Skip to main content
Download PDF

Early child stimulation, linear growth and neurodevelopment in low birth weight infants

BMC Pediatrics volume 22, Article number: 586 (2022) Cite this article

Abstract

Background

Children with low birth weight (LBW) are at risk of linear growth faltering and developmental deficits. Evidence suggests that early child stimulation and care reflected as responsive caregiving and opportunities for learning can promote development. The current analysis aimed to measure the extent to which linear growth and early child stimulation modify each other’s association with neurodevelopmental outcomes among LBW infants.

Methods

This is a secondary data analyses from a randomized controlled trial on the effect of community-initiated kangaroo mother care in LBW infants on their neurodevelopment at 12 months of corrected age. Bayley Scales of Infant and Toddler Development was used to assess cognitive, motor and language scores. Stimulation at home was assessed by the Pediatric Review of Children’s Environmental Support and Stimulation (PROCESS) tool. PROCESS scores were categorized into three groups: < Mean-1SD (low stimulation); Mean ± 1 SD (moderate stimulation) and > mean + 1SD (high stimulation).

Results

A total of 516 infants were available for neurodevelopment assessments. Interactions were observed between length for age z-score (LAZ) and PROCESS score categories. In the low stimulation group, the adjusted regression coefficients for the association between LAZ and cognitive, motor and language scores were substantially higher than in the moderate and high stimulation group. Stimulation was positively associated with neurodevelopmental outcomes in both stunted and non-stunted infants; however, the association was twice as strong in stunted than in non-stunted.

Conclusion

Moderate to high quality stimulation may alleviate the risk of sub-optimal development in LBW infants with linear growth deficits.

Clinical trial registration

The primary trial whose data are analysed is registered at clinicaltrials.gov (https://clinicaltrials.gov/ct2/show/NCT02631343).

Peer Review reports

What is already known on this topic?

  • Linear growth and quality of stimulation and nurturance are independently known to influence neurodevelopment, especially in children born with low birth weight (LBW).

  • The extent to which growth and stimulation influence each other’s association with cognitive, motor and language scores is unknown.

What this study adds?

  • High quality stimulation and nurturance could protect LBW infants with growth deficits from poor development scores.

  • Association of stimulation with neurodevelopmental outcomes was twice as strong in stunted than in non-stunted infants.

Introduction

The first 1000 days i.e., from conception through age 24 months, are foundational for brain development [1]. Both adverse and positive experiences during this period may critically shape children’s developmental trajectories [2, 3]. Children born with low birth weight (LBW) are at risk of linear growth faltering, cognitive and motor deficits as well as lower academic performance and behavioural problems compared to their normal birth weight counterparts [4,5,6,7]. Linear growth faltering in the first 2 years of life has been shown to be negatively associated with cognitive performance in childhood [8, 9]. There is also strong evidence that a child’s positive home environment reflected as responsive caregiving and opportunities for early learning, can promote development [10,11,12].

Less is known on whether linear growth and quality of stimulation/responsive caregiving at home influence each other’s association with cognitive, motor and language scores. Using a sample of 513 infants from rural India, Black et al. showed that a nurturant home environment attenuated associations between linear growth and fine motor and receptive language development [13]. Similarly, another multicentre study from Burkina Faso, Ghana and Malawi did not detect significant association between linear growth faltering and child development in the context of a high-quality developmental stimulation [14]. A study from rural Vietnam noted a modest beneficial effect of early child development interventions on cognition among children with declining height-for-age Z-scores or those that were stunted [15]. These findings indicate that in the presence of an environment characterized by nurturance and learning opportunities, children with low length-for- age z score (LAZ) can acquire developmental skills at the same level as their peers. Contrasting these findings, recent studies from Malaysian and Jamaican infants found no significant influence of home environment quality on the association between LAZ status and cognitive outcomes [16, 17]. More evidence is required on the interactive effects of linear growth and home environment in relation to developmental outcomes, particularly for the vulnerable subset of LBW infants. Further, evidence is required on whether in a setting with socio-economic constraints, a moderate to high-quality home environment can protect LBW infants with growth deficits from poor development scores and whether there is a differential effect of stimulation on developmental outcomes based on whether the LBW infant is stunted or not. The present analysis was aimed at providing insights on these pertinent issues of global importance.

