diabetes research and clinical practice 177 (2021) 108898 Contents available at ScienceDirect Diabetes Research and Clinical Practice journal homepage: www.elsevier.com/locat e/dia bre s Defective involuntary attention to novelty in type 1 diabetes and impaired awareness of hypoglycaemia Nicole Stantonyonge a,e,1, Saul Martinez-Horta b,c,d,1, Frederic Sampedro b,c,d, Marı́a Belén Sánchez-Saudinós b,c,g, Ana Chico a,e,f,* a Department of Endocrinology and Nutrition, Hospital de Santa Creu i Sant Pau, Barcelona, Spain b Biomedical Research Institute Sant Pau (IIB-Sant Pau), Barcelona, Spain c Centro de Investigación Biomédica en Red-Enfermedades Neurodegenerativas (CIBERNED), Madrid, Spain d Movement Disorders Unit, Neurology Department, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain e Department of Medicine, Autonomous University of Barcelona, Barcelona, Spain f CIBER-Bioengineering, Biomaterials and Nanotechnology (CIBER-BBN), Instituto de Salud Carlos III, Madrid, Spain g Memory Unit, Neurology Department, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain A R T I C L E I N F O A B S T R A C T Article history: Aim: To determine if there are differences in terms of neurophysiology and neurocognitive Received 19 March 2021 functioning in a group of type 1 diabetes (T1D) patients regarding hypoglycaemia aware- Received in revised form ness. 25 May 2021 Methods: 27 patients with T1D were classified according to Clarke score as having impaired Accepted 1 June 2021 awareness of hypoglycaemia (IAH; n = 11) or normal awareness to hypoglycaemia (NAH; Available online 4 June 2021 n = 16). We measured several clinical and sociodemographic variables and cognitive perfor- mance using neuropsychological tests. Electroencephalography was assessed during an auditory oddball task. We compared the groups in terms of clinical/sociodemographic vari- Keywords: ables as well as two event-related brain potentials (ERPs): The P3a which is associated with Type 1 diabetes automatic orientation of attention to novelty, and the P3b which is associated with target Hypoglycaemia detection and processing. Impaired awareness Results: The IAH group performed significantly worse on the Trail Making Test part A (TMT- Neurophysiology A) (p = 0.05). Compared to the NAH group, P3a and P3b amplitudes in the frontal-central sites were significantly lower in the IAH group (p < 0.05). The P3a was strongly associated with worse performance on the TMT-A in the IAH group (r = 0.540; p < 0.005) Conclusion: IAH is accompanied by decreased neurophysiological activity in ERPs associ- ated with information processing and with the automatic orientation of attention to nov- elty and environmental changes. These findings suggest a possible framework to better understand the cognitive origin of IAH in this patient population. Ó 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY- NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). * Corresponding author at: Department of Endocrinology and Nutrition, Hospital de Santa Creu i Sant Pau Mas Casanovas, 90 - 08041 Barcelona, Spain. E-mail address: achicob@santpau.cat (A. Chico). 1 These authors have contributed equally. https://doi.org/10.1016/j.diabres.2021.108898 0168-8227/Ó 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 2 diabetes research and clinical practice 177 (2021) 108898 1. Introduction neural activity associated with various cognitive processes [30]. Of these, the P300 (or P3b) elicited during auditory odd- In type 1 diabetes (T1D), long-term insulin treatment is not ball paradigms have been extensively studied in healthy indi- only essential due to the insulinopenic state that defines this viduals and in patients with several different conditions [31– group of patients, but it is also required to control glucose 34]. The P3b is a positive deflection appearing with maxim levels to reduce the risk of complications derived from amplitude in central-parietal sites around 300–600 ms after chronic hyperglycaemia [1,2]. However, insulin treatment is detection of a target. The functional meaning of the P3b is associated with an increased risk of hypoglycaemic events [2]. quite heterogeneous, and the morphology of this component These events are characterized by several symptoms of appears to be modulated by several variables (i.e.: older age, varying severity including spontaneous sweating, dizziness, disease, etc). However, the P3b is undoubtably involved in tar- tremor, increased hunger, and irritability, among others that get detection, processing speed and contextual update of serve to warn the patient [3]. Patients able to identify these information. The P3a or ‘‘novelty P3” displays maximum signs can act accordingly and take appropriate measures to amplitude over frontal-central sites with a latency around address these episodes. Awareness of these signs is vitally 250–500 ms in response to involuntary processing and orien- important given that failure to prevent and manage