Assessment of dietary intake of children with chronic kidney disease

It is well recognized that nutritional monitoring and intervention is indispensable in pediatric CKD management, yet prospectively collected data in this area are scarce. CKiD is one of the largest cohorts of pediatric CKD patients, and the prospectively collected dietary data permits an analysis of the impact of dietary intake on CKD progression and the development of comorbidities. Here we describe our baseline data and compare them with the recommended nutritional levels from the KDOQI and KDIGO guidelines.

CKiD participants consumed much more dietary sodium than the recommended limits. The excessive intake of sodium was most dramatic in the oldest age group. More than 25 % of adolescents aged 14–18 years consumed >5150 mg of sodium daily, which is more than double the top end of the recommended range. Interestingly, sodium intake was similar to data obtained from the general population. Median daily sodium intake reported in the National Health and Nutrition Examination Survey (NHANES) 2011–2012 among those aged 2–5 years, 6–11 years, and 12–19 years were 2295 mg, 3081 mg, and 3593 mg, respectively [14], which is similar to the corresponding levels of 2222 mg, 3028 mg, and 3577 mg among CKiD participants. A significant portion of pediatric CKD diagnoses comprises conditions under the category of congenital anomalies of kidney and urinary tract, which are commonly salt wasting and frequently require sodium supplementation. We found that children with a salt-wasting condition consumed slightly more salt than those without the condition, although this difference was not significant statistically. An analysis of dietary sodium consumption data and its relationship with blood pressure in the CKiD cohort is currently underway. Our results indicate that greater effort is required to reduce dietary sodium consumption in subgroups of children with CKD—such as those with the non-salt-wasting condition or with hypertension—given its potential adverse effects. Most dietary sodium comes from added salt during food processing or cooking [3]. Therefore, behavioral strategies like avoiding processed food and reading food labels should be encouraged. In fact, the National Kidney Foundation provides practical suggestions on how to reduce dietary sodium intake [34].

Dietary potassium consumption was associated with both GFR and age. That daily potassium intake decreased as CKD advanced was not surprising, as children with lower GFR may have reduced appetite and may have received more intense nutritional counseling and intervention, both of which can lead to a decrease of potassium intake. Unlike sodium, data relating dietary potassium intake and outcome among CKD populations is limited. A study investigating dietary potassium intake among adult hemodialysis patients demonstrated that patients in the highest quartile of potassium intake have the highest risk of mortality (adjusted hazard ratio 2.4, with 95 % confidence interval 1.1–7.5) compared with those in the lowest quartile [34]. Although restriction of potassium intake is recommended for children with CKD stages 2–5 and 5D who are at risk of hyperkalemia [1], there is no specific guideline to suggest potassium intake levels for children of various ages and CKD staging who require restriction. Fruits and vegetables are the main sources of dietary potassium [35], and cautious consumption of these food items is recommended for those who require potassium restriction.

Hyperphosphatemia is associated with significant morbidity and mortality in CKD patients [36, 37]. Although a clear association between dietary phosphorus intake and serum phosphorus level has not yet been demonstrated among early-stage CKD patients [38], dietary phosphorus limitation is still commonly advised among selected subgroups of patients. Our result of high phosphorus intake may be partly explained by the high protein intake in our cohort, especially among younger kids, as the daily protein intake level far exceeded the recommended level in all age groups, with the most dramatic difference observed among the youngest participants. This can also be reflected by the finding that no patient in the younger age groups met the suggested phosphorus consumption levels. Nevertheless, protein levels were still lower than in the general population [39]. This also points out the difficulty of achieving a nutritional balance in clinical management of young children with CKD. Food items rich in protein contain high phosphorus load, and protein restriction is usually not recommended for children with CKD in order to promote growth. While on one hand avoiding protein energy wasting is advocated, on the other hand, excessive dietary protein intake may contribute to CKD progression and impact the control of metabolic bone disease [36].

Children in the CKiD cohort consumed more than the recommended caloric intake. Of note, young children with CKD, especially those younger than 2 years, are vulnerable to malnutrition and often require supplementary enteral feeding (such as tube feeding or gastrotomy feeding) for preservation of normal growth [1, 7]. Yet, excessive caloric consumption occurred in all age groups. When we calculated and compared our data to figures obtained from the general population, we found that the daily caloric intake levels were similar to their peers in the US, as reported in NHANES [14]. NHANES data was obtained from the community-based general population and do not represent children with CKD. Therefore, our report represents a new finding in the CKD population. NHANES used one-day dietary recall interviews conducted by trained interviewers to derive nutritional intake data [14], which differs from data we obtained from FFQ in CKiD. Therefore, results may not be directly comparable. Despite that, our study emphasizes the common problem of excessive dietary caloric intake and obesity in the pediatric population with CKD, and the risk factors of obesity compounds the increased risk of accelerated CKD progression.

Macronutrient distribution in this study was consistent with the recommended levels in all age groups, in which carbohydrates were the main source of energy, followed by fat and protein. Although we did not specifically analyze the components of each macronutrient, such as distribution of dietary cholesterol, trans fatty acids, saturated fatty acids, or types of carbohydrates, results still indicate a reassuring picture in terms of energy distribution.

There are several limitations of our study. First, we recognize there are inherent drawbacks of employing an FFQ in assessing nutrient intake, which include the lack of a direct and quantitative method to assess nutrient consumption, inadequate coverage of all food items that an individual may consume, and an imprecise categorization of food items causing under- or overestimation of nutrients consumed [40]. Second, the FFQ used in this study was not validated in every age or ethnic group. Despite these issues, the large number of systematically collected FFQs with the standardized method of analysis employed in the CKiD cohort provides valuable information. Although misclassification is possible, it is likely to occur randomly, though systematic under- or overreporting of some food items is possible. Third, we used observed body weight to determine our estimated protein intake, whereas the reference values adopted from KDOQI guidelines suggest using ideal body weight instead [1]. While the use of ideal body weight can overcome the problem of overestimating protein intake in obese children and will improve the accuracy of protein intake estimation, there is currently no consensus on how to determine ideal body weight [41]. Prior to this report, few rigorous assessments of nutritional intake in a large nationally representative cohort of children with CKD have been published. We tried to minimize the inclusion of FFQs with implausible responses by setting strict exclusion criteria. Fourth, as supplementation may have significant impact on nutritional intake in this group of participants, we performed a sensitivity analysis by excluding the 27 patients who consumed nutritional supplements. Results were essentially unchanged. Finally, our study is a cross-sectional analysis that characterizes nutritional intake at study entry. We cannot infer that decline in GFR or increase in age causes the changes we observed in nutritional intake. We did not assess how nutritional intake changes over time as kidney disease progressed or how nutritional intake impacted disease progression and the development of various comorbidities. Those issues will require further analysis of longitudinal data from follow-up in the CKiD. Regardless, these data provide a unique insight into nutritional intake and potential points of modification to improve health outcomes for children with CKD.