Normative Fecal Calprotectin Concentrations in Guatemalan Preschoolers Are High Relative to Children Reported Elsewhere

Abstract

Background: 

Calprotectin is a fecal marker of intraintestinal inflammation derived from activated enteric neutrophils and macrophages. It is useful as a clinical marker in inflammatory bowel diseases; furthermore, it may have a role in public health epidemiology.

Objectives: 

The aim of the study was to describe the distribution of fecal calprotectin in Guatemalan preschool children sharing a common institutional diet; to relate it collectively to pediatric distributions in other geographic settings, and individually to concomitant indicators of intestinal infection or colonization and other descriptive features of the child.

Methods: 

Fecal samples were collected in 87 subjects, ages 2 to 7 years across 3 daycare centers sharing a common institutional menu, but from different ecological settings. Stools were examined, variously by routine light microscopy, quantitative egg counts, and a Giardia antigen test, for microbiological diagnosis, and an ELISA assay for fecal calprotectin (CalproLab).

Results: 

The median fecal calprotectin value was 58 mg/kg, with a mean of 98 ± 136 mg/kg and a range from 10 to 950 mg/kg; 61% of values were above the manufacturer's cut-off for elevated concentration and 51% exceeded an age-adjusted criterion. There were no associations between sex, age, growth indicators, or fecal microbiological findings by microscopy or ELISA assays, alone or in combination. The central tendency (mean or median) and distribution were generally shifted to the right in relation to comparable reports from children across the world literature.

Conclusions: 

Although specific, low-grade intestinal infections do not define calprotectin subgroups, right-shifted fecal calprotectin status in this population may reflect a general and diffuse stress of adverse environmental sanitation.

What Is Known

• Fecal calprotectin is a protein derived from neutrophils and tissue macrophages in the gut and is known as a useful clinical marker in inflammatory bowel diseases.
• Viral and bacterial diarrheal pathogens have been associated with increased calprotectin responses.

What Is New

• Children raised in low-income environments in Guatemala present a constellation of intestinal microbial burden, and environmental enteric dysfunction as a consequence.
• Low-income preschoolers in Guatemala manifest the highest central-tendency calprotectin values so far reported from developed or developing countries.
• High prevalence of enteric inflammation should be considered as a factor for the poor linear growth in Guatemala.

Fecal calprotectin is a protein derived from neutrophils and tissue macrophages in the gut. It is a low-molecular weight, 36 kDa, dimeric peptide, with 2 calcium-binding sites of a molecular weight related to the S-100 series; it can constitute 30% to 60% of cytosolic proteins of an intestinal neutrophil. It is recognized under several synonyms including S1008 and S1009 ^(1,2). It is considered to be involved in the antibacterial action and cellular chemotactic activities ⁽³⁾, and has an apoptotic effect on exogenous cells ⁽⁴⁾. Calprotectin is well recognized in the area of diagnostics, where it provides a noninvasive biomarker of the inflammatory status of the intestinal tract.

The intestinal exudate that forms in inflammatory bowel diseases (IBD), including ulcerative colitis and Crohn disease, gives rise to a fecal excretion of calprotectin that permits its differentiation from noninflammatory functional bowel disorders and allows monitoring of flares and remissions in the activity of IBD ^(5–8). In addition, calprotectin increases in the feces in cystic fibrosis ^(9,10), cow's milk protein allergy (in neonates and throughout the first year of life, fecal calprotectin is higher than later in life ^(11–13)), and is nurtured by the exclusivity of breast-feeding ^(14,15); nevertheless, intestinal disease can push calprotectin even higher in the infancy year ⁽¹³⁾. In fact, normative values for fecal calprotectin follow an upward course throughout the first 4 years of life, stabilizing out beyond the 48th month ⁽¹⁶⁾.