Methods

Study design and participants

This secondary data analysis was conducted using data from an individually randomized controlled trial (RCT) aimed to evaluate the effect of community-initiated Kangaroo Mother Care (ciKMC) on neurodevelopmental outcomes of infants born low birth weight at 12 months of corrected age (ClinicalTrials.gov identifier NCT02631343) [18]. The study was conducted in resource constrained settings of rural and semi-urban Haryana, North India. In this study population, ciKMC was not associated with the neurodevelopment measures at 12 months [18]. A total of 552 stable preterm or small for gestational age term infants identified within 72 hours of birth and weighing between 1500 and 2250 g were included in the trial and followed up till 12 months of age. In the primary trial, infants weighing between 1500 and 1800 g, as per the government recommendations, were initially referred to a health facility for evaluation. These infants were considered for inclusion only if the families refused to take the baby to the health facility, or if the baby was taken but the medical doctor/paediatrician did not recommend admission or if admission was done, it was for less than 72 hours [18]. The primary trial excluded infants who were unable to feed, had difficulty in breathing, had less than normal movements and those with gross congenital malformations. As this was a trial assessing the efficacy of Kangaroo Mother Care (KMC) initiated at home/community, those infants who had KMC initiated at the health facility were excluded [18].

Details of the trial have been published elsewhere [18, 19]. Ethical clearances for the primary trial were obtained from the Institutional Ethics Review Committee of Society for Applied Studies, New Delhi (SAS/ERC/KMC-GCC/2015), the World Health Organization (WHO) Ethics Review Committee, Geneva (ERC0002629) and the Regional Committee for Medical and Health Research Ethics in Norway. In the primary trial, written informed consent was obtained from all subjects and/or their legal guardian(s).

Exposure and outcomes

Baseline information was collected on maternal and paternal age and education, birth order, parity and sex of the infant. Gestational age was documented from an ultrasound report, hospital records or maternal recall, whichever was available, in the given order of preference. The wealth of the family was determined by an index created through a principal component analysis based on household assets [20]. Information on vital status, illnesses (including any hospitalization) along with anthropometric measurements (weight and length) were captured by an independent trained team during their home visits at infant age 1, 3, 6 and 12 months. Caregivers were asked about illness (es) and hospitalization(s) in the 2 weeks preceding the visit. Length was measured using infantometers reading to the nearest 0.1 cm. Exclusivity of breastfeeding was assessed at 1, 3 and 6 months of infant age through a structured questionnaire.

Developmental outcomes were ascertained in the study clinic by trained psychologists using the Bayley Scales of Infant and Toddler Development, 3rd Edition (BSID-III) at 12 months of corrected age [21]. The BSID-III was adapted for use in the study setting. Details of the adaptation have been provided elsewhere [18]. Child stimulation at home was assessed at 12 months of age by trained psychologists using “Pediatric Review of Children’s Environmental Support and Stimulation (PROCESS)” questionnaire [22,23,24]. PROCESS was created for use with parents of children 2–18 months of age and can be administered in a clinic or in a home setting [22]. It consists of three components: a parent questionnaire, clinical observation, and a toy checklist. The parent questionnaire includes 24 items about the physical environment, household organization, and stimulation practices for development. The 20 observational items focus primarily on the emotional quality of parent-child interactions and the toy checklist consists of 40 items. Total scores are summed across the three sections [22]. Higher scores reflect better stimulation and support to infants. PROCESS scores have been shown to have a good correlation (r = 0.84) with the most widely used measure of the household environment i.e., Home Observation for Measurement of the Environment (HOME) scores [23, 24].

Plan of analyses

All analyses were done using STATA version 16.0 and R version 3.3.3 (2017-03-06). Baseline characteristics were summarized as mean (SD) or proportion. Length-for-age z score (LAZ) was calculated based on the WHO Child Growth Standards, using the zanthro package in STATA [25]. Stunting was defined as LAZ < -2, based on the standard WHO definition [25]. Length measurements were done at 1, 3 and 6 and 12 months of infant age. For this analysis, we preferred to use the length measurements at 6 months instead of 12 months as we wanted to look at the interactions in a cohort approach rather than cross sectionally. Another related premise for adopting such an approach in mid-infancy was that if we could show that linear growth and stimulation at home interacted with each other and influenced each other’s association with neurodevelopment outcomes at 12 months of age, this would provide a reasonable time frame for the caregivers with infants having growth failure to invest in stimulation at home for improving their child’s neurodevelopment. PROCESS scores, reflecting stimulation environment at home, were categorized into three groups: Low stimulation group (< Mean-1SD); moderate stimulation group (Mean ± 1 SD) and high stimulation group (> Mean + 1SD). The mean (SD) PROCESS score was 124 (18).