hypogly- tation of attention to novel or relevant stimulus [33,35]. These caemia may lead to hypoglycaemia unawareness and severe measures have proven sensitive for the early detection of hypoglycaemia (SH) [4,5]. neuronal damage, even in the absence of clear indicators of SH is defined as a hypoglycaemic event associated with cognitive decline [36–38]. This suggests that evaluation of severe and potentially harmful symptoms including loss of the morphology of these ERPs in T1D-IAH may be an objective consciousness, seizure, or coma requiring the assistance of and robust approach to exploring the potential presence of a third person [6]. It has been reported a median rate of 1.1 signs of neuronal damage in this patient population, and to episodes for patient-year and a yearly prevalence of 4–36% understand the cognitive mechanisms involved in lack of of SH episodes in T1D [7–10]. The main risk factor for develop- awareness about hypoglycaemia. Thus, the presence of elec- ing an SH-related episode is impaired awareness of hypogly- trophysiological differences in the absence of a significant caemia (IAH), defined as the inability to perceive the cognitive impairment is especially relevant, since these differ- symptoms of hypoglycaemia [11]. Compared to individuals ences could be predictive biomarkers of cognitive decline. with a normal awareness of hypoglycaemia (NAH), T1D In this context, the main aim of the present study was to patients with IAH have a much greater risk of presenting assess and compare the morphology of the P3a and P3b com- more SH episodes, and thus of developing complications ponents in patients with TD1-IAH and NAH. A second aim associated with recurrent episodes of hypoglycaemia [12]. was to explore potential associations between neurophysio- It is widely accepted that diabetes mellitus is associated logical findings and clinical parameters. with an increased risk for cognitive impairment [13]. This association has been extensively studied in patients with type 2. Material and Methods 2 diabetes, in whom various mechanisms contribute to neu- ronal damage and concurrent cognitive deterioration [14,15]. 2.1. Clinical assessments Although several studies have evaluated the impact of T1D on cognitive function, the results are highly heterogenous The following clinical and sociodemographic data were [13,16,17]. The large, prospective Diabetes Control and Com- obtained and recorded for all patients: age; years of educa- plications Trial (DDCT) found no significant differences tion; disease duration; body mass index (BMI); type of insulin between patients with T1D and healthy controls in cognitive treatment (multiple insulin injections [MDI] or continuous function over a mean follow-up of 18 years [18]. By contrast, subcutaneous insulin infusion [CSII]); basal and prandial several cross-sectional studies and meta-analyses have found insulin dose (UI/kg/day); mean HbA1c and number of self- that, compared to healthy controls, patients with T1D gener- reported SH episodes during last year and last five years ally perform worse on several cognitive domains [16,19–21]. defined as a hypoglycaemic event requiring the assistance While numerous factors are believed to be involved in the of a third person; glucose levels during study; and LDL association between T1D and cognitive deterioration, a prior cholesterol. history of SH episodes appears to play a key role [22–24]. The Clarke scale was used to assess the participants’ Although the exact mechanisms contributing to this associa- awareness of hypoglycaemia [39]. The Clarke test is an 8- tion are unknown, it is reasonable to assume that SH episodes item questionnaire (validated for the Spanish population) in may be associated with some degree of neuronal damage, which the patient self-evaluates his or her awareness of which contributes to cognitive deterioration. Given that IAH hypoglycaemia. Based on the Clarke scores, participants were is associated with an increased recurrence of episodes of classified as NAH (Clarke score  2) or IAH (Clarke score  4) SH, it is plausible to hypothesize that IAH may also be associ- [39]. A scored of 3 points on the test was considered indeter- ated with an increased neuronal damage load and thus a minate and thus these patients were excluded from the study. greater risk of cognitive deterioration [25–29]. Mood and cognition were assessed by a battery of neu- Event-related brain potentials (ERPs) permit the assess- ropsychological assessments. The Beck Depression Inventory ment, with high temporal resolution, of the time course of (BDI) was used to determine the presence and severity of diabetes research and clinical practice 177 (2021) 108898 3 depressive symptoms [40]. Global cognitive status was stimulus. Epochs were baseline-corrected, subtracting the assessed with the Mini Mental- State Evaluation (MMSE) mean amplitude in the 100 ms before stimulus presentation. [41], a commonly used instrument that screens for global cog- A