According to the World Health Organization, 45% of deaths among children under 5 years of age are caused by malnutrition, which impacts more than 224 million children globally and makes this condition a critical global health concern (UNICEF et al., 2021). The situation will likely be exacerbated by the COVID-19 pandemic according to UNICEF (2021) due to deteriorations in household wealth and disruptions to the availability and affordability of nutritious food and essential nutrition services. Protein-energy malnutrition (PEM) or severe acute malnutrition (SAM) is a specific type of malnutrition defined as an acute caloric deficit due to deficiency of all macronutrients, and micronutrients in some cases (see Galler et al., 2021 for a review of the definitions).

While malnutrition has deleterious effects on cognitive abilities (Champakam et al., 1968; Galler et al., 1983b,1984a; Upadhyay et al., 1989a; Mendez and Adair, 1999; Berkman et al., 2002), it has also been associated with various effects on motor skills, social abilities and mental health (Galler et al., 1983a,1984b,1985; Upadhyay et al., 1989b), with many of these effects persisting into adolescence and even adulthood (Hoorweg and Stanfield, 1976; Walker et al., 2007; Galler et al., 2012b,c; Waber et al., 2014a,b). Using electroencephalography (EEG), brain function alterations have also been linked to early childhood malnutrition (for a review see Gladstone et al., 2014). Although these findings were replicated in several studies (Stoch et al., 1963; Baraitser and Evans, 1969; Griesel et al., 1990), few of them assessed the children at later timepoints in childhood. Bartel et al. (1979) studied 6- to 12-year-old children hospitalized for the treatment of kwashiorkor during the first 27 months of life and confirmed that their rest EEGs still exhibited significantly diminished alpha and increased theta and delta frequencies compared to a control group. It is clear from these behavioral and EEG findings that malnutrition at a young age has important acute deleterious effects on early brain function, cognition and behavior. However, the effects of early childhood malnutrition on brain function in adulthood are still unknown. Moreover, to our knowledge, no study has investigated brain function during a cognitive task.

The Barbados Nutrition Study (BNS) is a 55-year longitudinal study that has followed a Barbadian cohort hospitalized during the first year of life for a single episode of moderate to severe protein-energy malnutrition (PEM; Galler et al., 1983a,1987a) based on the Gomez classification (Gomez et al., 1955), and matched healthy controls who were former classmates of the PEM participants. The objectives of this study are to characterize the medical, neuropsychological, behavioral and brain effects of early PEM over the lifespan. The cohort comprises participants that were originally recruited between 1967 and 1972 when they were hospitalized as infants and subsequently enrolled in a government program that provided health and nutrition monitoring until the children reached 12 years of age. The malnutrition episode was restricted to the first year of life and all children in the PEM group achieved complete catch-up in physical growth by adolescence (Galler et al., 1987a,b,c). Control participants were classmates of the PEM children and were matched for age, sex, and handedness. Neuropsychological and psychiatric assessments revealed many cognitive, behavioral and mental health impairments associated with early childhood PEM, including lower IQ, conduct problems and higher prevalence of affective and depressive symptoms, as well as attention deficits (Galler et al., 1983a,b, 2010, 2012a). Additionally, most participants in the PEM group exhibit persistent attention and executive problems during childhood, adolescence and adulthood compared to the control group (Galler et al., 1983b,1990, 2012b; Waber et al., 2014a). In a recent study, brain function alterations were reported in the resting state activity of previously malnourished 5–11 years old children from the BNS cohort: specifically, an excess of theta, alpha2 and beta frequencies, and a decrease of alpha1, all suggesting a maturational lag in cortical development despite complete nutritional rehabilitation and demonstrated catch-up growth (Taboada-Crispi et al., 2018). However, little is known about the long-term effects of early childhood malnutrition on brain function in adults from this cohort.

Our aim was to study the persistence of brain function alterations in adults who experienced moderate to severe PEM during the first year of life by comparing the brain activity in adults from the PEM and the control groups of the BNS cohort using Evoked-Related Potentials (ERPs). Considering the attention and executive impairments previously reported in the PEM group, we administered a Go-No-Go attention task during EEG recordings to specifically isolate those altered processes. In a Go-No-Go task, two ERPs components are typically studied, namely the N2 and P3 components. N2 is defined as the largest fronto-central negative polarity deflection between 200 and 450 ms, and P3 is defined as the largest fronto-central positive polarity deflection between 350 and 600 ms (McLoughlin et al., 2010; Woltering et al., 2013; Grane et al., 2016; Schmüser et al., 2016). N2 and P3 are respectively, associated with conflict monitoring (i.e., conflict between the prepotent response and required response) and inhibition response (i.e., cancelation of a planned or prepotent response) processes in an attention/inhibition control task such as Go-No-Go, Stop signal, Stroop, or Flanker task (Braver et al., 2001; Smith et al., 2008). Those components have been widely studied and are thought to be altered in populations with attention and executive impairments like ADHD (McLoughlin et al., 2010; Johnstone et al., 2013; Woltering et al., 2013; Grane et al., 2016; Schmüser et al., 2016) and schizophrenia (Doege et al., 2010; Araki et al., 2016). In the current study, we hypothesized that the PEM group would show smaller N2 and P3 peak amplitudes absolute values compared to the healthy control group. This is the first study to investigate ERP adult brain function following early childhood malnutrition. It is also the first to examine the impact of childhood malnutrition on brain function during a cognitive task.