Neurodevelopmental outcomes consisted of cognitive, motor and language composite scores assessed by BSID-III at 12 months of corrected age. We first measured the association of LAZ at 6 months and PROCESS scores with scores obtained on BSID-III. We selected covariates for adjustment in the model based on their biological plausibility to influence the exposure and the outcomes and purposive selection principle i.e., covariates that brought at least 15% change in the univariate beta-coefficient were included in the multivariable model [26, 27].

We assessed the interaction between LAZ scores at 6 months of age and the PROCESS scores using likelihood ratio test comparing models with and without interaction terms. Analyses were stratified following the identification of a possible interaction. We initially did a screening where a P-value for interaction of less than 0.20 was investigated further [28]. The investigation was focussed on examining whether the magnitude of association between LAZ and outcome(s) of interest differed between the subgroups based on PROCESS score categories. Stratified results were presented at differing levels of PROCESS scores (low, moderate and high stimulation). For each of the categories of PROCESS score, we used linear regression with the composite scores for cognition, motor or language as an outcome and LAZ score as the exposure variable. Selection of variables for adjustment in the models was based on biological plausibility and purposive selection principle [26, 27].

Similarly, to assess whether the association between PROCESS scores and neurodevelopmental outcomes was modified by whether the babies were stunted or not, the interaction between the PROCESS score categories and stunting status was assessed using likelihood ratio test comparing models with and without interaction terms. In instances where the P-value of interaction was less than 0.20, the analyses were stratified and the effect sizes for the association between PROCESS categories and outcome(s) of interest were presented by the stunting categories. We used generalized additive models (GAM) in the mgcv package in R statistical package to depict non-linear associations between PROCESS score, LAZ and outcome scores (composite cognitive, motor and language scores) [29].

Ethics approval

No ethical approval was required for this secondary data analysis. However, the authors obtained written permission from the principal investigator of the primary trial to use the data for this secondary analysis.

Results

Characteristics of the sample

The primary trial enrolled 552 infants of which 516 infants had their neurodevelopment assessment at 12 months of age. The remaining 36 infants either died (n = 29) or the families had moved out of the study area (n = 7). Baseline characteristics of the 516 infants that were included in this analysis have been presented in Table 1. Supplementary Table 1 presents the comparison of baseline variables between infants with neurodevelopment assessment at 12 months of age and those that did not have the assessment and indicates no statistically significant differences. The infants studied belonged to economically constrained settings as reflected by some of the indicators such as proportion below poverty line (around 23%; national figure of around 15%) and median yearly family income (1316 USD; for some of the developed countries like United States, of around 67,000 USD) [30, 31].

Table 1 Baseline characteristics of the infants included in this secondary data analysis (N = 516)
Full size table

The mean (SD) composite cognitive, motor and language scores of the sample were 102.1 (11.8), 90.2 (10.4) and 84.9 (9.1) respectively. A total of 52.5% (271/516) of the infants were stunted at 6 months of age. As the exposures of interest i.e., linear growth at 6 months of age and PROCESS scores at 12 months of age were measured after the original intervention (ciKMC) was delivered, we attempted to understand whether ciKMC influenced these exposures. The mean (SD) PROCESS score at 12 months of age was statistically similar in the intervention [123.0 (16.6)] and control [125.0 (16.5)] groups (P = 0.16). Further, the mean (SD) LAZ at 6 months of age was also statistically similar in the intervention [− 2.12 (1.04)] and control [− 2.09 (1.06)] groups (P = 0.72). The ciKMC intervention did not have any significant association with the cognitive, language and motor outcomes at 12 months of adjusted age [18].

LAZ, PROCESS score and cognitive outcome

LAZ and PROCESS scores were associated with cognitive scores (Table 2).

Table 2 Association of length for age z score (LAZ) and PROCESS score with cognitive, motor and language scores at 12 months of corrected age (N = 516)
Full size table

There was an interaction between LAZ and PROCESS score categories for the cognitive composite score (P = 0.08) (Table 3).