Variable levels of fecal calprotectin have been documented in samples of free-living children in the small number of reports available in the literature ^(17–23), with only 1 from a low-income developing country, namely, Uganda ⁽²³⁾. In Chinese infants, fecal calprotectin levels were higher in a rural milieu, as compared to urban, and negatively associated with linear growth ⁽²⁴⁾. Finally, viral and bacterial diarrheal pathogens have been associated with increased calprotectin responses ⁽²⁵⁾. In a project based in 3 government-subsidized daycare centers in the Western Highlands of Guatemala, we collected fecal samples for assessment of fecal calprotectin levels and its relation to the evidence available on intestinal protozoa and helminthes infestation. We present here the results of this nested survey.

SUBJECTS AND METHODS

Geographical Setting and Sites

The study was conducted in the city of Quetzaltenango, the capital of the province of Quetzaltenango, located at about 2600 m above sea level. The provincial population is primarily of Mayan ascendancy, with a minority of mixed European-indigenous ancestry (mestizo or ladino), the latter concentrated in urban areas. Homes of low-income families are often of rudimentary constructions, poorly ventilated with indoor cooking stoves. Household piped water and electricity have become universal, but sewage disposal remains unsatisfactory.

The observations, monitoring, and collections were conducted in 3 local units of the Secretariat of Social Programs of the First Lady (SOSEP for the initials of the original Spanish: Secretaria de Obras Sociales de la Esposa del Presidente). They sponsor a nationwide system of daycare centers (Hogares Comunitarios) providing low-income families with the opportunity for caretakers to work outside of the home while their preschool children receive care, instruction, recreation, and 4 repasts. The feeding is based on a standard 40-day rotating menu, cycling every 8 workweeks, consisting of a breakfast, mid-morning snack, lunch, and afternoon snack.

The sites of the study were either inside or close to the urban limits of Quetzaltenango, all within a 20-km radius. Center A is in a semiurban setting in a suburb, La Esperanza, of the main city. Center B is situated in a marginal-urban area of the city itself, in the barrio of La Puerta del Llano. Finally, center C is located in a truly rural, agricultural zone in the hamlet of San Martín Chile Verde.

Enrollment of Subjects

Children, ages 2 to 7 years, attending the 3 previously described SOSEP daycare centers in or near metropolitan Quetzaltenango, Guatemala were sought for enrollment into the study. Most were of Mayan indigenous ascent, with a minority of ladino origin.

Exclusion Criteria

Children were excluded if they did not attend 1 of the 3 selected daycare centers at least 80% of the center's working days during the 40-day within the study period, or whose parents refused to sign the consent form. Eligible subjects were apparently healthy and with no restrictions in consuming the foods and beverages offered on the SOSEP menu.

Ethical Considerations

Ethical approval was granted by the Human Subjects Committee Center for the Studies of Sensory Impairment, Aging, and Metabolism. A parent or legal guardian signed a written informed consent form after the nature, purpose, risks, and benefits were explained. Authorization was also obtained from the district director of SOSEP for the Quetzaltenango area. A physician responded to the findings of the diagnostic tests (eg, hemogram, stool, and urine tests), and delivered deworming treatment along with medical prescription, for example, for oral iron supplementation or antibiotics for urinary tract infection.

Anthropometric Measurements

Children's height was measured in the 7th week of data collection using a wall stadiometer to the nearest 0.5 cm and with children standing in an erect posture without shoes with their gaze in the Frankfort plane. Weight was measured without shoes, but with normal daily attire, to the nearest 100 g on a calibrated Tanita Model BC522 digital scale (Tanita, Tokyo, Japan).

Collection of Fecal Specimens

Two samples of feces were collected; the first during the 7th, and the second one during the 8th week of the cycle. Feces collected during the 7th week of the cycle were used to determine fecal calprotectin and Giardia intestinalis. Specimens collected during the 8th week were used to perform the fecal light microscopy, quantitative microscopic helminth-egg counts (Kato-Katz), and a second determination of G intestinalis. Feces for calprotectin were immediately processed, homogenized, and stored in a −20°C freezer when analyses could be performed after storage.