two-step artifact rejection procedure was then applied. The nitive status. Scores < 26 points on the MMSE are suggestive first epochs were rejected if the signal in any of the 19 chan- of the presence of cognitive impairment of variable severity nels showed amplitude values > ±300 lV. Subsequently, addi- while scores > 26 suggest cognitive normality. The word tional epochs were excluded if amplitude values were > learning subtest of the Consortium to Establish a Registry ±40 lV in Fz, Cz or Pz channels. After these preprocessing for Alzheimer’s Disease (CERAD) was used to examine imme- steps, three types of trials—epochs containing frequent, infre- diate learning, delayed recall, and recognition [42]. Attention quent, and novel stimuli—were averaged separately. These and executive functions were assessed with the Trail Making averages were obtained for each participant and the ERP com- test (TMT) parts A and B [43], the phonetic verbal fluency ponents were identified and quantified. The P3b was the most (words starting with letter P) during one minute, and the positive deflection in the ERP between 300 and 650 ms post- semantic verbal fluency (animals) during one minute. Finally, stimulus in the infrequent trials while the P3a was the most confrontation naming was assessed with the short version of positive deflection in the ERP between 240 and 500 ms post- the Boston Naming Test [44]. Raw scores were adjusted using stimulus in the novel trials. The mean amplitude was calcu- the normative data for Spanish population [45]. lated in the time-window defining each ERP and introduced into the statistical analysis. 2.2. Auditory oddball task 2.5. Statistical analysis A standard auditory oddball task with a set of frequent, infre- quent, and novel stimuli was used. Frequent stimuli (1500 Hz, For the clinical and sociodemographic variables, independent duration: 60 ms) occurred with a probability of 0.8, infrequent t-test comparisons between the two groups were performed stimuli (1620 Hz; 60 ms) occurred with a probability of 0.1, and for continuous variables and the v2 test for categorical vari- novel sounds (e.g., a key or a door closing) occurred with a ables. ERP effects were quantified for the three midline elec- probability of 0.1 at 60 dB for 60 ms. Over a period of trodes (Fz, Cz, and Pz); these data were then evaluated by 15 min, participants were requested to respond as quickly repeated-measures analysis of variance (ANOVAs), applying as possible to infrequent (target) sounds with their right index the Greenhouse-Geisser correction when necessary. Post- finger and to ignore all other sounds. The percentage of cor- hoc comparisons were performed using paired and indepen- rect responses to the target sound and the percentage or dent t-test comparisons. responses to non-target novel stimuli were registered. 2.3. EEG recording 3. Results An electroencephalogram (EEG) was recorded with the Brai- The sample was comprised of 27 adults (12 females and 15 nAmp system and Brain Vision Recorder Software (Brain males), with a mean ± standard deviation (SD) age of 58.5 Products GmbH, Germany) at 19 standard scalp sites (Fp1/2, ± 10.8 years. The mean disease duration was 30.1 ± 9.9 m F3/4, C3/4, T3/4, T5/6, P3/4, O1/2, F7/8, Fz, Cz, Pz) referenced onths. Mean basal daily insulin dose (UI/kg) was 0.28 ± 0.11 to the two mastoid leads. Vertical eye movements were and mean prandial daily insulin dose (UI/kg) was recorded using a bipolar montage with two electrodes linked 0.27 ± 0.17. together and placed below each eye, which were referenced to The mean Clarke score was 2.56 ± 2. Of the 27 patients, 16 a third electrode placed centrally above the eyes. To monitor were classified as IAH and 11 as NAH. The two groups were horizontal eye movements, two electrodes were placed on similar for nearly all sociodemographic and clinical variables the external canthi of each eye. Electrode impedances were of interest (Table 1). However, the IAH group presented signif- kept below 5 kOhm. The electrophysiological signals were fil- icantly higher values on several parameters, as follows: mean tered with a bandpass of 0.1–35 Hz and digitized at a rate of basal daily insulin doses [t(27) 2.88; p < 0.01]; total number of 250 Hz. SH episodes in the last year [t(27) = 2.1; p < 0.05] and last five To maximize the information available for the subsequent years [t(27) = 2.57; p < 0.05].Table 2. ERPs, raw EEG signals were subjected to an ocular artifact Patients in the IAH group scored significantly higher on the minimization algorithm based on an eigenvalue decomposi- BDI (t(27) < 0.005), indicating worse depressive symptoms. On tion of time-delayed covariance matrices. After identifying the cognitive measures, no significant between-group differ- the source signals associated with eye movements, we ences were observed in the MMSE and moreover, the scores obtained corrected EEG signals from the remaining compo- obtained in this test were in the normal range (MMSE > 26). nents. The algorithm was implemented using Brain Vision Similarly, no significant differences were observed in the total Analyzer (Brain Products GmbH; Germany). corrected scores or in the percentage of cases in each group scoring in the clinical range (z < 1.5) for any of the measures. 