Single-ventricle CHD is the most critical subtype of CHD that requires multi-staged palliative surgeries with the culmination being the Fontan procedure. Reference Pike, Evangelista, Doering, Eastwood, Lewis and Child1 Although these surgeries are essential for long-term survival, these children face multiple comorbidities associated with chronic Fontan physiology (e.g., Fontan-associated liver disease and protein-losing enteropathy). Reference Pike, Roy and Moye2 Two common concerns in this population have been the ability to gain weight pre-Fontan and cognitive deficits. Approximately, 25% of children with single-ventricle CHD have known growth and nutritional deficits. Reference Power, Schultz and Dennis3 Factors contributing to these deficits include a high metabolic demand, inadequate caloric intake, altered intestinal pathology, and genetic and extracardiac abnormalities. Reference Medoff-Cooper and Ravishankar4–Reference Pike, Evangelista, Doering, Eastwood, Lewis and Child6
The cognitive deficits often become more apparent in adolescence where higher-level executive functioning skills are required. The aetiology of these cognitive deficits is unclear but presumed to be multifactorial (e.g., brain maturation, cyanosis, surgical, and perioperative factors). Reference Pike, Roy and Moye2,Reference Cabrera-Mino, Roy and Woo7 However, it is unknown if nutritional status (e.g., body mass index, pre-albumin, and thiamine) contributes to cognitive deficits in adolescents with single-ventricle CHD.
Thiamine, also known as vitamin B1, is a water-soluble vitamin found in meat, wholegrains, fish, and nuts and is necessary for the normal functioning of the nervous system. Symptoms of thiamine deficiency include fatigue, memory loss, depression, nausea, cardiomyopathy/heart failure (“wet beriberi”), and muscle cramps. Humans are dependent on dietary intake to fulfil their thiamine requirements as very little thiamine is stored in the body, and depletion can occur within 14 days. Thiamine deficiency has been reported in adult heart failure and was associated with loop diuretic use Reference Katta, Balla and Alpert8 and memory deficits in other clinical conditions. Reference Listabarth, Vyssoki and Marculescu9,Reference Qian, Zhao and Pan10 Thiamine deficiency was also noted in 27% of young children in heart failure on loop diuretics with a ventricular septal defect both before and after surgical closure. Reference Shamir, Dagan, Abramovitch, Abramovitch, Vidne and Dinari11 Although loop diuretics enhance the excretion of thiamine and may lead to its deficiency, low levels of thiamine can develop secondary to inadequate intake from the diet. Reference Katta, Balla and Alpert8,Reference Rahman, Jafry, Jeejeebhoy, Nagpal, Pisani and Agarwala12,Reference Attaluri, Castillo, Edriss and Nugent13
Pre-albumin is known to be a biomarker of protein nutritional status (e.g., malnutrition) with lower levels being associated with reduced survival and higher readmission rates in adult heart failure. Reference Lourenço, Silva and Friões14,Reference Murphy, Gray and Joyce15 There continues to be variability related to body mass index reported in the Fontan population with many identified as under to normal weight and felt to be associated with decreased intestinal perfusion and absorption of nutrients contributing to a cardiac cachexia clinical picture. Reference Pike, Evangelista, Doering, Eastwood, Lewis and Child1,Reference Cohen, Zak and Atz16–Reference Zaqout, Vandekerckhove and Michels18 Conversely, other studies have shown overweight and obesity secondary to decrease functional status related to worsening clinical condition or Fontan failure. Reference Yogeswaran, Anigwe and Blissett19–Reference Byrne, Weingarten and Clark21
The relationship of these nutritional biomarkers to cognitive function has not been assessed in adolescents with single-ventricle CHD post-Fontan procedure. The identification and treatment of dietary deficiencies could potentially enhance cognitive performance and ultimately self-care. Reference Attaluri, Castillo, Edriss and Nugent13 Therefore, the purpose of this study was to evaluate nutritional status (pre-albumin, body mass index, and thiamine) in adolescents with single-ventricle CHD post-Fontan and healthy controls to identify possible associations with impairment of cognitive function.
Materials and methods
Study design, sample, and setting
In this cross-sectional, comparative pilot study, 70 adolescents (40 single-ventricle CHD and 30 controls), aged 14 to 21 years, were recruited via flyers or provider referral from paediatric cardiology clinics in Southern California and healthy controls from the local community. The single-ventricle heart disease group included adolescents who had undergone surgical palliation with Fontan completion and were excluded for severe developmental delay that precluded the ability for self-report (e.g., cerebral palsy). Controls were excluded for any chronic medical or psychiatric condition.
