Impaired neuropsychological functioning in patients with hypopituitarism

Abstract Background Treatment of pituitary pathology mostly does not result in complete recovery of impairment in cognitive functioning. The primary aim of the current study was to assess cognitive impairment in patients with stable replacement therapy for hypopituitarism during the last 6 months prior to inclusion. It was expected that patients showed subjective and objective subnormal scores on neuropsychological functioning. Methods Forty‐two patients (40% men, 49 ± 15 years) treated for hypopituitarism conducted a neuropsychological test battery, including the Cognitive Failures Questionnaire (CFQ), 15‐Word test (15‐WT), Cambridge Neuropsychological Test Automated Battery (CANTAB) Motor Screening Task (MOT), Spatial Working Memory (SWM) and Affective Go/No‐go (AGN). Results were compared to reference values of healthy norm groups. Results Male and female participants scored significantly worse on the CFQ (P < .01, d = 0.91‐4.09) and AGN mean correct latency (P < .01, d = 1.66 and 1.29, respectively). Female participants scored significantly worse on 15‐WT direct recall (P = .01, d = 0.66), 15‐WT delayed recall (P = .01, d = 0.79), SWM total errors (P = .05, d = 0.41), SWM strategy (P = .04, d = 0.43), AGN errors of commission (P = .02, d = 0.56) and omission (P = .04, d = 0.41). Conclusion This study shows that subjective cognitive functioning is worse in patients treated for hypopituitarism compared to reference data. Also, female participants treated for hypopituitarism score worse on objective aspects of memory and executive functioning compared to reference data. Besides worse focus attention, this objective cognitive impairment was not found in male participants. It is recommended to conduct additional research, which focuses on the design and evaluation of a cognitive remediation therapy, aimed at compensation of impairments in different aspects of memory and executive functioning.


| INTRODUC TI ON
Hypopituitarism is a rare condition characterized by a deficit of one or more of the pituitary hormones. Clinical manifestations of hypopituitarism vary widely since it represents a diverse group of conditions, including differences in aetiology, pathology and hormonal deficiencies. Treatment of hypopituitarism consists of hormone substitution therapy. Unfortunately, current substitution methods do not completely mimicry the endocrine homeostasis and they are titrated based on plasma variables and not on action at tissue level. 1 Therefore, subtle derangements in hormonal status may appear in patients treated with substitution therapy, which in turn can lead to vague complaints and thereby a decreased quality of life. 1 Ehrnborg et al 2 found that patients with hypopituitarism, compared to the general population, had more health-related costs, took more sick leave days and were more likely to claim a disability pension. Studies also found a decreased mood state in patients treated for hypopituitarism. 3,4 Altogether, patients with hypopituitarism still experience problems and effects of their illness, even after treatment of their hormonal disturbances.

Most of the pituitary (stimulating) hormones and their products
can cross the blood-brain barrier and connect to specific binding sites throughout the central nervous system. This finding suggests that hormones act on the brain and that hormone deficiencies can have an effect on brain functioning and structure. Studies have indeed shown that hormonal deficiencies negatively affect neuropsychological functioning, such as memory and executive functioning. [5][6][7] The finding of a compromised neuropsychological functioning in patients with pituitary disease can still be found after adequate replacement therapy. Scientific research showed impairments in memory, 8,9 executive functions, 8-10 attention 8, 11 and visuospatial processing 8 in patients following successful treatment of Cushing's disease (CD). Structural brain abnormalities in these patients may attribute to their impaired neuropsychological functioning. [12][13][14] Hypopituitarism in patients with effective treatment of CD was associated with mildly impaired executive functioning. 9 For patients with successful treatment of acromegaly, a decrease in memory is also found, which is associated with a decreased activity in the left medial temporal cortex. 15 The same study of Martín-Rodríguez et al 15 showed a trend of decrease in performance on a visual memory task and less activity in the left medial temporal cortex for patients with hypopituitarism after treatment of acromegaly. Besides a decrease in memory, patients in remission from acromegaly also show a decrease in executive functioning. 16 It could be stated that this neuropsychological deterioration is the result of previous hormonal overproduction in CD and acromegaly, but the same neuropsychological deterioration is found in patients treated for nonfunctioning pituitary macroadenoma (NFPA), 17 a condition without hormonal overproduction. Another study on patients treated for NFPA found that these patients show a high self-reported prevalence and severity of cognitive dysfunction. 18 Hence, it seems that cure of pituitary disorder does not result in complete recovery of neuropsychological function, which can be due to structural brain changes by previously hormonal over-and/or underproduction and/or shortcomings in treatment.
Some of the long-term effects of hypopituitarism on cognitive functioning are known for different aetiologies of hypopituitarism, as described by the studies mentioned above. Most research in this field has been directed on a restricted range of cognitive domains within specific subgroups of hypopituitarism. However, during clinical practice, endocrinologists face patients with all different kinds of aetiologies of hypopituitarism. The aim of the current study, therefore, was to assess a spectrum of cognitive functions in a heterogeneous group of patients with hypopituitarism with subjective and objective measures, at least six months after adequate replacement therapy. These insights are of importance because they can provide essential information for clinical practice about the course of hypopituitarism and the expected long-term problems on cognitive functioning, which patients might encounter in general. Also, no other studies have been performed on both subjective and objective cognitive functions within this patient group. This study might provide more insight in what specific subdomains patients treated for hypopituitarism could underperform. It was expected that these patients show impaired scores on neuropsychological tasks. and/or their end-products were above cut-off scores defined by the Endocrine Society Clinical Practice Guideline. 19 Patients who were mentally retarded, suffering from dementia or severe (cognitive) illness or chronic users of medication that influences consciousness were not included.