2.4. P3a and P3b analysis However, a slight difference [t(27) = 2.03; p = 0.05] was found in the TMT-A regarding the percentage of cases scoring in In the auditory oddball task, the continuous EEG recording the clinical range. This difference was found in relation to a was segmented in epochs of 1000 ms starting 100 ms before higher proportion of patients scoring in the clinical range in the stimulus presentation and continuing until 900 ms post- the IAH group (IAH = 27.3% vs NAH = 6.3%). 4 diabetes research and clinical practice 177 (2021) 108898 Table 1 – Clinical and sociodemographic data. IAH (n = 11) NAH (n = 16) p Age (years) 57 ± 12.8 59.5 ± 9.4 0.570 Gender (% female) 58.3 41.7 0.096 Education (years) 13.2 ± 6.3 14.7 ± 4.5 0.478 BMI (kg/m2) 25.7 ± 6.8 27.5 ± 3.7 0.407 Disease duration (years) 29.6 ± 12.6 30.5 ± 8.1 0.818 MDI/CSII (%) 61.1/38.9 73.7/26.3 0.315 Clarke score 4.8 ± 0.8 1 ± 0.7 <0.001 Basal insulin dose (UI/kg/day) 0.21 ± 0.4 0.32 ± 0.1 0.020 Prandial insulin dose (UI/kg/day) 0.22 ± 0.1 0.31 ± 0.21 0.129 HbA1c (%) 7.7 ± 1 6.8 ± 3.2 0.364 Last year 7.5 ± 0.9 7.5 ± 0.7 0.949 Last 5 years 7.5 ± 0.8 7.5 ± 0.5 0.811 SH episodes in previous year 0.6 ± 1 0.06 ± 0.2 0.097 SH episodes in last 5 years 2.1 ± 3 0.1 ± 0.3 0.059 DKA episodes in previous year 0.1 ± 0.5 0.07 ± 0.2 0.829 DKA episodes in last 5 years 0.3 ± 0.6 0.1 ± 0.3 0.262 Retinopathy (% prevalence) 23.1 35.3 0.377 Glycaemia at study (mg/dL) 134.5 ± 85.1 154.5 ± 80.1 0.546 LDL cholesterol (mg/dL) 186.3 ± 29.9 159.7 ± 51.7 0.138 MMSE (score) 29.5 ± 0.6 29.3 ± 1.2 0.676 BDI (score) 10 ± 5.5 4.3 ± 3.7 0.004 Abbreviations: IAH, impaired awareness of hypoglycaemia; NAH, normal awareness to hypoglycaemia Body mass index; 2Multiple daily insulin injections/continuous subcutaneous insulin infusion; SH, severe hypoglycaemia; DKA, diabetic ketoacidosis; MMSE, Mini-mental State Examination; BDI, Beck Depression inventory. Table 2 – Neuropsychological assessment data. IAH (n = 11) NAH (n = 16) p CERAD word list memory 8.18 ± 1.6 7.8 ± 1.8 0.660 Corrected standardized score - - % cases scoring z < -1.5 18.2 12.5 0.545 CERAD word list recall 6.2 ± 1.8 5.5 ± 2.3 0.408 Corrected standardized score – – % cases scoring z < -1.5 9.1 12.5 0.643 CERAD word list recognition 19.4 ± 1 19.4 ± 1.2 0.864 Corrected standardized score – – % cases scoring z < -1.5 0 6.3 0.398 TMT-A2 52.9 ± 20.3 42.4 ± 15.9 0.144 Corrected standardized score 6.73 ± 2.5 9.2 ± 3.4 0.053 % cases scoring z < -1.5 27.3 6.3 0.131 TMT-B2 125.8 ± 81.3 95.1 ± 60.8 0.271 Corrected standardized score 7.5 ± 2.1 9 ± 2.5 0.119 % cases scoring z < -1.5 9.1 12.9 0.643 Phonetic verbal fluency 15.3 ± 4.5 15.1 ± 4.9 0.899 Corrected standardized score 10 ± 1.6 9.5 ± 2.7 0.569 % cases scoring z < -1.5 0 12.5 0.342 Semantic verbal fluency 19.8 ± 4.9 19.6 ± 5.7 0.929 Corrected standardized score 9.1 ± 2.8 8.5 ± 2.7 0.537 % cases scoring z < -1.5 9.1 18.8 0.488 Boston naming test 12.3 ± 1.2 13 ± 1.6 0.282 Corrected standardized score 9.9 ± 1.8 10.8 ± 3.1 0.368 % cases scoring z < -1.5 0 6.3 0.398 Abbreviations: IAH, impaired awareness of hypoglycaemia; NAH, normal awareness to hypoglycaemia; BMI, body mass index; CERAD. Con- sortium to Establish a Registry for Alzheimer’s Disease; TMT, Trail Making Test parts A and B. 3.1. Auditory oddball task using the factors ‘‘electrode” (Fz, Cz, Pz) and ‘‘condition” (fre- quent/infrequent) showed significant effects at Pz [F No between-group differences in the number of averaged (1,26) = 15.9; p < 0.001] driven by significant increased positiv- epochs or correct responses (response to target signal) were ity at Pz [t = -3.99; p < 0.001] around 400 ms after delivery of an observed. Repeated measures ANOVA in the overall sample infrequent stimulus (P3b). When novel stimuli were added to diabetes research and clinical practice 177 (2021) 108898 5 the factor ‘‘condition”, significant effects were found at Fz [F Repeated measures ANOVA using the factors ‘‘group” (IAH; (1,26) = 27.1; p < 0.001], Cz [F (1,26) = 11.2; p < 0.005], and Pz NAH) and ‘‘condition” (frequent; novel) showed a significant [F (1,26) = 21.08; p < 0.001], driven by a significantly increased group  condition interaction for the P3a at Fz [F positivity at Fz [t = 5.2; p < 0.001], Cz [t = 3.3; p < 0.005] and (1,26) = 5.02; p < 0.05], and Cz [F (1,26) = 4.59; p < 0.05]. Post- Pz [t = 4.5; p < 0.001] around 250 ms–400 ms following the hoc t-test comparisons showed that this effect was driven novel stimulus (P3a) (Fig. 1) (see Fig. 2.). by a significant reduction in P3a in the IAH group at Fz [t Fig. 1 – Stimulus-locked grand average ERPs for the whole sample. Grand average at Fz and Pz electrodes depicting the