Fecal Microbiology and G intestinalis Assays

For the fecal microscopy, a conventional microscopic preparation of feces was made in the clinical laboratory of the La Democracia Hospital in Quetzaltenango, immediately after collection. An experienced clinical microscopist examined the specimens to identify and report helminth ova, motile protozoan trophozoites, and protozoan cyst forms. They were reported in qualitative terms of present or absent in the fecal smear.

The Kato-Katz quantitative fecal helminthes egg count method using sedimentation and light microscopy was performed by a member of the research team (MERA). Only a total of 80 samples had the proper consistency to perform the quantitative, Kato-Katz thick-smear technique. A standard preparation of fecal samples with malachite green was set to dry under a hood and then covered with a cover slip to produce a permanent prep. Light microscopy was performed on a UNICO microscope (model G380LED, Dayton, NJ), examining the entire surface of the 1.0 specimens. The total number of worm eggs counted is multiplied by a factor of 10 for equivalency to the ova per gram of stool.

G intestinalis prevalence and intensity were derived in Guatemala using the ProSpecT-Giardia-EZ microplate ELISA assay (Oxford, Cambridge, UK) and read on a coanalyzer STAT FAX 303-plus ELISA plate reader (Awareness Technologies, Inc., Palm City, FL). The absorbance value was used as a proxy for the intensity recognizing that values >300 nm optical density (OD) cannot be further resolved.

Measurement and Interpretation of Fecal Calprotectin

Concentration of calprotectin in the fecal samples from the 7th-week collection was measured using the CalproLab Calprotectin ELISA Test Kit (CALP-0170, Calpro AS, Oslo, Norway) following the manufacturer's instructions for previous preparation, storage, and determination by ELISA procedure. Absorbance was read at an OD of 450 nm on the STAT FAX 303-plus ELISA reader.

We have used 2 established criteria for abnormal elevation in calprotectin concentration. The first is that of >50 mg/kg of feces, derived from the package insert and used generically in the literature. The second are age-adjusted concentration criterion derived from the 97.5th percentile of healthy Norwegian children stratified by age groups up to 4 years. The authors present their criteria as follows: “As a result, 3 cut-off levels were established based on the 97.5% percentiles of fecal calprotectin in different age groups: 538 mg/kg (1–6 months), 214 mg/kg (6 months–3 years), and 75 mg/kg (3–4 years)” ⁽¹⁶⁾. For our subjects younger than 48 months, >75 mg/kg was applied as an alternate criterion.

Data Handling and Statistical Analysis

Descriptive statistics of the median, mean, standard deviation, and minimum to maximum limits were calculated for the assorted variables. For calprotectin, with a non-Gaussian distribution, the median is the preferred expression, but the arithmetic mean was computed for comparison with some literature reports. Other variables were characterized with central tendency expressions appropriate to the normality of the distribution. The Mann-Whitney U test and Kruskal-Wallis test were used for nonparametric analysis of the medians between or among groups, and Student t test for parametric analysis of means within a normal distribution. Associations among variables were assessed using the Spearman rank-order correlation coefficient.

z Scores for weight-for-age (WAZ), weight-for-height (WHZ), and height-for-age (HAZ) were calculated as growth indicators using the WHO Anthro software ⁽²⁶⁾ for children ages up to 59 months, and Anthro Plus software ⁽²⁷⁾ for children 60 months and older.

RESULTS

Characteristics of the Subjects

Table 1 provides the descriptive statistics of mean, standard deviation, median and minimal and maximal limits for age, height, and the anthropometric growth indicators WAZ, WHZ, and HAZ for the 87 children, and disaggregated for the 42 female and 45 male participants. In general, the children were very short in both absolute (cm) and relative (by the HAZ growth indicator) terms. More than half of the subjects would be classified as stunted, with an HAZ below −2 SD. The rate of underweight was considerably lower, with the median WAZ of −1.4. There was virtually no wasting, with the WHZ median value close to the 0.00 z score level.