| Short-and long-term memory
The 15-Word test (15-WT) 23 was used to measure different aspects of verbal memory: short-term memory (direct recall) and long-term memory (delayed recall). In the direct recall condition, the investigator read 15 words out loud, after which the patient had to remember and repeat as much words as possible. This was done five times, and the score on the direct recall was the sum of the correctly remembered words for the five repeats (max score of 75). After 20 minutes, the delayed recall was administered, in which the patient was asked to name all the words on the list they remembered (max score of 15).

| Motor Screening Task
The first administered CANTAB ® test was the Motor Screening Task (MOT), which was used to introduce the software to the participant and to screen for sensorimotor deficits or lack of comprehension.

| Spatial working memory and executive function
During the CANTAB ® Spatial Working Memory Task (SWM), participants were asked to fill up an empty column on a touchscreen with blue "tokens", which they found under a series of coloured boxes. The more boxes on the screen, the more boxes the participants had to search through and the harder the test was. Next to working memory (eg, retention and manipulation of visuospatial information), this task also demanded executive functioning (eg, strategy). Administration time was estimated at 4 minutes.
Outcome measures were double errors, total errors and strategy.
A high score on double or total errors meant poorer working memory and executive functioning, and high scores on strategy represented worse executive functioning. Age-and gender-related reference values were derived from normative data provided by CANTAB.
There were missing data for one participant on the AGN so this participant was excluded for results on the AGN.

| Subjective cognitive functioning
Subjective cognitive functioning data are shown in Table 3. Both

| Short-and long-term memory
More than 50% of the female participants had a score below the 30th percentile on both memory subscales, and about 30% had a score below the 10th percentile. Scores of males were less often in these low percentiles, as shown in Table 4. Women scored significantly lower on the 15-WT direct recall and 15-WT delayed recall when compared to their reference sample, with medium effect sizes (d = 0.66 and 0.79, respectively). No significant differences were found for men.

| CANTAB tests
On the Motor Screening Task (MOT), all but 2 patients were faster and all but one made fewer errors than reference data. This means that, in general, there was no presence of sensorimotor deficits or lack of comprehension.
As shown in Figure 1, female participants scored significantly worse on Spatial Working Memory (SWM) total errors and SWM strategy, with small to medium effect sizes (d = 0.41 and 0.43, respectively) when compared to reference data. No differences were found on the SWM between errors subtest. Also, no differences with reference data were found for male participants.
Affective Go/No-Go (AGN) data are shown in Table 5. Mean affective bias was negative in both men and women, which means that participants had faster reaction times during positive blocks than during negative blocks (a positive bias), which was comparable to reference data. When compared to reference data, men

| DISCUSS ION/CON CLUS ION
The aim of the current study was to assess with objective measures a wide spectrum of cognitive functions in patients with different ae-

TA B L E 5 Scores on the Affective
Go/No-Go (AGN) and comparisons to reference data functioning compared to reference data. These results were not found in male participants. On affective information processing bias, both men and women had slower reaction times and women made more commission and omission errors compared to reference data.
This shows that especially in women, focus attention and general competence in inhibiting behavioural responses is worse. Since no differences between set-shift and nonshift blocks within the participants were found, it seems that participants had no problem in their abilities to suppress and reverse stimulus-reward associations.
Overall, participants reacted faster on positive than negative words, implying a positive bias. A possible explanation for this is that a positive attentional bias could be part of a coping mechanism for dealing with (chronic) illness. Previous research of Caprara et al 31  Since the highest levels of self-reported cognitive dysfunction and severity were found in NFPA, the study indicates that cognitive dysfunction can still be present while hormonal levels are within normal range. A limitation to this cited study is that no objective cognitive variables were measured, so outcomes solely depended on patient's self-perception. To our knowledge, our study is the first to assess both subjective and objective cognitive functioning in hypopituitarism. This is of importance since discrepancies between self-per- Despite the intentional selection of the heterogeneous group of patients with hypopituitarism, this also forms a limitation.
Aetiologic groups are represented in small numbers: that is ten participants with nonfunctioning pituitary adenoma, five with acromegaly, five with Cushing's disease and four prolactinoma.
Therefore, the results of this heterogeneous group are difficult to translate to the situation of individual patients. However, we in-

| CON CLUS IONS
Our study is the first to point out that neuropsychological functioning, regarding different aspects of memory and executive functioning, is impaired in a heterogeneous group of patients with treated hypopituitarism. Subjective dysfunction seemed impaired in both men and women but when objectifying the cognitive domains, only women appear to be affected. Hence, it seems that adequate replacement therapy in hypopituitarism does not result in complete recovery of neuropsychological function in females. This study is of importance because it provides essential information for clinical practice about the course of hypopituitarism and the expected long-term problems on specific subdomains of cognitive functioning, which patients might encounter in general. Current treatment mainly focuses on hormonal disturbances 46 rather than functional outcomes. Therefore, it is recommended to conduct additional research, which focuses on the design and evaluation of a cognitive remediation therapy. Our study proposes that such an intervention should be directed on compensation of impairments in different aspects of memory and executive functioning. By developing such an intervention, neuropsychological complaints in treated hypopituitarism may decline, which in turn may lead to an improvement of quality of life.

ACK N OWLED G EM ENTS
The authors would like to thank all study participants for their time and effort.

CO N FLI C T S O F I NTE R E S T
The authors have no conflicts of interest to declare.

DATA AVA I L A B I L I T Y S TAT E M E N T
Research data are not shared.