ERPs for frequent (thin line), infrequent (grey line) and novel (black line) stimuli. The P3a was identified as the most positive ongoing deflection between 240 and 500 ms (grey area) post-stimulus in the novel trials. The P3b was identified as the most positive deflection in the ERP between 400 and 8000 ms (grey area) post-stimulus in the infrequent trials. The topographical map shows the frontal-central distribution of the P3a, and the central-parietal distribution of the P3b. 6 diabetes research and clinical practice 177 (2021) 108898 Fig. 2 – A) Stimulus-locked (grand average) ERPs in each group. B) Topographical maps showing the scalp distribution of the mean voltage of the ERPs in each group. A decreased voltage is seen in the IAH group both for the P3a and P3b. C) Difference wave-form for the novelty P3a and scatter-plot showing the association between the amplitude of the difference wave-form and performance in the TMT-A. (27) = 2.63; p < 0.05] and Cz [t(27) = 2.39; p < 0.05]. In parallel, T1D patients with IAH presented significant neurophysiologi- the comparisons made with the difference waveform (novel - cal differences in both the P3a and P3b signals compared to frequent) also corroborated a lower amplitude of P3a in the patients with NAH. Importantly, both groups were compara- IAH group at Fz [t(27) = 2.61; p < 0.05] and Cz [t(27) = 2.36; ble in terms of almost all sociodemographic and clinical vari- p < 0.05]. Repeated measures ANOVA using the factors ables, but the IAH group had higher scores for depressive ‘‘group” (IAH; NAH) and ‘‘condition” (frequent; infrequent) mood and worse performance in the TMT-A, a processing showed a significant group  condition interaction for the speed and attention task. P3b at Fz [F (1,26) = 4.3; p < 0.05]. Post-hoc t-test comparisons T1D patients with IAH exhibited significantly decreased showed that this effect was driven by a significant reduction amplitudes in both the P3a and P3b components in frontal- in the P3b in the IAH group at Fz [t(27) = 2.1; p < 0.05], also seen central locations, and a tendency to elicit the P3a with a at Fz [t(27) = 2.1; p < 0.05] using the difference waveform (in- delayed latency. Although no between-group differences were frequent – frequent). observed in cognitive parameters assessed through neuropsy- In terms of ERP latency, a trend towards significance was chological assessment, performance in the TMT-A was worse found for the P3a [t(27) = 1.8; p = 0.076], resulting from an in the IAH group. Interestingly, the amplitude of the P3a was increased latency in the IAH group (IAH = 424 ± 34 ms vs. strongly associated with performance of these tasks, suggest- NAH = 391 ± 53 ms). A bivariate correlation analysis showed ing the involvement of disrupted neurophysiological mecha- that the P3a amplitude was significantly associated with the nism in the IAH group. normalized score on the TMT-A (r = 0.540; p < 0.005), with a Our findings suggest that IAH is associated with disrupted non-significant trend with the Clarke score (r = 0.344; neuronal processes that have minimal effects on cognitive p = 0.085). functioning, at least at the point in time that the study was performed. Nevertheless, as our findings show, dysfunctional 4. Discussion neuronal processes can be detected through the measure- ment of neurophysiological differences. Although the func- The aim of the present study was to characterize the effects of tional meaning of these findings cannot be determined from impaired awareness of hypoglycaemia in patients with T1D the data obtained in the present study, these results suggest using specific neurophysiological parameters associated with several interesting possible interpretations. First, given the attention and cognitive processing. We also sought to deter- association between IAH and more frequent recurrence of mine the association between the presence of IAH and other severe hypoglycaemia, it is reasonable to suppose that there measures of neuropsychological functioning. We found that may be a causal relationship between these recurrent epi- diabetes research and clinical practice 177 (2021) 108898 7 sodes and the concurrent presence of subtle neuronal dam- episodes. Given the between-group differences in depressive age [22,29,46]. Depending on the individual’s clinical stage, symptom severity, it would be interesting to explore the pos- these patterns may not accumulate sufficient neuronal dam- sible relationship between depression, attention, and aware- age load to produce evident cognitive deficits, even though ness of hypoglycaemia. In fact, it is widely recognized that may be detectable through neurophysiological measures. the amplitudes of p3a and p3b are reduced in patients with Thus, it is also reasonable to assume that, in the long term, depressive symptoms, with a negative impact on attention continued and recurrent SH episodes will cause progressive, and processing speed [50,51]. Thus, our findings may reflect