Severe acute malnutrition (SAM), currently affecting nearly 20 million pre-school-aged children globally, is a significant factor in approximately one-third of the nearly 8 million deaths of under-five children(1). In India, almost 7·5 % under-five children are severely wasted (weight-for-height <–3 sd) and 21 % are wasted (weight-for-height <–2 sd) according to the National Family Health Survey-4 (NFHS-4) data(2).
SAM is a clinical syndrome due to an imbalance between the demand and supply of energy content, proteins and micronutrients, with a complex interplay of various pathological mechanisms. Severely malnourished children have increased total body water and sodium, while there is deficiency of K, Mg and phosphate stores(Reference Halliday3–Reference Hother, Girma and Rytter5). However, there is paucity of data regarding serum Ca levels in severely malnourished children. Chisti et al. (Reference Chisti, Salam and Ashraf6) found that hypocalcaemia was prevalent in 26 % of severely malnourished children aged <5 years admitted to an urban hospital in Bangladesh, and was associated with serious consequences such as seizures and death(Reference Chisti, Salam and Ashraf6). Abdominal distension, acute watery diarrhoea, vomiting, lethargy and severe sepsis were also more commonly seen in severely malnourished children with hypocalcaemia(Reference Chisti, Salam and Ashraf6–Reference Rudick and Sheehan9).
According to the WHO and Indian Academy of Pediatrics (IAP) guidelines on hospital-based management of severely malnourished children, all severely malnourished children should be given supplemental K, Mg, vitamin A, folic acid, Zn, Cu, Fe and multivitamins routinely(Reference Dalwai, Choudhury and Bavdekar10,Reference Bhatnagar, Lodha and Choudhury11) . The multivitamin supplement should preferably contain vitamins A, C, D, E and B12 besides thiamine, riboflavin and nicotinic acid. There is no recommendation on routine supplementation of Ca or vitamin D, although the F-75 and F-100 diets in the stabilisation and nutritional rehabilitation phases, respectively, contain approximately 320 mg Ca/l, and ready-to-use therapeutic food (RUTF) contains some vitamin D (15 μg per sachet, with each child receiving 1·5–5·0 sachets per day according to body weight)(12). Intake from these sources may not be sufficient to consistently elevate the circulating concentrations of serum 25-hydroxycholecalciferol (25-(OH)D) in children with SAM, considering a high prevalence of vitamin D deficiency in this group(Reference Ejaz and Latif13–Reference Berkley, Ngari and Thitiri15). Systemic inflammatory response may also dysregulate vitamin D metabolism and increase requirement of the same(Reference Jones, Hachmeister and Khasira16).
A better understanding of the prevalence and clinical predictors of hypocalcaemia in children with SAM shall contribute to more focused strategies, reducing hypocalcaemia-related morbidity, especially in resource-poor settings. We conducted the current study to estimate the prevalence of hypocalcaemia and determine its predictors in under-five children with SAM. Secondary outcome measures included serum vitamin D status and clinical rickets.
Methods
This was a cross-sectional study conducted in children aged between 1 and 59 months hospitalised with SAM in the paediatrics ward of a tertiary care hospital in Delhi, between November 2017 and April 2019. SAM was defined as weight-for-length/height <–3 z-score of the WHO growth standards, or bilateral pedal oedema of nutritional origin, or mid-upper arm circumference (MUAC) <11·5 cm (for children aged 6–59 months)(17). Children taking Ca or vitamin D supplements at the time of admission or within the last 2 weeks prior to admission and those with chronic illnesses such as heart defects, renal diseases, cancer and congenital anomalies were excluded.
Sample size
Using an estimated prevalence (P) of hypocalcaemia in SAM at 26 % as per data from Chisti et al. (Reference Chisti, Salam and Ashraf6) and an acceptable absolute precision (d) of ±7 % (acceptable relative precision of 25 %) with 95 % CI, the sample size was estimated to be 150.
Patient details and laboratory measurements
Patient’s information, case history and relevant examination, laboratory investigations, treatment received, duration of admission and final diagnosis at the time of discharge/death were recorded in a predesigned case record form. Biochemical parameters analysed included serum Ca, serum phosphate, serum alkaline phosphatase (ALP), serum 25-(OH)D, serum Mg, serum albumin, serum Na and serum potassium. Venous blood (3 ml) was collected, and the sample centrifuged within 15 min from the time of collection, and the separated serum was stored at −20°C.