Table 3 Association between length for age z score and neurodevelopmental outcomes, by PROCESS score categories
Full size table

In the low stimulation group, the adjusted regression coefficient (b = 3.63, 95% CI; 1.22, 6.03) was substantially higher than in the moderate stimulation group (b = 1.41, 95% CI; 0.25, 2.56) and the high stimulation group (b = 1.69, 95% CI; − 1.15, 4.52) (Table 3). The GAM plot supports the findings obtained in regression models (Fig. 1). The GAM plot shows that at lower PROCESS scores, the cognitive scores tend to decrease with decrease in LAZ scores whereas at higher PROCESS scores, the relation between LAZ and cognitive score has low variability. Further, with an increase in the PROCESS scores, the cognitive scores increased, more so in those with LAZ less than − 2 SD. An interaction was observed between stunting and PROCESS score categories (Table 4).

Fig. 1
figure 1

GAM plot depicting the relationship between length-for-age Z score, PROCESS score and cognitive, motor and language composite score

Full size image
Table 4 Association between PROCESS score categories and neurodevelopmental outcomes, by stunting status
Full size table

In both stunted and non-stunted infants, PROCESS scores were associated with cognitive scores with a clear dose response relationship (Table 4). The adjusted regression coefficient was comparatively higher in stunted infants.

LAZ, PROCESS score and motor outcome

LAZ and PROCESS scores were associated with motor scores (Table 2). There was an interaction between LAZ and PROCESS score categories for the motor composite score (P = 0.03) (Table 3). In the low stimulation group, the adjusted regression coefficient (b = 4.08, 95% CI; 1.69, 6.46) was higher than in the moderate stimulation (b = 1.54, 95% CI; 0.50, 2.58) and the high stimulation group (b = 1.05, 95% CI; − 1.14, 3.25) (Table 3). The GAM plot confirmed the findings obtained in regression models (Fig. 1). An interaction was observed between stunting and PROCESS score categories (Table 4). In stunted infants, PROCESS scores were associated with motor composite scores with a dose response relationship. In non-stunted infants, the adjusted regression coefficient was comparatively lower and did not reach statistical significance.

LAZ, PROCESS score and language outcome

LAZ and PROCESS scores were associated with language scores (Table 2). A potentially relevant interaction (P = 0.12) was observed between LAZ and PROCESS score categories (Table 3). In the low stimulation group, the adjusted regression coefficient (b = 2.47, 95% CI; 0.56, 4.38) was substantially higher than in the moderate stimulation (b = 1.02, 95% CI; 0.21, 1.86) and high stimulation group (b = 0.40, 95% CI; − 1.78, 2.58) (Table 3). The GAM plot confirmed the findings obtained in regression models (Fig. 1). An interaction was observed between stunting and PROCESS score categories (P = 0.05) (Table 4). In both stunted and non-stunted infants, PROCESS scores were associated with language scores with a distinct dose response relationship. The adjusted regression coefficient was comparatively higher in stunted infants.

Discussion

The current analyses aimed at providing answers to questions with programmatic implications, specifically whether within a setting with socio-economic constraints, a moderate to high-quality home environment can alleviate the risk of low development scores in LBW infants with linear growth deficits. We observed a weakening of the association between growth deficits and negative neurodevelopment outcome with increase in stimulation and nurturance at home. Additionally, we also observed that while stimulation at home was associated with neurodevelopmental outcomes in both stunted and non-stunted infants, the association was stronger in stunted than non-stunted infants.

Our findings corroborate the studies done in Bangladesh, Vietnam and in African settings (Burkina Faso, Malawi and Ghana) where the authors noted that among non-low birth weight children, a nurturant home environment attenuated the association between linear growth and neurodevelopmental outcomes [13,14,15]. However, findings contrast with the results of the studies among the Malaysian and Jamaican children where no significant influence of home environment quality on the association between LAZ status and cognitive outcomes was noted. The observed difference might be due to fairly smaller sample sizes in these studies, thereby reducing the power to detect significant interactions [16, 17].