cumulative neural damage, eventually leading to significant mechanisms associated with the etiology of difficulties in cognitive impairment with important implications for the detecting hypoglycaemic symptoms in patients with IAH. patient’s quality of life [13,17,26]. Clearly, an in-depth study will be required to explore the The findings of this study underscore the need to avoid, or mechanisms involved and the implications. Assuming that at least minimize, recurrent episodes of SH in order to prevent there is an association between IAH, SH, and concurrent neu- potentially clinically meaningful deleterious consequences ronal damage, more research is needed to accurately eluci- for cognitive functioning. However, longitudinal studies are date and characterize the mechanisms that promote needed to confirm whether the observed neural damage is neuronal damage in patients with T1D-IAH. Similarly, if we cause or consequence of IAH and cognitive alterations. It is assume that there is a causal association between defective also important to consider that we examined only a limited attention and the development of IAH, it will be necessary number of ERPs and neuropsychological tasks in this study. to determine why some T1D patients present alterations in Thus, the apparent absence of clinically relevant neuropsy- brain processes related to the orientation of attention that chological deficits and/or differences between groups may make them unable to detect hypoglycaemia symptoms. be attributable to two factors: a) the selected tests may not The present study has several important limitations, par- be sufficiently sensitive to detect existing deficits and/or b) ticularly the small sample size, which precludes the general- the patients could have symptoms in other domains or pro- ization of these findings. In addition, the neuropsychological cesses that have not been explored. A second possible inter- examination performed in this study does not cover all of the pretation of our findings is related to the frontal P3a, or potential cognitive domains of interest. Consequently, future novelty P3, which has been extensively associated with the studies should use a larger battery of neuropsychological engagement of attention resources, especially those involved assessments. By contrast, the main strength of this study is in the orientation of attention to relevant environmental the description of a novel, potentially explanatory and theo- changes or novelty [32,33]. The P3b neurophysiological com- retical framework to understand the cognitive origin and ponent has been extensively studied in recent decades and implications of IAH in patients with T1D. shown to be closely associated with information processing, In conclusion, the findings of the present study suggest cognitive workload, and context update. Even though the that the presence of neural damage associated with IAH in neural source of both components is only partially under- T1D patients, as evidenced by neurophysiological alterations, stood, it is accepted that the P3a is more closely related to could act as novel biomarkers in this group of patients. neuronal populations in the prefrontal cortex as well as temporal-parietal regions while the P3b involves a complex circuitry of parietal and temporal regions, including hip- CRediT authorship contribution statement pocampal formation [35,47,48]. Nicole Stantonyonge: Conceptualization, Data curation, IAH is phenomenologically characterized by a lack of Funding acquisition, Investigation, Methodology, Writing- awareness or conscious detection of a set of symptoms. How- original draft, Writing-review and editing. Saul Martinez- ever, the extent to which IAH is due to the failure of endoge- Horta: Conceptualization, Data curation, Investigation, nous mechanisms of detection of inner relevant information Methodology, Writing-original draft, Writing-review and edit- is not clear [49]. Unfortunately, is not possible to determine ing. Frederic Sampedro: Investigation, Writing-review and the kind of causal relationship between IAH and the differ- editing. Marı́a Belén Sánchez-Saudinós: Investigation, ences observed in the P3a and P3b components. Therefore, Methodology, Writing-review and editing. Ana Chico: Concep- new studies are needed to explore in greater depth the extent tualization, Investigation, Supervision, Writing-review and to which the observed differences in P3a and P3b morphology editing. are a consequence of the accumulation of neuronal damage caused by recurrent SH episodes. It is not clear whether the origin of IAH is due to a failure of the involuntary and auto- Declaration of Competing Interest matic mechanisms of orientation of attention to relevant and novel inner information, or a combination of both fac- The authors declare that they have no known competing tors. In this regard, the lack of correlation between the differ- financial interests or personal relationships that could have ences in the ERPs and the number of SH episodes during the appeared to influence the work reported in this paper. last year and five years seems to suggest that the accumula- tion of SH episodes may not be wholly responsible for the pat- Acknowledgement tern of neural damage, which may or may not promote these differences. In turn, this suggests that the mechanisms that The authors acknowledge the editorial assistance of Bradley give rise to these differences in neurophysiological signals Londres from Londres Medical Writing, Editing and may already exist in patients with IAH prior to onset of these Translation. 