Serum Ca was estimated using a UniCel DxC 600 automatic analyser (Beckman Coulter India Pvt. Ltd) by ion-selective electrode (ISE) method, measured in mg/dl, corrected for serum albumin levels and then converted into mmol/l for classification of hypocalcaemia. Formulae for calculated parameters are as follows(Reference Sava, Pillai and More18):
-
1. Corrected total Ca (mg/dl) = Total Ca (mg/dl) + 0·8 × (4 – serum albumin in g/dl)
-
2. Corrected total Ca (mmol/l) = Corrected total Ca (mg/dl) × 0·25
Hypocalcaemia was defined as serum Ca (albumin-adjusted) <2·12 mmol/l(Reference Fong and Khan19). Serum 25-(OH)D levels were estimated with an immunoassay analyser by chemiluminescence method and commercially available kits (Beckman Coulter Inc.). Serum 25-(OH)D between 30 and 50 nmol/l was considered as vitamin D insufficiency and <30 nmol/l as vitamin D deficiency(Reference Sahota20).
Severely malnourished children with hypocalcaemia were compared with those without hypocalcaemia with regard to gender, age, socioeconomic status, duration of exclusive breastfeeding, diarrhoea, vomiting, lethargy, shock, convulsion, respiratory rate per minute, diastolic blood pressure, sepsis, abdominal distension, hypoxaemia (spO2 <90 %), anthropometric parameters, serum Na, serum potassium, serum phosphate, serum 25-(OH)D and serum Mg. Sepsis was defined as the presence of inflammation (abnormal leucocytes count (>12 000 or <4000/mm3) or band and neutrophil ratio (≥0·1)) plus the presence or presumed presence of infection with thermo-instability (hypothermia (≤35°C) or hyperthermia (≥38·5°C)) and tachycardia(Reference Chisti, Duke and Robertson21,Reference Dellinger, Levy and Rhodes22) . Rickets was defined based on clinical examination (presence of rachitic rosary, frontal bossing, wrist widening, wide-open fontanel, pot-belly abdomen, Harrison’s sulcus, genu valgum and/or double malleoli) aided by biochemical tests and confirmed by radiographs.
Patient management
All enrolled children were managed as per WHO and IAP guidelines(Reference Dalwai, Choudhury and Bavdekar10–12), including supportive care, intravenous fluids, oral rehydration salts solution, antibiotics depending on the local resistance pattern and availability, oxygen therapy, blood transfusions if required, frequent monitoring and assessment and nutritional supplements. We used local preparations of F-75 and F-100 diets in the stabilisation and rehabilitation phases. In the emergency treatment of children presenting with hypocalcaemic seizures, intravenous 10 % calcium gluconate was administered slowly at a dose of 2 ml/kg (maximum 10 ml) followed by an infusion of calcium gluconate (1–5 ml/kg per day) titrated to maintain eucalcaemia. Those children with asymptomatic hypocalcaemia were supplemented with oral Ca (500–800 mg/d)(Reference Khadilkar, Khadilkar and Chinnappa23). For the treatment of nutritional rickets and vitamin D deficiency, daily doses of 50 μg of vitamin D were given to children between 1 and 3 months; a single dose of 1·25 mg or daily doses of 50 μg for children between 3 and 12 months; and a single dose of 3.75 mg or daily doses of 75–150 μg for children between 12 and 59 months(Reference Khadilkar, Khadilkar and Chinnappa23,Reference Munns, Shaw and Kiely24) . Oral vitamin D doses were given along with oral Ca 500–800 mg/d, regardless of age or weight, for a minimum duration of 3 months(Reference Khadilkar, Khadilkar and Chinnappa23,Reference Munns, Shaw and Kiely24) .
Statistical analysis
Data and findings of the current study were entered into a personal computer using SPSS for Windows (version 20.0). The prevalence of hypocalcaemia in under-five children hospitalised with SAM was calculated. Differences in proportions (for categorical data) and in means (for continuous data) were calculated using χ 2 test and Student’s t test, respectively. Clinical predictors of hypocalcaemia in severely malnourished children were identified as per the following process: sociodemographic and clinical characteristics were analysed initially in a univariate model, and further a logistic regression model was constructed taking hypocalcaemia as a dependent variable and co-variates with P < 0·1 (on univariate analysis) as independent variables. P < 0·05 was considered significant.
Results
We enrolled 150 children, out of which seventy-six (50·7 %) were boys; 129 (86 %) hailed from urban areas, nineteen (12·7 %) from rural areas and two (1·3 %) from urban slums. A majority (n 126, 84 %) of children were aged <24 months.
Hypocalcaemia with corrected serum Ca <2·12 mmol/l was documented in thirty-nine (26 %) out of 150 children with SAM. Of these, eleven were <6 months of age. The prevalence of hypocalcaemia was comparable between children <6 months (11/41, 26·8 %) and those between 6 and 59 months (28/109, 25·7 %) (P = 0·887). The mean and median values of corrected serum Ca in severely malnourished children were 2·23 (sd 0·26) and 2·27 (IQR 2·09, 2·38) mmol/l, respectively.
