Zinc Status and Growth in Infants and Young Children with Cystic Fibrosis

Abstract

Background:

Zinc deficiency is associated with poor growth in children without cystic fibrosis (CF), but its impact on growth in children with CF is unknown.

Objective:

To determine the prevalence of low serum Zn (sZn) and its relationship with growth in the first three years of life in children with CF.

Methods:

We utilized data from infants with CF who were enrolled in a longitudinal study of nutrition and lung health and had sZn measured as part of clinical care. Cross-sectional correlations between sZn levels and growth z-scores were assessed by Pearson’s correlation coefficient. To identify factors associated with sZn status and its association to longitudinal growth patterns, multiple regression analysis with repeated measures were performed using generalized estimating equations.

Results:

A total of 106 sZn measurements from 53 infants were identified. Seventeen infants (32%) had intermittent Zn insufficiency, defined as at least one sZn <70 mcg/dL in their first 3 years of life. There were no significant cross-sectional associations between sZn and growth z-scores. However, analysis of longitudinal growth patterns revealed that weight- and length-for-age z-scores in children with intermittent Zn insufficiency were lower during early infancy and their weight-for-length z-scores at age 3 years were also lower compared to those who were always Zn sufficient.

Conclusion:

Low sZn occurs in one-third of children with CF in the first 3 years of life. Cross-sectional and longitudinal analyses revealed discrepant associations between sZn and growth. Therefore, prospective studies are needed to understand the role of Zn in growth in CF.

INTRODUCTION

Cystic fibrosis (CF) is an autosomal recessive genetic disease caused by mutations in the CF transmembrane conductance regulator (CFTR) gene. Respiratory failure due to CF lung disease is the overwhelming cause of death in CF. Several studies have demonstrated a significant association between nutritional status in infants and young children and lung function later in life^[1–5]. Identifying modifiable risk factors for growth failure is important for preserving future lung function.

Zinc (Zn), an essential mineral that is critically important for many metabolic pathways and enzymatic processes in the human body, is one such potentially modifiable risk factor. Zn deficiency is associated with decreased growth velocity and growth failure in children without CF^[6]. Infants with CF are at high risk for Zn deficiency due to impaired Zn homeostasis and the increased Zn requirements associated with infancy and early childhood^[7–10]. There have been few studies examining Zn status and growth in CF, and none in children <3 years of age, the period during which growth and nutrition are most critical for long-term outcomes^{[1–3, 5]}. A prior study demonstrated that approximately 30% of infants diagnosed with CF by newborn screening (NBS) had Zn concentrations in the deficient range defined as <60ug/dL (<9.2umol/dL)^[7], but this study did not assess the relationship between Zn status and nutritional outcomes.

The objective of this study was to advance the understanding of Zn status in very young children with CF by utilizing data collected from a multi-center prospective longitudinal study that follows a cohort of infants with CF born during 2012–17 until the youngest subject reaches 6 years of age in 2023. We hypothesized that serum Zn status is suboptimal in infants and toddlers with CF and this may affect their growth. To test this hypothesis, we determined the prevalence of suboptimal serum Zn status, identified factors contributed to low serum Zn levels, and assessed the associations between serum Zn and growth in young children with CF in their first 3 years of life.

METHODS

Study design and population

The present study utilized a subgroup of 53 children who had serum Zn (sZn) measurement in their first 3 years of life from a parent study known as FIRST (Feeding Infants Right…from the Start), which is a multi-center prospective longitudinal study designed with two objectives: 1) to identify optimal infant feeding and preschool diet for young children with CF, and 2) to better elucidate the impact of early malnutrition on lung disease development in CF. The FIRST cohort consists of 183 infants born during 2012–17, enrolled after NBS at age 1.9 ± 1.1 months, and being followed at six CF Centers located in five states (Madison and Milwaukee, WI; Boston, MA; Indianapolis, IN; Salt Lake City, UT; and Chicago, IL) until all children reach age 6 years in 2023. As of 12/31/2020, the entire FIRST cohort has reached 3 years of age. Rigorous guidelines for study conduct have resulted in a very detailed and comprehensive collection of nutritional and pulmonary data. The FIRST project was approved by the Institutional Review Boards at the University of Wisconsin-Madison and all participating institutions. Informed written consent was obtained from the parents of all participating infants.