There is lack of consensus with regards to the consideration of P-value to indicate presence of an interaction. While some investigators propose to adhere to the conventional P-value of < 0.05, others suggest that usually the power to test for interactions is low in many epidemiologic studies and therefore, testing for interaction tests based solely on P-value of < 0.05 may be misleading and could probably miss out important effect modifications [28, 32,33,34]. Based on this consideration, the suggestion is to increase the type 1 error rate to 20% while assessing tests of interaction [28]. Some researchers argue that consideration of a P-value for interaction tests is a part of the entire spectrum of information to be utilized in the assessment of effect modification and other components should be considered such as stratum-specific measures and prior biological knowledge [35, 36]. In our study, we considered a P-value of less than 0.20 to further investigate for potential interaction. Subsequently, we placed emphasis on the magnitude of effect size within the subgroups and attempted to make careful interpretations. Our findings were also supported by the GAM plots that depicted non-linear relationships between LAZ, PROCESS score and neurodevelopment outcomes.

There are strengths and limitations of this secondary data analyses. The data utilized is from a robust and well conducted randomized controlled trial with very low attrition. The measurements, including the anthropometry, and outcome data were collected by trained and standardized study team. One of the limitations is that the study lacks reliable data on gestational age. Weight was measured within 72 hours of birth by trained study team and inclusion of infants with weight between 1500 and 2250 g meant that these infants would be either preterm or term small for gestational age. Therefore, the findings could be extended only to a specific population of LBW infants i.e., stable late preterm or term small for gestational age (SGA) infants. There could also be a possibility that in babies with poorer linear growth or smaller attained length at 6 months of age or rather the factors that lead to such growth faltering may lead to poorer home stimulation which is measured 6 months later by PROCESS. Measurement of home stimulation at one time point only i.e., at 12 months is also a limitation. We also acknowledge that this being an observational study, the results may be affected by unmeasured confounding.

Our findings support the promotion of stimulation to LBW infants in order to offset the negative effect of growth faltering on neurodevelopmental outcomes. It is likely that every child will benefit from this strategy and therefore, future studies should test this approach in normal/non-high-risk children as well. Our findings indicate that through focusing only on nutrition for growth, we may miss to capitalize the important developmental effects of early child stimulation and responsive caregiving. The findings call for a comprehensive approach with nutrition and nurturing care at the forefront. This approach underlies the comprehensive framework of Nurturing Care that incorporates health, nutrition, responsive caregiving, opportunities for early learning, and child protection as a way to help children not only survive but also thrive [37].

Conclusion

The findings suggest that a moderate to high-quality stimulation at home may alleviate the risk of poor development scores in LBW infants with linear growth deficits. Early child stimulation may particularly be beneficial for LBW infants with linear growth deficits/stunting. Efforts for improving child development should be comprehensive with promotion of adequate nutrition and optimal nurturing care as integral components.

Availability of data and materials

The dataset analysed in the present study is publicly available at: https://figshare.com/s/c932d11ff5101e2268bf.

Abbreviations

LBW:

Low birth weight

PROCESS:

Pediatric Review of Children’s Environmental Support and Stimulation tool

LAZ:

Length for age Z score

RCT:

Randomized controlled trial

CiKMC:

Community initiated Kangaroo Mother Care

HBPNC:

Home Based Post Natal Care

BSID-III:

Bayley Scales of Infant and Toddler Development, 3rd Edition

HOME:

Home Observation for Measurement of the Environment

GAM:

Generalized additive models

SGA:

Small for gestational age

AGA:

Appropriate for gestational age

References

  1. Fox SE, Levitt P, Nelson CA 3rd. How the timing and quality of early experiences influence the development of brain architecture. Child Dev. 2010;81(1):28–40.

    Article  Google Scholar 

  2. Wachs TD, Georgieff M, Cusick S, McEwen BS. Issues in the timing of integrated early interventions: contributions from nutrition, neuroscience, and psychological research. Ann N Y Acad Sci. 2014;1308:89–106.

    Article  CAS  Google Scholar 

  3. Bick J, Nelson CA. Early adverse experiences and the developing brain. Neuropsychopharmacology. 2016;41(1):177–96.

    Article  CAS  Google Scholar 

  4. Upadhyay RP, Naik G, Choudhary TS, Chowdhury R, Taneja S, Bhandari N, et al. Cognitive and motor outcomes in children born low birth weight: a systematic review and meta-analysis of studies from South Asia. BMC Pediatr. 2019;19(1):35.