TABLE 1: Characteristics of the subjects by age, height, growth indicators, and fecal calprotectin concentration

Descriptive Statistics of Fecal Calprotectin

The final column of Table 1 provides the descriptive statistics of calprotectin concentrations for the whole sample of 87 preschoolers, and the subgroups of 42 girls and 45 boys. The aggregate calprotectin concentration had a median of 57 mg/kg, with an arithmetic mean of 98 ± 136 mg/kg and a min-max range of 10 to 950 mg/kg. There was no significant difference in concentrations by sex between the median concentration (P = 0.417).

In terms of diagnostic interpretation, using the criterion of the manufacturer (CalproLab) for active inflammation, as in an IBD, >200 mg/kg of stool, only 9 children (10%) would be classified as with severe inflammation. Using the manufacturer's generic criterion for “noninflamed” of <50 mg/kg, applicable to all ages, 32 subjects (37%) were so classified, leaving 55 (63%) in moderate or severe intestinal inflammation categories. The newer—and age-specific—Norwegian classification ⁽¹⁶⁾ uses a normative cut-off for children younger than 48 months of age of <75 mg/kg, leaving the <50 mg/kg for 48 months or older. Partitioning our 36 children younger than 4 years, and the remaining 51 above that age for application of these age-specific cut-offs, the recalculated aggregate rate of elevated fecal calprotectin declines to 51%.

Association of Fecal Calprotectin With Sex, Anthropometric Status, and Site

Table 2 compares median fecal calprotectin across anthropometric status categorical groups by the Mann-Whitney U test where applicable. The distribution for WHZ was too narrow and normal to be compared in this way. None of the probability values reached the 5% criterion for statistical significance. For the age in months and all of the anthropometric z score indices, associations with individual calprotectin were calculated as the Spearman rank-order correlation coefficient. Once again, the P values did not achieve statistical significance.

TABLE 2: Fecal calprotectin concentrations compared by and across age and growth indicators

Descriptive statistics for fecal calprotectin across the 3 daycare centers are shown in Table 3. As it can be seen, there was a 9 mg/kg gradient in calprotectin concentration from the semiurban to the rural setting, but without statistical significance (P = 0.499).

TABLE 3: Fecal calprotectin concentrations compared by the site and setting of collection

Description of the Microbiological and Parasitological Status of the Population

In terms of prevalence of helminthes, the 2 methods were combined; only 12 (15%) of the 80 subjects with a fecal sample suitable for analysis had any type of ova. Regarding prevalence and intensity, specifically by Kato-Katz technique, Ascaris lumbricoides was found in 9 samples, in 3 of which Trichuris trichiura was also found; 3 additional specimens had only hookworm species. The concentration of nematode eggs was modest, ranging from 30 to 13,320 per gram of feces.

The second approach was nonquantitative fecal smear for helminth ova and protozoal trophozoites and cysts, as performed by the microscopist of the collaborating hospital laboratory. One or another of the reportable fecal protozoa and helminths for the laboratory (A lumbricoides, Hymenolepis nana, G intestinalis, and Entamoeba histolytica (alone or combined with other reportable pathogenic or nonpathogenic parasites) were found in 32 of 87 stool specimens submitted. The third mode of diagnosis was for fecal G intestinalis with an ELISA method based on a giardia antigen. A total of 48 of 87 subjects tested (55%) were positive for an asymptomatic giardiasis by this method. In sum, cumulatively, across the 3 modalities of microbiological diagnosis, 68 subjects (78%) had 1 or more findings recorded in the database.

Interaction of Fecal Calprotectin With Fecal Microbes and Pathogens

We performed several analyses between the 2 domains of fecal diagnosis: calprotectin concentration and microbiology. The subgroup comparisons are presented in Table 4 based on prevalence. We also compared the median fecal calprotectin concentration in 4 comparative formats: any helminth eggs versus none (P = 0.676); any light microscopic findings versus none (P = 0.775); Giardia positive versus Giardia negative by antigen assay (P = 0.983); and any microbial prevalence versus none (P = 0.770). None reached statistical significance. Finally, the median 54 mg/kg of the 40 individuals who had more than 1 microbiologic species diagnosed across the microscopic and antigen-based examinations compares with a median 60 mg/kg for the 47 remaining subjects with 1 or no species detected. The P value was 0.565.