Serum Zn measurement

Nutritional biomarkers measured in the FIRST study include fat-soluble vitamins A (retinol), D (25-hydroxyvitamin D), E (α-tocopherol), and essential fatty acid composition. Serum Zn concentrations are not being measured in the FIRST study blood specimens. However, all blood measurements obtained during routine CF center visits at each study site (referred to as “local labs”) are retrieved from electronic medical records on an ongoing basis into the FIRST study database.

A total of 106 sZn concentrations measured during the first 3 years of life from 53 subjects were identified from the FIRST study database. More specifically, 20 (38%), 18 (34%), 12 (22%), 1 (2%) and 2 subjects (4%) had one, two, three, four and five serum Zn measurements, respectively. In addition, 98 measurements (92%) were found to be from one study site that routinely measure sZn. Review of from the subjects’ clinical laboratory reports showed that sZn were measured by ARUP Laboratories (Salt Lake City, UT) using quantitative inductively coupled plasma-Mass Spectrometry. In our study, suboptimal Zn status was defined as sZn <70 μg/dL^{[11, 12]}.

Growth assessment

The FIRST study visits were designed to coincide with recommended schedule for clinical care, i.e., monthly after diagnosis until 6 months of age, bi-monthly from 6 to 12 months of age, and every 3 months thereafter^[13]. Growth is measured at each visit by the CF center staff according to the Anthropometry Procedures Manual used in the 2009–10 National Nutrition and Health Examination Survey^[14]. Recumbent length (before 2 years of age) and standing height (after 2 years of age) are measured in triplicates to the nearest 0.1 cm, and weight is measured in duplicates to the nearest 0.01 kg. Average lengths/heights and weights are used in all data analyses. Intra-visit variabilities as indicated by coefficient of variations (CV) were 0.24 ± 0.44% for weights (13.8 ± 1.5 kg) and 0.22 ± 0.41% for heights (91.2 ± 4.2 cm) based on >400 sets of replicate measurements in the 3^rd year of life. In the present study, a total of 934 growth measurements were utilized to calculate age- and sex-specific percentiles and z-scores for weight-for-age (WFA), length-for-age (LFA), weight-for-length (WFL) and BMI-for-age by using the WHO references for age 0–24 months and CDC references for age 25–36 months [height-for-age (HFA) in place of LFA and weight-for-height (WFH) in place of WFL] as recommended by the CDC and the American Academy of Pediatrics^{[15, 16]}

Indicator of intestinal malabsorption in CF

CF is characterized by intestinal malabsorption due to pancreatic insufficiency (PI), which must be accounted for in studies of nutrient status such as serum Zn. Meconium ileus (MI), a neonatal presentation of intestinal obstruction, is the most severe manifestation of PI, occurs in approximately 15% of infants with CF^[17]. Another 60–70% of patients with CF have no MI but are PI, while the remaining 10–20% are typically pancreatic sufficient (PS). In routine clinical settings, pancreatic functional status in CF is assessed and defined by fecal elastase-1 concentration; <100 mcg/g indicates PI and ≥200 mcg/g defines PS^[18]. In the FIRST study, fecal specimens are collected at approximately 2, 4, 6, 8, 12 mo and annually thereafter. Children with CF are classified as PS if their fecal elastase-1 concentrations are consistently >200 mcg/g in the first 2 years of life.

Other relevant covariates

In the FIRST study, sex, race, and ethnicity were recorded by the research coordinators at enrollment. Race and ethnicity were termed and categorized according to that reported in the human subjects enrollment required by NIH, namely, 7 categories for race (White, Black or African American, Asian, American Indian or Alaska Native, Native Hawaiian or other Pacific Islander, more than one race, or unknown) and 2 categories for ethnicity (Hispanic or Latino and Not Hispanic or Latino). Gestational age, birth weight and birth length were collected from electronic medical records and verified with newborn discharge reports. Mutations for the CFTR gene were collected from electronic medical records and verified with newborn screening reports.

Statistical analysis

SAS (version 9.4, SAS Institute, Inc, Cary NC, 2016) was used for data processing and statistical analyses. For continuous measures, Analysis of Variance (ANOVA) were used to compare differences when Normally distributed and a non-parametric rank sum was used when the data were skewed and transformations did not work. Comparison in proportions were assessed by Chi-square test Fisher’s Exact test being used when expected cell counts were small.

Cross-sectional correlations between sZn and growth z-scores were assessed by Pearson’s correlation coefficient. To identify factors associated with sZn concentrations and the associations between sZn and growth in the first 3 years of life, multiple regression analysis with repeated measures were performed using generalized estimating equations (GEE) with a working assumption of independence among observations^[19], using the PROC GENMOD procedure in SAS.