The main objective of this study was to compare the brain activity of adults who experienced moderate to severe PEM during the first year of life and healthy controls with no histories of malnutrition using a Go-No-Go inhibition task. We hypothesized that the PEM group would show altered neural response associated with attention and inhibition (N2 and P3 components) during the task. The hypothesis was partially confirmed, since we observed a reduction of N2 amplitude during the No-Go condition in the PEM group compared to the control group, but there was no difference detected in P3. Overall, results of the Go-No-Go task revealed a main effect of Condition for both components (N2 and P3) in both PEM and control groups. This is a typical result for a Go-No-Go task, indicating that the No-Go condition induces a genuine inhibition response which delays the onset and amplifies the N2 and P3 components (Eimer, 1993; Bokura et al., 2001; Schmüser et al., 2016).

The N2 amplitude was smaller in the PEM group compared to the control group during the No-Go condition. This result has also been found in adults and children with ADHD (Brandeis et al., 2002; Fallgatter et al., 2004; Gow et al., 2012; Woltering et al., 2013). Of note, attentional deficits have also been reported in the PEM participants in earlier studies (Galler et al., 1983a,1990, 2012b; Waber et al., 2014a). There is debate amongst the scientific community on the cognitive process underlying the N2 component. Although some authors argue that N2 is related to the inhibition process (i.e., cancelation of a planned or prepotent response; Eimer, 1993; Falkenstein et al., 1999), more recent studies point to an association between N2 and conflict monitoring (i.e., conflict between the prepotent response and required response; Donkers and van Boxtel, 2004; Enriquez-Geppert et al., 2010; Huster et al., 2013; Groom and Cragg, 2015). According to the conflict monitoring hypothesis, an inhibition task such as Go-No-Go should evoke an N2 component because of the unbalanced ratio between Go and No-Go trials, which would lead to the creation of a prepotent response (Go) that conflicts with the infrequent inhibition of this response (No-Go) and not because of the inhibition process per se (Braver et al., 2001). However, conflict monitoring is closely related to attention, since it is responsible for triggering cognitive control changes by adjusting attention levels to optimize performance and prevent subsequent conflict (Botvinick et al., 2001). According to these models, our results can be interpreted as showing impairment in conflict monitoring and/or attention following early childhood malnutrition.

Surprisingly, the ERP results showed no difference between the two nutrition groups in P3 amplitude or latency. P3 component is usually considered to be a marker of response inhibition processing and evaluation (Huster et al., 2013; Groom and Cragg, 2015). In studies of ADHD using a Go-No-Go task, both components, N2 and P3 are typically altered (Johnstone et al., 2013; Woltering et al., 2013). However, the reduced N2 amplitude in the PEM group and similar P3 amplitude in the PEM and control groups is in line with the behavioral results. Indeed, in the current study, Go accuracy was significantly lower in the PEM group compared to the control group, revealing that adults who suffered from early childhood malnutrition committed more omission errors than controls. No difference in commission error rate was found. In this respect, it was previously reported in this cohort that inattention symptoms were more prevalent in the PEM group than impulsivity, and that omission errors differentiated the nutrition groups better than commission errors on the Connors Continuous Performance Task-II (Galler et al., 2012b). Omission errors are usually attributed to impairments in attention and vigilance whereas commission errors are associated with inhibition deficits. Therefore, early childhood malnutrition may be more closely associated with diminished attention and vigilance. Furthermore, no differences in reaction times were found between our groups. Slower reaction times have been consistently reported in populations with attentional and inhibition difficulties and are associated with hyperactivity and impulsivity (Wiersema et al., 2006; Rubia et al., 2007; Barkley et al., 2008; McLoughlin et al., 2010; Tamm et al., 2012; Schmüser et al., 2016).

The behavioral results are in line with the persistent attention deficits previously reported in our cohort during childhood, adolescence and adulthood (Galler et al., 1983a,1990, 2012b; Waber et al., 2014a). Inattention has also been reported in other studies assessing the neurocognitive profile of malnourished populations (Richardson et al., 1972) and following famine exposure in early childhood (Wang et al., 2016).