    Article  Google Scholar 

  5. Arpi E, Ferrari F. Preterm birth and behaviour problems in infants and preschool-age children: a review of the recent literature. Dev Med Child Neurol. 2013;55(9):788–96.

    Article  Google Scholar 

  6. Weindrich D, Jennen-Steinmetz C, Laucht M, Schmidt MH. Late sequelae of low birthweight: mediators of poor school performance at 11 years. Dev Med Child Neurol. 2003;45(7):463–9.

    Article  Google Scholar 

  7. Danaei G, Andrews KG, Sudfeld CR, Fink G, McCoy DC, Peet E, et al. Risk factors for childhood stunting in 137 developing countries: a comparative risk assessment analysis at global, regional, and country levels. Plos Med. 2016;13(11):e1002164.

    Article  Google Scholar 

  8. Sudfeld CR, McCoy DC, Danaei G, Fink G, Ezzati M, Andrews KG, et al. Linear growth and child development in low- and middle-income countries: a meta-analysis. Pediatrics. 2015;135(5):e1266–75.

    Article  Google Scholar 

  9. Adair LS, Fall CH, Osmond C, Stein AD, Martorell R, Ramirez-Zea M, et al. Associations of linear growth and relative weight gain during early life with adult health and human capital in countries of low and middle income: findings from five birth cohort studies. Lancet. 2013;382(9891):525–34.

    Article  Google Scholar 

  10. Hartinger SM, Lanata CF, Hattendorf J, Wolf J, Gil AI, Obando MO, et al. Impact of a child stimulation intervention on early child development in rural Peru: a cluster randomised trial using a reciprocal control design. J Epidemiol Community Health. 2017;71(3):217–24.

    Article  Google Scholar 

  11. Yousafzai AK, Obradović J, Rasheed MA, Rizvi A, Portilla XA, Tirado-Strayer N, et al. Effects of responsive stimulation and nutrition interventions on children's development and growth at age 4 years in a disadvantaged population in Pakistan: a longitudinal follow-up of a cluster-randomised factorial effectiveness trial. Lancet Glob Health. 2016;4(8):e548–58.

    Article  Google Scholar 

  12. Worku BN, Abessa TG, Wondafrash M, Lemmens J, Valy J, Bruckers L, et al. Effects of home-based play-assisted stimulation on developmental performances of children living in extreme poverty: a randomized single-blind controlled trial. BMC Pediatr. 2018;18(1):29. Published 2018 Feb 5. https://doi.org/10.1186/s12887-018-1023-0.

    Article  PubMed  PubMed Central  Google Scholar 

  13. Black MM, Yimgang DP, Hurley KM, Harding KB, Fernandez-Rao S, Balakrishna N, et al. Mechanisms linking height to early child development among infants and preschoolers in rural India. Dev Sci. 2019;22(5):e12806.

    Article  Google Scholar 

  14. Prado EL, Abbeddou S, Adu-Afarwuah S, Arimond M, Ashorn P, Ashorn U, et al. Linear growth and child development in Burkina Faso, Ghana, and Malawi. Pediatrics. 2016;138(2):e20154698. https://doi.org/10.1542/peds.2015-4698.

    Article  PubMed  Google Scholar 

  15. Watanabe K, Flores R, Fujiwara J, Tran LT. Early childhood development interventions and cognitive development of young children in rural Vietnam. J Nutr. 2005;135(8):1918–25.

    Article  CAS  Google Scholar 

  16. Nurliyana AR, Mohd Shariff Z, Mohd Taib MN, Gan WY, Tan KA. Early growth and home environment are associated with cognitive development in the first year of life of Malaysian infants. Early Hum Dev. 2019;140:104890 [published online ahead of print, 2019 Oct 23].

    Article  Google Scholar 

  17. Walker SP, Grantham-Mcgregor SM, Powell CA, Chang SM. Effects of growth restriction in early childhood on growth, IQ, and cognition at age 11 to 12 years and the benefits of nutritional supplementation and psychosocial stimulation. J Pediatr. 2000;137(1):36–41.

    Article  CAS  Google Scholar 

  18. Taneja S, Sinha B, Upadhyay RP, Mazumder S, Sommerfelt H, Martines J, et al. Community initiated kangaroo mother care and early child development in low birth weight infants in India- a randomized controlled trial. BMC Pediatr. 2020;20:150. https://doi.org/10.1186/s12887-020-02046-4.