A large proportion of the enrolled children (n 126, 84 %) were either deficient (n 98, 65·3 %) or insufficient (n 28, 18·7 %) in serum 25-(OH)D. Of thirty-nine children with hypocalcaemia, thirty (76·9 %) were deficient in serum 25-(OH)D, four (10·2 %) were insufficient and five (12·8 %) had normal serum 25-(OH)D; corresponding values for children with normocalcaemia (n 111) were sixty-eight (61·3 %), twenty-four (21·6 %) and nineteen (17·1 %), respectively. Clinical rickets were observed in sixty-three (42 %) of 150 children.
Results of a univariate analysis comparing hypocalcaemic with normocalcaemic children for various sociodemographic and clinical characteristics are shown in Table 1. Results of a logistic regression analysis are depicted in Table 2. The presence of clinical rickets had 6·6 times higher odds of having hypocalcaemia compared with those without rickets (95 % CI 2·54, 17·15). The odds of hypocalcaemia were 4·5 and 2·6 times higher in children with abdominal distension and sepsis, respectively.
Univariate analysis of predictors of hypocalcaemia in children with severe acute malnutrition (n 150)
25-(OH)D, 25-hydroxycholecalciferol.
* Children belonging to the lower and upper-lower socioeconomic class as per modified Kuppuswamy scale.
† Till the date of enrolment in infants <6 months.
‡ n 21 and 81, respectively, in hypocalcaemic and normocalcaemic groups.
Results of logistic regression to explore the independent clinical predictors of hypocalcaemia in under-five children with severe acute malnutrition (n 150)
Discussion
Hypocalcaemia was present in about one-fourth of severely malnourished children, the prevalence being comparable between children under and over 6 months of age. The presence of clinical rickets, abdominal distension and sepsis was recognised as an independent predictor of hypocalcaemia in severely malnourished children.
Our results are consistent with those reported by Chisti et al. (Reference Chisti, Salam and Ashraf6) A cross-sectional, descriptive study conducted in Kathmandu, Nepal, by Mishra et al. (Reference Mishra, Bastola and Jha25) has found that hypocalcaemia was prevalent in 46·7 % of malnourished children, which is higher than that reported by the current study. This difference in findings despite similar age groups of participants might be because of the difference in the selection of participants, in which Mishra et al. (Reference Mishra, Bastola and Jha25) included sixty children (aged 6–59 months) with protein energy malnutrition (PEM) (weight-for-height below the 80 % of 50th percentile of weight-for-height based on the growth chart of the National Centre of Health Statistics), while our study included children aged 1–59 months hospitalised with SAM (defined as per WHO 2013)(Reference Mishra, Bastola and Jha25). In another study from Uganda on children aged 6–24 months(Reference Nabeta, Kasolo and Kiggundu26), the prevalence of hypocalcaemia was 43·5 %. However, the participants included children with both SAM and moderate acute malnutrition (MAM). The prevalence of rickets and the cut-off used for defining hypocalcaemia (2·2 mmol/l) were higher in these children. Rickets in African settings is more likely to be due to dietary insufficiency of Ca(Reference Edwards, Thiongó and Van den Bergh27–Reference Özkan30).
We observed vitamin D deficiency in 65·3 % of severely acute malnourished children. This is much higher than the findings of a cross-sectional study by Nabeta et al. (Reference Nabeta, Kasolo and Kiggundu26) (14·6 %). It is possible that our children might have had combinations of dietary deficiencies of Ca and vitamin D and limited sun exposure, considering that a major proportion (86 %) hailed from urban areas and 84 % were between 1 and 24 months of age (with infants and younger children less likely to be out in sunshine). These inconsistent findings between the two studies can be explained by the differences in age groups, criteria of malnutrition used for recruitment of cases, and the cut-offs used for defining vitamin D deficiency. Nabeta et al. (Reference Nabeta, Kasolo and Kiggundu26) recruited 117 children in Uganda between the ages of 6 and 24 months with malnutrition defined as SAM (weight-for-length/height <–3 sd) or MAM (weight-for-length/height <–2 and up to –3 sd), and cut-off used was ≤20 ng/ml. Moreover, vitamin D status can be influenced by geography, climate, food production and availability, or religious and cultural practices, skin pigmentation and burden of infectious and chronic diseases in different regions(Reference Prentice, Schoenmakers and Jones31).