Data collection
After Institutional Review Board approval from the University of California, Los Angeles (#16-000433), parental permission and assent were obtained for participants under the age of 18 years, and informed consent was obtained from participants aged 18 years and over. All participants completed a demographic form (e.g., age, gender, and ethnicity), food frequency questionnaire, the Montreal Cognitive Assessment, and a blood sample with laboratory analysis to assess pre-albumin and thiamine levels. All blood samples and laboratory analyses were performed at the same outpatient centre at UCLA. Clinical information was extracted from the medical records (e.g., type of single ventricle and Fontan procedure, number of surgeries, current oxygen saturation, height, weight, body mass index, and medications [including over-the-counter drugs and multivitamins]).
Measurements
Montreal Cognitive Assessment
The Montreal Cognitive Assessment is an administered screener used to measure multiple domains of cognitive function, such as visuospatial, attention/concentration, executive function, language, delayed memory recall, and naming. Reference Nasreddine, Phillips and Bedirian22 The visuospatial and executive function tasks are written items (e.g., three-dimensional cube, Watson Clock Drawing Task, and alternating Trail Making Task Part B), and the remaining are scored based on verbal response. Reference Nasreddine, Phillips and Bedirian22 The total score ranges from 0 to 30, and a score < 26 is considered abnormal. The Montreal Cognitive Assessment has been validated in adolescents both with and without CHD and has a Cronbach’s alpha of 0.8. Reference Pike, Poulsen and Woo23
Body mass index
All participants’ height and weight were obtained in order to calculate their body mass index value. Body mass index was assessed based on the weight-for-age and gender growth charts. 24 (Center for Disease Control and Prevention [CDCP], 2017).
Pre-albumin and thiamine laboratory tests
Venous blood was collected at the same lab for all participants and assayed. The adolescent normal values for thiamine (normal 70–110 nmol/L) and pre-albumin levels (normal 23–45 mg/dL)] were used based on the UCLA outpatient laboratory reference ranges. Participants were notified if levels were below the normal range.
Thiamine Food Frequency Questionnaire
The Thiamine Food Frequency Questionnaire was developed by the investigator due to a lack of instruments available that assess the intake of foods rich in thiamine and fast-food consumption. This was developed based on the validated Food Frequency Index which captures dietary patterns and not portions consumed, a Likert scale format, and relied on a 7–30-day recall period Reference Freisling, Elmadfa, Schuh and Wagner25 . Since thiamine has a short half-life, the questionnaire was based on foods consumed in a given week. A seven-item, questionnaire was used to assess how often vegetables and fruits, breads and grains, meats, nuts and seeds, dairy, beans and legumes, and fast foods were consumed in a given week. Response options were never, once a week, 1–3 times a week, 4–5 times a week, once a day, or twice or more times a day with corresponding scores of 0–5. A single-item question was asked regarding intake of multivitamin use. The Thiamine Food Frequency Questionnaire was pilot-tested in three adolescents to assess face validity and a nutritional expert for content validity. Total scores ranged from 0 to 36 with higher scores indicating increased dietary intake of thiamine.
Statistical analyses
Sample characteristics are presented as means with standard deviations or medians with interquartile ranges for continuous variables and Chi-square for categorical variables. The variables were examined for normality (thiamine and pre-albumin normally distributed, and cognitive scores not normally distributed) by the Shapiro–Wilks tests. Thus, non-parametric statistics (Mann–Whitney U and Kruskal–Wallis tests) were used to assess group differences (single-ventricle CHD versus control), Montreal Cognitive Assessment group (low versus high scores), compared to body mass index (Center for Disease Control and Prevention guidelines – underweight, normal, and overweight), thiamine, and pre-albumin levels. Spearman’s rho correlation coefficients were used to assess variables associated with total Montreal Cognitive Assessment scores. Only variables with significant correlations (p = 0.05) were entered into the multivariate analysis using linear regression to assess predictors of cognitive scores. All analyses were conducted with the Statistical Package for the Social Sciences version 23.0 (IBM; Somers, NY) with a significance set at a p-value < 0.05.
