-
301
)
Poca
M.
Sahuquillo
J.
Barba
M.
Añez
J.
Arikan
F.
Prospective study of methodological issues in intracranial pressure monitoring in patients with hydrocephalus
Journal of Neurosurgery
 , 
2004
, vol. 
100
 
2
(pg. 
260
-
265
)
Poca
M. A.
Mataro
M.
Del Mar Matarin
M.
Arikan
F.
Junque
C.
Sahuquillo
J.
Is the placement of shunts in patients with idiopathic normal-pressure hydrocephalus worth the risk? Results of a study based on continuous monitoring of intracranial pressure
Journal of Neurosurgery
 , 
2004
, vol. 
100
 
5
(pg. 
855
-
866
)
Poca
M. A.
Mataro
M.
Matarin
M.
Arikan
F.
Junque
C.
Sahuquillo
J.
Good outcome in patients with normal-pressure hydrocephalus and factors indicating poor prognosis
Journal of Neurosurgery
 , 
2005
, vol. 
103
 
3
(pg. 
455
-
463
)
Poca
M. A.
Sahuquillo
J.
Mataró
M.
Update on diagnosis and treatment of normotensive hydrocephalus (chronic hydrocephalus of the adult)
Neurología
 , 
2001
, vol. 
16
 
8
(pg. 
353
-
369
)
Pujari
S.
Kharkar
S.
Metellus
P.
Shuck
J.
Williams
M. A.
Rigamonti
D.
Normal pressure hydrocephalus: long-term outcome after shunt surgery
J Neurol Neurosurg Psychiatry
 , 
2008
, vol. 
79
 
11
(pg. 
1282
-
1286
)
Raftopoulos
C.
Deleval
J.
Chaskis
C.
Leonard
A.
Cantraine
F.
Desmyttere
F.
, et al.  . 
Cognitive recovery in idiopathic normal pressure hydrocephalus: A prospective study
Neurosurgery
 , 
1994
, vol. 
35
 
3
(pg. 
397
-
404
discussion 404–395
Reitan
R. M.
The relation of the trail making test to organic brain damage
J Consult Psychol
 , 
1955
, vol. 
19
 
5
(pg. 
393
-
394
)
Sahuquillo
J.
Rubio
E.
Codina
A.
Molins
A.
Guitart
J. M.
Poca
M. A.
, et al.  . 
Reappraisal of the intracranial pressure and cerebrospinal fluid dynamics in patients with the so-called “normal pressure hydrocephalus” syndrome
Acta Neurochirurgica (Wien)
 , 
1991
, vol. 
112
 
1–2
(pg. 
50
-
61
)
Savolainen
S.
Hurskainen
H.
Paljarvi
L.
Alafuzoff
I.
Vapalahti
M.
Five-year outcome of normal pressure hydrocephalus with or without a shunt: Predictive value of the clinical signs, neuropsychological evaluation and infusion test
Acta Neurochirurgica (Wien)
 , 
2002
, vol. 
144
 
6
(pg. 
515
-
523
discussion 523
Solana
E.
Poca
M. A.
Sahuquillo
J.
Benejam
B.
Junque
C.
Dronavalli
M.
Cognitive and motor improvement after retesting in normal-pressure hydrocephalus: a real change or merely a learning effect?
J Neurosurg
 , 
2010
, vol. 
112
 
2
(pg. 
339
-
409
)
Stambrook
M.
Cardoso
E.
Hawryluk
G. A.
Eirikson
P.
Piatek
D.
Sicz
G.
Neuropsychological changes following the neurosurgical treatment of normal pressure hydrocephalus
Archives of Clinical Neuropsychology
 , 
1988
, vol. 
3
 
4
(pg. 
323
-
330
)
Strauss
E.
Sherman
E. M. S.
Spreen
O.
A compendiu of neuropsychological test. Administration, norms, and commentary
 , 
2006
3rd ed.
New York
Oxford University Press
Thomas
G.
McGirt
M. J.
Woodworth
G.
Heidler
J.
Rigamonti
D.
Hillis
A. E.
, et al.  . 
Baseline neuropsychological profile and cognitive response to cerebrospinal fluid shunting for idiopathic normal pressure hydrocephalus
Dementia and Geriatric Cognitive Disorders
 , 
2005
, vol. 
20
 
2–3
(pg. 
163
-
168
)
Vanneste
J. A.
Diagnosis and management of normal-pressure hydrocephalus
Journal of Neurology
 , 
2000
, vol. 
247
 
1
(pg. 
5
-
14
)
Wechsler
D.
A standardized memory scale for clinical use
J Psychol
 , 
1945
, vol. 
19
 (pg. 
87
-
95
)
Wikkelso
C.
Andersson
H.
Blomstrand
C.
Lindqvist
G.
Svendsen
P.
Normal pressure hydrocephalus. Predictive value of the cerebrospinal fluid tap-test
Acta Neurologica Scandinavica
 , 
1986
, vol. 
73
 
6
(pg. 
566
-
573
)