TABLE 4: Fecal calprotectin concentrations compared by different diagnostic criteria for microbial infestation of the intestinal tract

We next moved from the interaction in the binary sense of positive or negative for parasitic infection (prevalence-based analysis) in relation to elevated or normal calprotectin, to explore the possibility an examination based on a graded association involving the individual burden of parasites (intensity-based analysis). For helminth burden, we took advantage of the quantitative ova counts by the Kato-Katz microscopic technique. Only the 12 subjects with helminth eggs were included in this subanalysis; a nonsignificant rank-order correlation (r = 0.168, P = 0.602) was found between egg count and calprotectin values. For the Giardia burden, we use an intensity index previously used in a thesis performed at CeSSIAM to assess protozoa burden with growth ⁽²⁸⁾. This involves using the OD values for absorbance in the ELISA giardia antigen test as a continuous, scaling variable. Using all available assays (n = 87), a nonsignificant Spearman correlation (r = 0.112, P = 0.450) was found between the proxy for antigen burden and the amount of calprotectin in the stool.

DISCUSSION

Technology has emerged to allow for accessible diagnostic assessment of intestinal inflammation using a number of fecal biomarkers: lactoferrin, S100A12, and calprotectin ^(8,9,29,30). The latter is the most widely available and the most thoroughly studied. We took advantage of a field study conducted in preschool children to measure fecal calprotectin among various useful variables of growth, nutritional status, and intestinal health to construct a descriptive and comparative profile for this age group living in low-income circumstances in the region.

Apart from the specific illnesses associated with the activation of inflammatory cell in the intestine, graded dietary and environmental stimuli may be responsible for the higher or lower concentrations of fecal inflammatory biomarkers. In a geographic, comparative sense, we were able to identify 7 pediatric series in at-large populations or healthy controls in clinical studies in the literature (Table 5). The age variation makes direct comparisons among sites difficult to interpret. They are not perfectly age matched across the spectrum, a factor of consideration, given the age-sensitive calprotectin biology in early childhood ⁽¹⁶⁾. The median value of 55 mg/kg for our older children is not substantially different from Norwegian children of 5 years ⁽¹⁹⁾; however, our distribution is shifted to the right, with a high value of 950 mg/kg and 11 individuals exceeding the maximal concentration of 176 mg/kg seen in 5-year-olds in Norway. A comparison of the median of all 87 of our subjects with the medians for the broadest age ranges for the UK ^(17,22), Italy ⁽²¹⁾, and Sweden ⁽²⁰⁾ finds both our median and distribution of calprotectin values displaced to the right. As compared to the only other tropical country in the series, Uganda ⁽²³⁾, the comparison may be the most difficult of all, as it is heavily loaded with children who are both younger and older than the 2 to 7 age range of the Guatemalan series. What we can see in Table 5 is a median fecal calprotectin concentration of 28 mg/kg, about half of our 55 mg/kg value when comparing their 159 subjects above the age of 4 years with our 36 children older than 5 years.

TABLE 5: Comparison of central tendencies for fecal calprotectin in free-living populations of children

As with the reference literature (data not shown), our study failed to find differences between sexes in calprotectin status. Age is clearly a factor, with decreasing fecal calprotectin with increasing age, at least through 4 or 5 years ^{(16,18,19,23)}. We found a numerical tendency with a 10 mg/kg variation across medians above and below 5 years, but without statistical significance.