RESULTS

From the entire FIRST cohort, we identified 53 subjects who had a total of 106 sZn measurements. Seventeen infants (32%) had intermittent Zn insufficiency (defined as <70 mcg/dL on ≥1 measurement), while 36 infants (68%) were always Zn sufficient. Table 1 shows the demographic and clinical features of the total study cohort and a comparison between those infants who were always Zn sufficient and who were intermittently insufficient. A significant gender difference was observed in sZn status; about two-thirds of the always Zn sufficient children were females, compared to one-third of the intermittently insufficient Zn group. There was a trend towards lower birth weight and length in the intermittently Zn insufficient group and the proportion of F508del homozygous infants was higher, although the P value for the genotype difference was just above the prespecified threshold of 0.05. Consistent with this observation, the intermittently Zn insufficient group also showed a trend of proportionately more children with a history of MI and no PS children compared to the always Zn sufficient group. Five infants (9% of the total) had an sZn≤60 mcg/dL, and 6 of the 13 measurements (46%) from these infants were also ≤60 mcg/dL.

Table 1.

Characteristics of the study population

	All	Zn always sufficient (all values >70 mcg/dL)	Zn intermittently insufficient (≥1 value <70 mcg/dL)	p-value
# Subjects	53	36	17
Serum Zn status
# of measurements	106	65	41
age at measurement, mean ± SD	14.1 ± 10.1	14.1 ± 10.2	13.9 ± 10.2	0.94
concentration (mcg/dL), mean ± SD	85.3 ± 17.3	90.2 ± 14.1	77.5 ± 19.1	<0.001
insufficient (<70 mcg/dL), # measurements (%)	19 (18%)	0	19 (46%)
Demographics
Female sex, n (%)	29 (54%)	23 (64%)	6 (35%)	0.05
White race, n (%)	51 (96%)	36 (100%)	15 (88%)	0.10
Not Hispanic or Latino, n (%)	53 (100%)	36 (100%)	17 (100%)
Gestational age
weeks, mean ± SD (n)	38.3 ± 1.8 (51)	38.5 ± 1.5 (34)	38.0 ± 2.2 (17)	0.44
Premature (< 37 weeks), n (%)	9 (17%)	5 (14%)	4 (24%)
Birth weight
gram, mean ± SD	3047 ± 467	3081±465	2975 ± 477	0.44
z-score, mean ± SD	−0.57 ± 1.08	−0.47 ± 1.07	−0.77 ± 1.09	0.35
Low birth weight (<2500 g), n (%)	6 (11%)	4 (11%)	2 (11%)	1.00
Birth length
centimeter, mean ± SD (n)	48.7 ± 2.3 (51)	48.9 ± 2.3 (34)	48.4 ± 2.2 (17)	0.43
z-score, mean ± SD	−0.44 ± 1.16	−0.32 ± 1.15	−0.67 ± 1.17	0.31
Low birth length (z-score<-2), n (%)	4 (8%)	2 (6%)	2 (12%)	0.59
CFTR genotype by F508del status, n (%)				0.06
F508del/F508del	27 (51%)	15 (42%)	12 (71%)
F508del/other	24 (45%)	20 (56%)	4 (24%)
Other/other	2 (4%)	1 (3%)	1 (6%)
CF gastrointestinal phenotype, n (%)				0.33
Meconium ileus (MI)	10 (19%)	5 (14%)	5 (29%)
Pancreatic insufficiency (PI)	41 (77%)	29 (81%)	12 (71%)
Pancreatic sufficiency (PS)	2 (4%)	1 (6%)	0 (0%)

Figure 1 shows sZn vs age in the first 3 years of life for the study cohort. There was an inverse relationship between sZn and age, although this relationship was primarily observed in the first year of life. Table 2 shows the results of the multiple regression analysis of factors associated with sZn concentration in the first 3 years of life. Consistent with the pattern observed in Figure 1, age was significantly and negatively correlated with sZn. In contrast, female sex and non-F508del CFTR genotype were positively associated with sZn.

Figure 1.

Serum Zn concentration in the first 3 years of life. Data are presented for all 53 subjects in Panel A. Panel B shows the initial measurement of Zn for all 53 subjects. Panels C and D show any additional measurements, with the second measurement in blue triangles and the third measurement in red circles.

Table 2.