Results
Sample characteristics between groups
A total of 70 adolescents, 40 single-ventricle CHD (mean age 16 ± 1.6, female 51%, Hispanic 44%, hypoplastic left heart syndrome 26%) and 30 controls (mean age 16.8 ± 1.9, female 52%, Hispanic 66%), were studied. Five were excluded from the analysis secondary to lab processing errors or inability to obtain blood samples after the second venipuncture attempt. No significant differences in age, gender, or ethnicity appeared between groups. Body mass index was higher in the controls compared to single-ventricle CHD group (24 versus 20; p = 0.005) (Table 1).
Demographic, cognitive, and laboratory results between groups (n = 70).
MoCA = Montreal Cognitive Assessment; N/A= not applicable; TFFQ = Thiamine Food Frequency Questionnaire; SD = standard deviation; SVHD = single-ventricle heart disease.
*Mann–Whitney U-test.
Thiamine, pre-albumin, body mass index, dietary intake, and cognitive scores between groups
Single-ventricle CHD group had lower median total Montreal Cognitive Assessment scores (23 versus 29, p < 0.001) including subscales except naming and pre-albumin levels (23 versus 27, p = 0.013) than controls. Single-ventricle CHD group had higher thiamine levels than controls (127 versus 103, p = 0.033) and a higher intake of multivitamins (Table 1). A comparison of dietary intake between groups using the Thiamine Food Frequency Questionnaire showed no statistically significant difference in individual food categories except for a higher intake of fast food in the control group (p = 0.049).
Cognitive scores associated with thiamine, pre-albumin, and body mass index
Lower pre-albumin levels (< 23) and underweight body mass index were associated with abnormal total Montreal Cognitive Assessment scores (p = 0.030) in both single-ventricle CHD and control groups (Table 2). There were no significant differences between groups with low pre-albumin levels and cognitive scores (p = 0.07). Pre-albumin (r = .262*; p = .037) and body mass index (r = .264*; p = .030) but not thiamine (r = −0.009, p = .256) were correlated with Montreal Cognitive Assessment scores. Of note, pre-albumin was correlated with visuospatial/executive function (r = .263*; p = .036) and attention (r = .261*; p = .025) subscales of the Montreal Cognitive Assessment. Lastly, low pre-albumin level was the only independent predictor of worse cognition (β =0.260, R2 = 0.67, adjusted R2 = 0.052, F = 4.480; p = 0.038).
Cognitive scores associated with thiamine, pre-albumin, and BMI.
BMI = body mass index.
Mann–Whitney U-test and Kruskal–Wallis test.
Discussion
Thiamine levels in adolescence with and without single-ventricle heart disease
Our study findings identified that thiamine was not deficient in adolescents with single-ventricle CHD and were higher than in controls. Of interest, the majority of the single-ventricle heart disease group were taking a daily multivitamin (usually 50–100 mg of thiamine) compared to only a few in the control group. Dietary intake of foods high in thiamine was not different between groups except the control group that had a higher intake of fast foods than the single-ventricle CHD group. The general adolescent diet is often high in processed or ultra-processed foods (e.g., fast food), which can lower micronutrient intake, in particular thiamine, and has been shown to impact cognitive function in adolescents. Reference Askari, Abbaszadeh and Saharkhiz26 This may be reflective of the lower thiamine levels in the control group.
Vigilant parenting behaviours and few adolescents on diuretic therapy may explain the higher thiamine levels seen in the single-ventricle CHD group compared to controls. Parents of children with severe CHD tend to be very observant and have learned to closely monitor dietary intake and weight gain to promote their child’s well-being. Reference Meakins, Ray, Hegadoren, Rogers and Rempel27–Reference Carey, Nicholson and Fox29 Consequently, these vigilant behaviours may continue into adolescence and young adulthood and be manifested by daily multivitamin intake and decreased fast-food consumption. Furthermore, only a few adolescents were on diuretic therapy greater than 10 years out from the Fontan procedure. Studies in adult heart failure acknowledge the excretory influence on thiamine levels which maintained on high-dose diuretics. Reference Katta, Balla and Alpert8 Despite higher thiamine levels in the single-ventricle CHD group, cognitive scores were worse compared to controls indicating other factors affecting cognitive function. In addition, there was no difference between groups with low pre-albumin and cognitive scores indicating the possibility for nutrition-induced changes.