An experience in China found increased evidence of intestinal inflammation with decreased height-for-age z scores in a cohort within the first year of life ⁽²⁴⁾. The truly unique nature of these highland preschoolers is their degree of linear growth retardation or stunting, consistent with the position of Guatemala, as the nation with the highest under-five stunting rate in the Western Hemisphere (49.8%) ⁽³¹⁾. The total and sex-specific HAZ values in Table 1 reflect this reality. Overall, stunting was present in 57 of the 87 participants, for a global prevalence of 66%. The respective HAZ values indicate stunting rates of the respective sites resolves to: center A, 36%; center B, 63%; and center C, 81% (data not shown).

The same Chinese study ⁽²⁴⁾ found more rural location correlated with more elevated fecal calprotectin in infants. As described, the 3 sites of the study had a rural to urban gradient. Our findings fail to confirm any such association within this geographic continuum of calprotectin status in our preschoolers (Table 3).

On the basis of the literature review in Table 5, the fecal calprotectin concentration for our Western Guatemala subjects seems to be distributed generally to the right as compared to the experience for children in developed countries, and even compared to a series from Uganda, in East Africa. We interpret this as stronger background inflammatory response in the intraintestinal milieu. The explanation may reside in the contextual factors from the Western Highlands of Guatemala. A burden of intestinal parasitoses is evident, represented primarily by protozoan organisms of both commensal and pathogenic nature, especially G lamblia. We grant that we were unable to find differential or gradient relations with prevalence or intensity of infestation and colonization after exhaustive analysis (Table 4). The prevalence of both of the aforementioned classes, however, attests to a profound fecal-oral transmission among the children. That helminthic infections and their egg counts, such as common roundworm and hookworm, are so modest is commensurate with the altitude of more than 2600 m and intermittent deworming efforts of which the children—or their family members—are beneficiaries. Although specific associations with neither the presence nor intensity of parasite infestations and fecal calprotectin concentrations were evident, the higher concentration of fecal calprotectin in Guatemalan supposes a greater inflammatory state in the intestine. The frequent hosting of parasitic intestinal species sets the present children apart from those in the other studies listed, with the exception of Uganda. In an aggregate—but not graded—manner, there is a great deal of abnormal intestinal colonization. Moreover, although not documented in these children, environmental enteric dysfunction ^(32,33) is likely to be prevalent in these preschool children. As early as the 1970s, Rosenberg et al ⁽³⁴⁾ documented the conditions for environmental enteric dysfunction (then known as tropical enteropathy) in Guatemalan children. Less than pristine intestinal health in a diffuse sense is undeniably the situation of the study sample.

Strengths and Limitations of the Study

With 87 subjects, the present study ranks third in terms of total sample size among the comparative studies listed in Table 5, and has the second largest sample of children younger than 7 years of age after the series in Uganda ⁽²³⁾. This is both a strength and a weakness, as the overall narrow age range of 5 years limits its comparability with the more widely aged groups in the pediatric literature. A further strength is an epidemiological focus with a systematic collection of fecal samples to describe a preschool group in yet a second developing country. The subjects in the rural-urban Chinese report ⁽²⁴⁾ were infants. As a cross-sectional, point-in-time study with a single stool sample for each individual, we are able to construct a distribution, but we have no understanding of day-to-day within-individual variation, especially in those individuals with fecal calprotectin values in excess of 200 mg/kg.

CONCLUSIONS

In comparison with other countries, developed or developing, in which at-large populations of children have been samples for fecal calprotectin, our central-tendency values are the highest so far reported. Five of 10 or 6 of 10 subjects had an elevated calprotectin concentration in the stool, depending on the diagnostic criterion, and 1 in 10 had a level corresponding to clinically significant, active inflammatory disease of the bowel; all of this is seen in their free-living setting between home and daycare attendance. Although we have been unable to associate the individual burden of G lamblia or the fecal microorganisms from light microscopy in a gradient fashion with the individual calprotectin level, this constellation of intestinal microbial burden across the population may condition a generalized chronic inflammatory condition. Any interference with absorption or retention of essential nutrients derivative from this prevalence of enteric inflammation must be considered as a factor, potentially remedial, for the poor linear growth at the core of the Guatemalan public health agenda.

UNCITED REFERENCE

^(11,12).