The presence of clinical rickets was identified as the most significant independent predictor of hypocalcaemia in children with SAM. Rickets is supposed to be uncommon in children with malnutrition(Reference Salimpour32–Reference Thirupuram, Srivastava and Gupta34), but recent studies have reported a strong association between the two(Reference Soliman, Madina and Morsi35–Reference Lulseged and Fitwi37). Severely malnourished children with clinical signs of rickets indicate a fluctuating nutritional course, with periods of infection and nutritional deprivation alternating with periods of recovery, while those without clinical signs of rickets have a chronic steady nutritional deprivation(Reference Soliman, Madina and Morsi35). Further, it is now well known that the aetiology of nutritional rickets ranges from isolated vitamin D deficiency to isolated Ca deficiency(Reference Özkan30,Reference Thacher, Fischer and Tebben38) .
Apart from nutritional rickets, the presence of abdominal distension and sepsis also contributed significantly to the predictors of hypocalcaemia in children with SAM. Abdominal distension in hypocalcaemia might indicate pot-belly abdomen, which is one of the classical findings of rickets in children(Reference Özkan30). In addition, a frequent observation of abdominal distension and hypocalcaemia in children with SAM indicates ileus in such children(Reference Chisti, Ahmed and Faruque8,Reference Rudick and Sheehan9) . The current study also indicated that severely malnourished children with hypocalcaemia had significantly higher proportion of sepsis compared with normocalcaemic ones. This corroborates with a finding of earlier studies that there is a strong association between hypocalcaemia and sepsis(Reference Chisti, Ahmed and Faruque8,Reference Rudick and Sheehan9) . Various other studies also postulated that severe pneumonia, gastrointestinal and ear infections are associated with vitamin D deficiency(Reference Thornton, Marín and Mora-Plazas39,Reference Haider, Nagi and Khan40) , and so with hypocalcaemia. Apart from its role in bone mineralisation, vitamin D has a vital role in regulating immune responses to infection(Reference Lang and Aspinall41).
Limitations of the current study include it being a cross-sectional hospital-based study with limited sample size and no control group. It is possible that vitamin D deficiency (and resultant hypocalcaemia) in enrolled children was just a reflection of deficiency in community rather than an accompaniment of malnutrition; however, a recent Comprehensive National Nutrition Survey(42) documented only 14 % of pre-schoolers (1–4 years) in India, and 32·5 % in Delhi to be deficient in 25-(OH)D, whereas in the current study, 65·3 % were deficient, reflecting a likely association of hypocalcaemia and vitamin D deficiency with SAM. Another limitation is a lack of measurement of ionised Ca, which may be a more accurate reflection of the physiologic Ca state. However, we had adjusted total serum Ca values for serum albumin levels; this reduces the chances of false low serum Ca levels in malnourished children due to hypoproteinaemia(Reference Sava, Pillai and More18,Reference Kragh-Hansen and Vorum43) . We feel that despite the limitations, the results still have clinical relevance in resource-poor settings where there is lack of facility to measure ionised Ca.
Our findings may help in decision-making regarding Ca supplements, particularly in resource-poor settings where routine estimation of serum Ca and serum 25-(OH)D is not immediately possible. We recommend routine supplementation of vitamin D in conjunction with oral Ca for a sufficient duration in severely malnourished children in view of the high prevalence of vitamin D deficiency/insufficiency in this group. The therapeutic approach recommended in the management of SAM contains vitamin D and Ca but not in sufficient quantities to take care of hypocalcaemia and vitamin D deficiency of the magnitude observed in the current study. Moreover, local preparations of therapeutic diets are often used, which usually lack these nutrients in adequate amounts. Also, the duration of administration of these diets is variable for different patients (ranging from a few days to 2–6 weeks) depending on the clinical condition and response of each child, after which we shift to home-based diets. We therefore feel that an intake of these sources for a short duration may not be sufficient enough to consistently elevate the concentrations of Ca and vitamin D. Further studies are required to establish appropriate dosages of Ca and vitamin D supplements in severely malnourished children and monitor for efficacy and safety.
Acknowledgements
Acknowledgement: None. Financial support: The current work was supported by the Indian Council of Medical Research (ICMR), New Delhi, under MD/MS/DM/MCH thesis financial assistance programme of ICMR, Human Resource Development grant no. 3/2/June-2017/PG-Thesis-HRD(36). Conflict of interest: There are no conflicts of interest. Authorship: Conception and design: P.G.; collection and assembly of data: C.S.; review of literature, intellectual inputs: C.S., P.B., R.S.A.; data analysis and interpretation: C.S., P.G., D.S.; manuscript writing: C.S., P.G., D.S., P.B., R.S.A.; final approval of manuscript: all authors are accountable for all aspects of the work. Ethics of human subject participation: The current study was conducted according to the guidelines laid down in the Declaration of Helsinki, and all procedures involving research study participants were approved by the Institutional Ethics Committee–Human Research (IEC-HR) of the University College of Medical Sciences, Delhi. Written informed consent was obtained from all guardians/caregivers of children.