Body mass index and pre-albumin related to cognition
Our findings identified lower body mass index and pre-albumin levels to be associated with lower cognitive scores. Unfortunately, there are no studies evaluating the association between cognitive scores with body mass index and/or pre-albumin levels in adolescents with CHD or specifically single-ventricle CHD. These finding may reflect their clinical condition or disease severity. Low serum albumin levels related to enteric protein loss are not uncommon post-Fontan. Reference Patel, Loomes, Goldberg, Mercer-Rosa, Dodds and Rychik30,Reference Bernardi, Moses, Barber, Witte and Seckeler31 Prolonged Fontan physiology results in decreased cardiac output and increased central venous pressures, which changes mesenteric blood flow. High central venous pressures post-Fontan can alter protein absorption and compromise weight gain (not fluid secondary to ascites or peripheral oedema) in patients with protein-losing enteropathy. Albumin and pre-albumin can also be low in patients with an acute or chronic inflammatory illness or conditions. A low-flow state of chronic heart failure results in the stimulation of the inflammatory system, which appears to be similar post-Fontan. Reference Ostrow, Freeze and Rychik32 Studies have shown that 30% of patients post-Fontan with protein-losing enteropathy have elevated inflammatory biomarkers (e.g., C-reactive protein) in the presence of low albumin. Reference Ostrow, Freeze and Rychik32 Despite only four adolescents in our study with documented protein-losing enteropathy, low normal pre-albumin levels may be inherent in the condition and reflective in our findings.
Poor linear growth in the face of high caloric intake is well documented in children pre-Fontan Reference Mancilla, Zielonka and Roizen33 with some identifying an association with neurodevelopment and cognitive deficits. Reference Medoff-Cooper and Ravishankar4 Our study identified executive function, visuospatial, memory, attention, and concentration deficits that are congruent with many studies assessing neurocognitive outcomes in this population. Reference Pike, Roy and Moye2,Reference Bellinger, Watson and Rivkin34,Reference Cassidy, White, DeMaso, Newburger and Bellinger35 However, these studies did not examine specific nutritional biomarkers or body mass index related to cognitive outcomes. Although the literature on low versus high body mass index and cognitive performance in children remains controversial, our findings are plausible based on established links between nutritional deficiencies and compromised cognitive development. Reference Engle and Fernàndez36,Reference Benton37 Sarcopenia associated with lean muscle mass deficits as in cardiac cachexia has been reported in Fontan patients and felt to be similar to patients with advanced heart failure from acquired heart disease. Reference Greutmann, Le and Tobler38,Reference Ouimet-Grennan, Guerrero-Chalela and Therrien39 Conversely, increased body mass index in adult Fontan patients has been associated with symptomatic heart failure or Fontan failure. Reference Martinez, Byku and Novak20,Reference Byrne, Weingarten and Clark21 Thus, our Fontan population is younger in age with few on diuretics or protein-losing enteropathy, which may reflect a more stable chronic condition. Future studies are needed to confirm these findings and investigate other potential contributing factors such as reduced cerebral blood flow associated with lower cardiac output with ageing and its effect on cognition. Furthermore, future studies are needed to assess if an increase in protein intake (preferably in liquid form for easier absorption) could improve weight gain and cognitive performance for selected patients.
Limitations
Our study should be viewed in light of some limitations. The sample size was relatively small which limits our ability to perform additional analyses based on diuretic and multivitamin use. We also did not anticipate the number in the single-ventricle CHD group who were taking daily multivitamins and a healthier diet and not taking diuretics. Furthermore, using the Thiamine Food Frequency Questionnaire which was adapted from a known validated instrument to assess dietary intake, we modified the instrument for brevity to assess dietary consumption of foods high in thiamine, and this altered the validity of the instrument. Finally, the cross-sectional nature of the study limits our ability to establish any causal relationship.
Conclusion
Lower pre-albumin was an independent predictor for worse cognition in adolescents with single-ventricle heart disease and controls. Lower pre-albumin levels may reflect chronic liver changes or protein-losing enteropathy seen in Fontan physiology. These findings highlight the possibility for nutrition-induced cognitive changes. Improved nutritional status, in particular protein intake, in adolescents with single-ventricle CHD should be assessed as a potential treatment to improve cognitive performance.
Acknowledgements
The authors gratefully acknowledge the assistance of Patty Chung and Paola Moreno for assistance with data collection.
Financial support
This work was supported by the National Institutes of Health grant R01NR016463 and University of California, Los Angeles, School of Nursing Funding.
Competing interests
None.
Ethical standard
This research did not involve human or animal experimentation. Approval was obtained from the institutional review boards.
Open access