    Article  PubMed  PubMed Central  Google Scholar 

  19. Mazumder S, Taneja S, Dube B, Bhatia K, Ghosh R, Shekhar M, et al. Effect of community-initiated kangaroo mother care on survival of infants with low birthweight: a randomised controlled trial. Lancet. 2019;394(10210):1724–36.

    Article  Google Scholar 

  20. Vyas S, Kumaranayake L. Constructing socio-economic status indices: how to use principal components analysis. Health Policy Plan. 2006;21(6):459–68.

    Article  Google Scholar 

  21. Bayley N. Bayley scales of infant and toddler development: administration manual. 3rd ed. United States of America: Psychorp; 2006.

    Google Scholar 

  22. Fernald LCH, Prado EL, Kariger PK, Raikes A. A toolkit for measuring early childhood development in low and middle income countries (English). Washington, D.C.: World Bank Group; 2017. Available at: http://documents.worldbank.org/curated/en/384681513101293811/A-toolkit-for-measuring-early-childhood-development-in-low-and-middle-income-countries

    Book  Google Scholar 

  23. Casey PH, Bradley RH, Nelson JY, Whaley SA. The clinical assessment of a child's social and physical environment during health visits. J Dev Behav Pediatr. 1988;9(6):333–8.

    Article  CAS  Google Scholar 

  24. Casey PH, Barrett K, Bradley RH, Spiker D. Pediatric clinical assessment of mother-child interaction: concurrent and predictive validity. J Dev Behav Pediatr. 1993;14(5):313–7.

    Article  CAS  Google Scholar 

  25. Onis M. WHO child growth standards based on length/height, weight and age. Acta Paediatr. 2006;95:76–85.

    Google Scholar 

  26. Hosmer DW, Lemeshow S, Sturdivant RX. Applied logistic regression. New York: Wiley; 2013.

    Book  Google Scholar 

  27. Bursac Z, Gauss CH, Williams DK, Hosmer DW. Purposeful selection of variables in logistic regression. Source Code Biol Med. 2008;3:17. https://doi.org/10.1186/1751-0473-3-17.

    Article  PubMed  PubMed Central  Google Scholar 

  28. Selvin S. Statistical analysis of epidemiologic data. 3rd ed. New York: Oxford University Press; 1996. p. 213–4.

    Google Scholar 

  29. Wood SN. Generalized additive models: an introduction with R. 2nd ed. New York: Chapman and Hall/CRC, Taylor and Francis Group; 2017.

    Book  Google Scholar 

  30. Newhouse D, Vyas P. Estimating poverty in India without expenditure data: A survey-to-survey imputation approach. Available at: https://documents1.worldbank.org/curated/en/866871560283271368/pdf/Estimating-Poverty-in-India-without-Expenditure-Data-A-Survey-to-Survey-Imputation-Approach.pdf

  31. Income and poverty in the United States: 2020. Available at: https://www.census.gov/library/publications/2021/demo/p60-273.html#:~:text=Median%20household%20income%20was%20%2467%2C521,median%20household%20income%20since%202011.

  32. Durand CP. Does raising type 1 error rate improve power to detect interactions in linear regression models? A simulation study. Plos One. 2013;8(8):e71079.

    Article  CAS  Google Scholar 

  33. Marshall SW. Power for tests of interaction: effect of raising the type I error rate. Epidemiol Perspect Innov. 2007;4:4.

    Article  Google Scholar 

  34. McClelland GH, Judd CM. Statistical difficulties of detecting interactions and moderator effects. Psychol Bull. 1993;114(2):376–90.

    Article  CAS  Google Scholar 

  35. Poole C. Low P-values or narrow confidence intervals: which are more durable? Epidemiology. 2001;12(3):291–4.

    Article  CAS  Google Scholar 

  36. Greenland S, Rothman KJ. Chapter 18: Concepts of Interaction. In: Modern Epidemiology 2 Edition (Edited by: Rothman KJ, Greenland S). New York: Lippincott-Raven; 1998. p. 329–42.