8 diabetes research and clinical practice 177 (2021) 108898 Data availability Control in U.S. Adult Patients With Type 1 Diabetes: A Real- World Study. Diabetes Care 2019;42(12):2220–7. [11] Cryer PE. Diverse causes of hypoglycemia-associated The datasets generated during and/or analysed during the autonomic failure in diabetes. N Engl J Med 2004;350 current study are available from the corresponding author (22):2272–9. on reasonable request. [12] Geddes J, Schopman JE, Zammitt NN, Frier BM. Prevalence of impaired awareness of hypoglycaemia in adults with Type 1 Funding information diabetes. Diabet Med 2008;25(4):501–4. [13] Cukierman T, Gerstein HC, Williamson JD. Cognitive decline and dementia in diabetes–systematic overview of This study was supported by a grant from the SPANISH DIA- prospective observational studies. Diabetologia 2005;48 BETES SOCIETY. (12):2460–9. [14] Srikanth V, Sinclair AJ, Hill-Briggs F, Moran C, Biessels GJ. Compliance with ethical standards Type 2 diabetes and cognitive dysfunction-towards effective management of both comorbidities. Lancet Diabetes Ethical approval: All procedures performed in studies involv- Endocrinol 2020;8(6):535–45. [15] Moran C, Beare R, Phan T, et al. Neuroimaging and its ing human participants were in accordance with the ethical Relevance to Understanding Pathways Linking Diabetes and standards of the institutional research committee and with Cognitive Dysfunction. J Alzheimers Dis 2017;59(2):405–19. the 1964 Helsinki declaration and its later amendments or [16] Brands AM, Biessels GJ, de Haan EH, Kappelle LJ, Kessels RP. comparable ethical standards. The effects of type 1 diabetes on cognitive performance: a Statement of informed consent: Informed consent was meta-analysis. Diabetes Care 2005;28(3):726–35. obtained from all individual participants included in the [17] Lincoln NB, Faleiro RM, Kelly C, Kirk BA, Jeffcoate WJ. Effect of study. long-term glycemic control on cognitive function. Diabetes Care 1996;19(6):656–8. [18] Diabetes C, Complications Trial/Epidemiology of Diabetes I, Complications Study Research G, et al. Long-term effect of R E F E R E N C E S diabetes and its treatment on cognitive function. N Engl J Med. 2007;356(18):1842–1852. [19] Nunley KA, Rosano C, Ryan CM, et al. Clinically Relevant Cognitive Impairment in Middle-Aged Adults With [1] American Diabetes A. 12. Older Adults: Standards of Medical Childhood-Onset Type 1 Diabetes. Diabetes Care 2015;38 Care in Diabetes-2020. Diabetes Care. 2020;43(Suppl 1):S152– (9):1768–76. S162. [20] Tonoli C, Heyman E, Roelands B, et al. Type 1 diabetes- [2] Diabetes C, Complications Trial Research G, Nathan DM, et al. associated cognitive decline: a meta-analysis and update of The effect of intensive treatment of diabetes on the the current literature. J Diabetes 2014;6(6):499–513. development and progression of long-term complications in [21] Shalimova A, Graff B, Gasecki D, et al. Cognitive Dysfunction insulin-dependent diabetes mellitus. N Engl J Med 1993;329 in Type 1 Diabetes Mellitus. J Clin Endocrinol Metab 2019;104 (14):977–986. (6):2239–49. [3] American Diabetes A. 6. Glycemic Targets: Standards of [22] Auer RN. Hypoglycemic brain damage. Metab Brain Dis Medical Care in Diabetes-2021. Diabetes Care. 2021;44(Suppl 2004;19(3–4):169–75. 1):S73–S84. [23] Auer RN, Wieloch T, Olsson Y, Siesjo BK. The distribution of [4] Brito-Sanfiel M, Diago-Cabezudo J, Calderon A. Economic hypoglycemic brain damage. Acta Neuropathol 1984;64 impact of hypoglycemia on healthcare in Spain. Expert Rev (3):177–91. Pharmacoecon Outcomes Res 2010;10(6):649–60. [24] Chaytor NS, Barbosa-Leiker C, Ryan CM, Germine LT, Hirsch [5] McCoy RG, Van Houten HK, Ziegenfuss JY, Shah ND, Wermers IB, Weinstock RS. Clinically significant cognitive impairment RA, Smith SA. Increased mortality of patients with diabetes in older adults with type 1 diabetes. J Diabetes Complications reporting severe hypoglycemia. Diabetes Care 2012;35 2019;33(1):91–7. (9):1897–901. [25] Stantonyonge N, Sampedro F, Mendez J, Martinez-Horta S, [6] American Diabetes A. 6. Glycemic Targets: Standards of Chico A, Gomez-Anson B. Structural Grey and White Matter Medical Care in Diabetes-2020. Diabetes Care. 2020;43(Suppl Differences in Patients with Type 1 Diabetes and Impaired 1):S66–S76. Awareness of Hypoglycaemia. J Clin Endocrinol Metab 2020. [7] Gimenez M, Lopez JJ, Castell C, Conget I. Hypoglycaemia and [26] Languren G, Montiel T, Julio-Amilpas A, Massieu L. Neuronal cardiovascular disease in Type 1 Diabetes. Results from the damage and cognitive impairment associated with Catalan National Public Health registry on insulin pump hypoglycemia: An integrated view. Neurochem Int 2013;63 therapy. Diabetes Res Clin Prac 2012;96(2):e23–5. (4):331–43. [8] Gruden G, Barutta F, Chaturvedi N, et al. Severe hypoglycemia [27] Aoki T, Sato T, Hasegawa K, Ishizaki R, Saiki M. Reversible and cardiovascular disease incidence in type 1 diabetes: the hyperintensity lesion on diffusion-weighted MRI in EURODIAB Prospective Complications Study. Diabetes Care hypoglycemic coma. Neurology. 