Multiple regression analysis assessing factors associated with serum Zn concentrations (mcg/dL) in the first 3 years of life

	Coefficient	p-values
Age (months) at serum Zn measurement	−0.4	0.04
Demographic characteristics
Sex, girls vs boys	7.7	0.02
Race, White vs all others	0.6	0.96
Birth characteristics
Birth weight z-score	1.0	0.59
Birth length z-score	−0.2	0.90
Gestational age, full-term vs premature	3.9	0.26
CFTR genotype
F508del/other vs F508del/F508del	3.6	0.20
Other/other vs F508del/F508del	18.6	<0.001
CF gastrointestinal phenotype
Pancreatic insufficiency vs meconium ileus	3.1	0.378
Pancreatic sufficiency vs meconium ileus	6.1	0.16
Additional Zn supplementation
Yes vs no	−15.0	0.13

The cross-sectional relationship between sZn and growth indices is shown in Figure 2. Overall, there were no significant associations between sZn and WFA, LFA/HFA, or WFL/WFH z-scores. However, the directions of the associations were in opposite trends in LFA/HFA and WFA z-scores comparing younger infants <6 month of age (panels B and C, Figure 2) to older infants and toddlers (panels E and F, Figure 2). In addition, when we analyzed the longitudinal growth patterns of the children who were always Zn sufficient compared to those with intermittent Zn insufficiency (Figure 3), both WFA and LFA z-scores were lower in the first 6 months of life in the latter group. At 3 years of age, WFA and HFA z-scores were lower in the intermittently insufficient cohort, although these differences were not statistically significant. On the other hand, WFH z-score was significantly lower in the intermittently insufficient group. It should be noted that we did not have sZn levels at 3 years of age for many children, precluding an analysis of the cross-sectional correlation between sZn and growth z-scores in these 2 groups at this time point.

Figure 2.

Cross-sectional correlation between the first serum Zn and growth z-scores by age (<6 & 12–36 months). At each panel, correlation between Serum Zn and Z-score is indicated by Pearson’s correlation coefficient (r) and p value

Figure 3.

Growth in the first 3 years of life in Zn sufficient (always >70) shown in black crosses and sometimes insufficient (at least 1 measurement <70) shown in black circles. The GEE p values are adjusted for birth weight, age, sex, race, prematurity, F508del status, pancreatic insufficiency, and Zn supplementation.

DISCUSSION

In this analysis of subjects from the FIRST study ages 1–39 months with at least one sZn measurement, we found a high prevalence of Zn insufficiency; one-third of this cohort exhibited Zn insufficiency at some point during their first 3 years of life. Zn insufficiency was more likely to be present in male infants and those who were F508del homozygous, and sZn levels declined with age. There was no significant cross-sectional correlations between sZn and WFA, LFA/HFA, or WFL/WFH z-scores, but children who had at least one low sZn measurement did have a lower WFH z-scores at 3 years of age compared to those who were always Zn sufficient.

To our knowledge this is the first study that has analyzed Zn status and its relationship to growth in a cohort of children with CF less than 3 years of age who were diagnosed through NBS. Krebs et. al. demonstrated approximately 30% of infants with CF diagnosed through newborn screening had Zn deficiency, but they did not study the relationship between Zn deficiency and growth^[7]. Although they used a lower threshold than our study (<60ug/dL), the overall prevalence of Zn insufficiency was similar in both our study and theirs. Maqbool et al performed a cross-sectional study of older children with CF ages 8–10 years old and also observed no association between growth and plasma Zn levels^[20]. Monge et. al. reported 17.6% of people with CF (PWCF) at their institution ages 2 to 31 years (median: 14.8 +/− 8 years) were Zn deficient, defined as <70mcg/dL in children <10 years of both sexes and in women aged ≥10 years and <74mcg/dL in men aged ≥10 years^[21]. The same study also found a positive association between sZn and WFH and BMI z-score. However, this study had a small sample size (7 males, 10 females) and included a wide age range of subjects. Compared to Monge, et al, both our study and Krebs, et al found a higher prevalence of Zn deficiency in infants. These findings are consistent with other studies suggesting infants are at higher risk of Zn deficiency^{[9, 22, 23]}.