The altered electrophysiological marker found in this study is also coherent with the literature on brain function effects following childhood malnutrition. Indeed, several studies show that childhood malnutrition is associated with the slowing of the EEG’s dominant rhythm in infancy (Stoch et al., 1963; Baraitser and Evans, 1969; Griesel et al., 1990; Gladstone et al., 2014) and electrophysiological alterations that persist in childhood even after nutritional rehabilitation (Bartel et al., 1979; Taboada-Crispi et al., 2018). Recently, our colleagues also found abnormal developmental resting-state EEG source activity when comparing childhood and adulthood EEGs in the BNS cohort, with PEM group showing an increase in low alpha, a decrease in high alpha and no changes in beta activity whereas control group showed a slight decrease in low alpha, an increase in high alpha and a decrease in beta activity (Bosch-Bayard et al., 2022). More specifically in ERP studies, ERP abnormalities were also found in concurrently malnourished infants with a higher relative abnormality of their auditory evoked potentials compared to controls (Barnet et al., 1978; McDonald et al., 2007). Long-lasting alterations in ERP components during a recognition memory task were also found in children 10 years after iron deficiency treatment (Congdon et al., 2012). We extend knowledge in this field by showing brain function deficits still perceptible in adulthood.

Experimental studies further confirm the findings of the current study. Prenatally malnourished rats also show a significant impairment in executive function and attentional tasks (McGaughy et al., 2014; Newman et al., 2019). Further, using 2-deoxyglucose to measure neural activity during an attentional task, it has been shown that malnourished and control rats recruit different brain pathways (Rushmore et al., 2021). Whereas prefrontal activity was a predictor of task performance in controls, malnourished rats relied on brain networks involving limbic structures such as the hippocampal formation, reflecting compensatory brain responses to malnutrition. As in the current study, animal research has also shown alterations in brain function when performing attentional and executive tasks.

The analyses presented in this study adjusted for childhood socioeconomic and home environmental conditions of the participants. Indeed, given the known impact of the SES on brain function (Farah, 2017; Yaple and Yu, 2020), it was important to control for the initial childhood socioeconomic differences between our two groups. Although SES effects did not modify the significant effects of early malnutrition on executive function in the BNS cohort (Galler et al., 2012b; Waber et al., 2014a), lower SES has been associated with impaired performance in executive functions and attention shifting tasks (Lawson et al., 2018) as well as differences in ERP components during attentional tasks (St. John et al., 2019; Hiran Perera et al., 2021). Thus, by controlling for the SES, the presented results suggest that malnutrition limited to the first year of life is associated with abnormalities in adult brain function that go beyond what could be attributable to socioeconomic factors.

This study has several limitations. First, the small sample size does not allow us to conclude with certainty that the results were not due to a lack of statistical power. Nevertheless, the effect size is moderate for the N2 amplitude difference between the groups (d = 0.55). Additionally, the results reported here are specific to the N2 and P3 electrophysiological markers and different analyses might reveal other electrophysiological differences between our groups. For example, the use of source localization analysis on our data would allow for a more precise exploration at the source level of the temporal dynamics of brain functioning in executive function. Also, spectral analysis or functional connectivity analysis could reveal interesting information on the brain functional organization. Correlations between physiological data and EEG data could be relevant as well, for instance to study the relationship between heart rate variability and cognitive decline (Forte et al., 2019). Further analyses will be performed in a near future to better characterize and compare brain functions in the two groups. We also plan on comparing our task-based EEG outcomes with the resting-state EEG outcomes found in this cohort by our collaborators (Taboada-Crispi et al., 2018; Bosch-Bayard et al., 2022).

In sum, this study shows that adults who experienced early childhood malnutrition limited to the first year of life have different brain response patterns during a response inhibition task compared to healthy controls. This adds to the existing literature on cognitive and neural outcomes following early childhood malnutrition, suggesting that attention and conflict monitoring, two cognitive control processes, remain altered in adulthood. Considering the impact of persisting cognitive alterations on the quality of life, more research is needed to better characterize early brain markers and clinical profiles associated with early childhood malnutrition in order to develop a disease progression model applicable to vulnerable populations.

The data that supports the findings of this study is available from the corresponding author AG, anne.gallagher@umontreal.ca upon reasonable request.

The studies involving human participants were reviewed and approved by Massachusetts General Hospital’s IRB, Sainte-Justine University Hospital’s Ethic Committee, Centro de Neurociencias de Cuba’s Ethic Committee, and the Barbados Ministry of Health. The patients/participants provided their written informed consent to participate in this study.

KR acquired the data, performed the analysis, wrote the original draft, and revised the manuscript. PV participated in data acquisition and data analysis and revised the manuscript. JT provided help in data analysis and revised the manuscript. MB participated in the conceptualization of the study and revised the manuscript. CB coordinated the data acquisition. AR helped the coordination of the data acquisition and revised the manuscript. PV-S and AG conceptualized the study, provided supervision, and revised the manuscript. JG conceptualized the study, coordinated the data acquisition, provided supervision, and revised the manuscript. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.