2004;63(2):392–3. 2012;35(7):1598–604. [28] Lee BW, Jin ES, Hwang HS, Yoo HJ, Jeong JH. A case of [9] Pedersen-Bjergaard U, Thorsteinsson B. Reporting Severe hypoglycemic brain injuries with cortical laminar necrosis. J Hypoglycemia in Type 1 Diabetes: Facts and Pitfalls. Curr Diab Korean Med Sci 2010;25(6):961–5. Rep 2017;17(12):131. [29] Kirchhoff BA, Lugar HM, Smith SE, et al. Hypoglycaemia- [10] Pettus JH, Zhou FL, Shepherd L, et al. Incidences of Severe induced changes in regional brain volume and memory Hypoglycemia and Diabetic Ketoacidosis and Prevalence of function. Diabet Med 2013;30(4):e151–6. Microvascular Complications Stratified by Age and Glycemic diabetes research and clinical practice 177 (2021) 108898 9 [30] TC H. Event-Related Potentials: A Methods Hand book. [42] Morris JC, Heyman A, Mohs RC, et al. The Consortium to Cambridge, MA: MIT Press; 2005. Establish a Registry for Alzheimer’s Disease (CERAD). Part I. [31] Courchesne E, Hillyard SA, Galambos R. Stimulus novelty, Clinical and neuropsychological assessment of Alzheimer’s task relevance and the visual evoked potential in man. disease. Neurology 1989;39(9):1159–65. Electroencephalogr Clin Neurophysiol 1975;39(2):131–43. [43] Crowe SF. The differential contribution of mental tracking, [32] Escera C, Alho K, Schroger E, Winkler I. Involuntary attention cognitive flexibility, visual search, and motor speed to and distractibility as evaluated with event-related brain performance on parts A and B of the Trail Making Test. J Clin potentials. Audiol Neurootol 2000;5(3–4):151–66. Psychol 1998;54(5):585–91. [33] Friedman D, Cycowicz YM, Gaeta H. The novelty P3: an event- [44] Lansing AE, Ivnik RJ, Cullum CM, Randolph C. An empirically related brain potential (ERP) sign of the brain’s evaluation of derived short form of the Boston naming test. Arch Clin novelty. Neurosci Biobehav Rev 2001;25(4):355–73. Neuropsychol 1999;14(6):481–7. [34] Hedges D, Janis R, Mickelson S, Keith C, Bennett D, Brown BL. [45] Pena-Casanova J, Blesa R, Aguilar M, et al. Spanish P300 Amplitude in Alzheimer’s Disease: A Meta-Analysis and Multicenter Normative Studies (NEURONORMA Project): Meta-Regression. Clin EEG Neurosci 2016;47(1):48–55. methods and sample characteristics. Arch Clin Neuropsychol [35] Polich J. Updating P300: an integrative theory of P3a and P3b. 2009;24(4):307–19. Clin Neurophysiol 2007;118(10):2128–48. [46] Lindvall O, Auer RN, Siesjo BK. Mechanisms of hypoglycemic [36] Solbakk AK, Reinvang I, Andersson S. Assessment of P3a and brain damage. Evidence against a significant role of the P3b after moderate to severe brain injury. Clin noradrenergic locus coeruleus system. Exp Brain Res 1988;73 Electroencephalogr 2002;33(3):102–10. (1):219–23. [37] Papaliagkas V, Kimiskidis V, Tsolaki M, Anogianakis G. [47] Soltani M, Knight RT. Neural origins of the P300. Crit Rev Usefulness of event-related potentials in the assessment of Neurobiol 2000;14(3–4):199–224. mild cognitive impairment. BMC Neurosci 2008;9:107. [48] Polich J, Criado JR. Neuropsychology and neuropharmacology [38] Papaliagkas VT, Kimiskidis VK, Tsolaki MN, Anogianakis G. of P3a and P3b. Int J Psychophysiol 2006;60(2):172–85. Cognitive event-related potentials: longitudinal changes in [49] Graveling AJ, Frier BM. Impaired awareness of mild cognitive impairment. Clin Neurophysiol 2011;122 hypoglycaemia: a review. Diabetes Metab 2010;36(Suppl 3): (7):1322–6. S64–74. [39] Clarke WL, Cox DJ, Gonder-Frederick LA, Julian D, Schlundt D, [50] Bruder GE, Kroppmann CJ, Kayser J, Stewart JW, McGrath PJ, Polonsky W. Reduced awareness of hypoglycemia in adults Tenke CE. Reduced brain responses to novel sounds in with IDDM. A prospective study of hypoglycemic frequency depression: P3 findings in a novelty oddball task. Psychiatry and associated symptoms. Diabetes Care 1995;18(4):517–22. Res 2009;170(2–3):218–23. [40] Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An [51] Klawohn J, Santopetro NJ, Meyer A, Hajcak G. Reduced P300 in inventory for measuring depression. Arch Gen Psychiatry. depression: Evidence from a flanker task and impact on ERN, 1961;4:561–71. CRN, and Pe. Psychophysiology 2020;57(4) e13520. [41] Folstein MF, Robins LN, Helzer JE. The Mini-Mental State Examination. Arch Gen Psychiatry 1983;40(7):812.