Our results suggest that there could be other factors which play a greater role in growth and nutrition outcomes during infancy and toddlerhood in PWCF besides sZn. Multiple studies have reported an association with growth in PWCF and exocrine pancreatic function^{[24, 25]}. In a prospective longitudinal multicenter study of 99 infant PWCF, Munck et al. found an association between WFA and LFA z-scores and the presences of initial pulmonary symptoms as well as an association between LFA z-score and hospitalization for respiratory symptoms and Staphylococcus aureus infection^[24]. In comparison, Patterson et al found an association between a single isolation of Pseudomonas aeruginosa and reduced weight and height in infant PWCF in the first 2 years of life^[25]. Similarly, the Baby Observational and Nutrition Study (BONUS) found lower weight or length in the first year of life was associated with P. aeruginosa infection, MI, PI, male sex, and histamine blocker use^[26]. Other potential factors include alterations in levels of insulin-like growth factor, growth hormone binding protein, essential fatty acids, and growth hormone receptor gene expression^[26–28].

It is also possible that sZn is not a good measure of total body Zn status. We used sZn because it is the only biomarker clinically available to evaluate Zn status, and it is currently the only laboratory test recommended by the WHO and other international organizations for the evaluation of Zn status. Serum Zn may also have low sensitivity to detecting marginal Zn status^{[6, 23]}. Multiple factors can affect sZn levels including age, sex, fasting status, time of day the sample is obtained, and the presence of acute infection or inflammation^{[23, 29, 30]}. These factors further affect the ability of sZn to accurately reflect steady state total body Zn statues. Albumin concentration can also affect sZn concentrations, since it is the primary carrier protein for Zn^{[12, 31]}. However, unlike for calcium, there are no validated equations for correcting or adjusting sZn based on albumin concentrations. Other measures of Zn status, such as red blood cell (RBC) Zn may be a better measure of Zn status, but this test is neither widely used nor available for clinical purposes^[29].

The role of Zn supplementation in infants and young children with CF remains unclear. There are limited data on the impact of Zn supplementation on growth in young PWCF, and reported outcomes have been inconsistent^[32–35]. US and European guidelines recommend considering a 6-month trial of Zn supplementation in infants under 2 years of age who are not adequately growing despite reasonable caloric intake and pancreatic enzyme replacement therapy^{[36, 37]} but these were derived by consensus and not evidence based. Given the high prevalence of intermittent serum Zn insufficiency observed in our study and others, it will be useful and important to conduct a prospective study utilizing a standardized protocol to measure RBC Zn in conjunction with serum Zn, and concurrently assess Zn intake and supplementation status. This will generate novel data needed to definitively determine the role of Zn in CF nutrition.

There are several limitations of our study. Since measurement of sZn was not part of the FIRST study protocol, not all study participants had sZn measurements obtained. As such, it is possible that our study was underpowered to detect significant associations between sZn and growth outcomes. The majority (85%) of the subjects with sZn measurements came from a single study site that routinely obtains sZn as part of clinical care on all CF patients, potentially introducing the possibility of bias. In comparison to the entire FIRST cohort, baseline characteristics were overall similar with the exception that infants from this study site had lower birth weight and birth length. We defined Zn insufficiency as an sZn<70 mcg/dL. There is no consensus definition of Zn insufficiency and other investigators have used higher and lower thresholds in their studies. If we had used a cutoff of 60 mcg/dL, the power of detecting any significant associations between Zn insufficiency and growth would be even lower since fewer infants fell below this threshold compared to the higher threshold we used. Since infants with CF are at high nutritional risk, we chose to use the higher cutoff.

In conclusion, Zn insufficiency was common in this cohort of infant and young PWCF. We did not observe significant cross-sectional associations between sZn and growth. However, longitudinal analyses revealed somewhat different growth patterns between young PWCF who experienced intermittent Zn insufficiency and those who were always Zn sufficient. These results suggest that the association between Zn and growth requires further investigation with larger sample size and more consistent sZn measurements, and that factors other than Zn may play a greater role in growth in infant and young PWCF. However, an alternative interpretation is that sZn is not a good measure of total body Zn status. More studies are needed to further our understanding of the role of Zn in growth as well as the best methods to measure total body Zn status in an individual.

Abbreviations:

FIRST

Feeding Infants Right from the Start

CF

cystic fibrosis

PWCF

people with CF

Zn

zinc

sZn

serum Zn

WFA

weight-for-age

LFA

length-for-age

HFA

height-for-age

WFL

weight-for-length

WFH

weight-for-height

CFTR

cystic fibrosis transmembrane conductance regulator

NBS

newborn screening

MI

meconium ileus

PI

pancreatic insufficiency

PS

pancreatic sufficiency