    Google Scholar 

  37. World Health Organization, United Nations Children’s Fund, World Bank Group. Nurturing care for early childhood development: a framework for helping children survive and thrive to transform health and human potential. Geneva: World Health Organization; 2018. Licence: CC BY-NC-SA 3.0 IGO. Available at: https://apps.who.int/iris/bitstream/handle/10665/272603/9789241514064-eng.pdf?ua=1

    Google Scholar 

Download references

Acknowledgments

None.

Funding

Open access funding provided by University of Bergen. The original study on which the current manuscript is based was funded by Grand Challenges Canada (Grant Number 0725–03) and the Research Council of Norway (RCN) through its Centers of Excellence Scheme (project number 223269) and by the University of Bergen through funding to the Centre for Intervention Science in Maternal and Child Health (CISMAC). The funding agency had no role in the designing the study; the collection, analysis, and interpretation of data; the writing of the manuscript; and the decision to submit the paper for publication.

The authors have no financial relationship relevant to this article to disclose RB and TD are staff members of the World Health Organization. The authors alone are responsible for the views expressed in this article and they do not necessarily represent the decisions, policy or views of the World Health Organization.

Author information

Authors and Affiliations

  1. Centre for Health Research and Development, Society for Applied Studies, 45, Kalu Sarai, New Delhi, 110016, India

    Ravi Prakash Upadhyay, Sunita Taneja, Sarmila Mazumder & Nita Bhandari

  2. Department of Global Public Health and Primary Care, University of Bergen, Bergen, Norway

    Ravi Prakash Upadhyay

  3. Department of Research, Innlandet Hospital Trust, Lillehammer, Norway

    Tor A. Strand

  4. Centre for Intervention Science in Maternal and Child Health (CISMAC), Department of Global Public Health and Primary Care, University of Bergen, Bergen, Norway

    Halvor Sommerfelt & Jose Martines

  5. Cluster for Global Health, Division for Health Services, Norwegian Institute of Public Health, Oslo, Norway

    Halvor Sommerfelt

  6. Department of Psychosocial Science, Faculty of Psychology, University of Bergen, Bergen, Norway

    Mari Hysing

  7. Department of Mental Health and Substance Use, World Health Organization, Geneva, Switzerland

    Tarun Dua

  8. Center for Effective Global Action (CEGA), School of Public Health, University of California, Berkeley, USA

    Patricia Kariger

  9. Department of Maternal, Newborn, Child and Adolescent Health, World Health Organization, Geneva, Switzerland

    Rajiv Bahl

Authors
  1. Ravi Prakash Upadhyay
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Sunita Taneja
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Tor A. Strand
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Halvor Sommerfelt
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Mari Hysing
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Sarmila Mazumder
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Nita Bhandari
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Jose Martines
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Tarun Dua
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Patricia Kariger
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Rajiv Bahl
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

RPU, ST, TAS and MH were involved in conceptualization, data acquisition, analysis, interpretation of data and drafting of the manuscript. ST, NB and SM were involved in designing and conducting the primary trial. NB and SM also provided administrative and technical support for this secondary data analysis exercise. HS, JM, TD, PK and RB provided critical feedback on the statistical analysis and drafting of the manuscript. All authors critically reviewed the manuscript, provided important intellectual content and approved the final manuscript as submitted. The authors agree to be accountable for all aspects of the work.

Corresponding author

Correspondence to Ravi Prakash Upadhyay.

Ethics declarations

Ethics approval and consent to participate

No ethical approval was required for this secondary data analysis. However, the authors obtained written permission from the principal investigator of the primary trial to use the data for this secondary analysis. Ethical clearances for the primary trial were obtained from the Institutional Ethics Review Committee of Society for Applied Studies, New Delhi, the World Health Organization (WHO) Ethics Review Committee, Geneva and the Regional Committee for Medical and Health Research Ethics in Norway.

Consent for publication

Not applicable.

Competing interests

No competing interest to declare.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Supplementary Table 1.

Comparison of baseline characteristics of the infants with neurodevelopment data at 12 months (N = 516) and those that did not have (N = 36).

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Upadhyay, R.P., Taneja, S., Strand, T.A. et al. Early child stimulation, linear growth and neurodevelopment in low birth weight infants. BMC Pediatr 22, 586 (2022). https://doi.org/10.1186/s12887-022-03579-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12887-022-03579-6

Keywords

BMC Pediatrics

ISSN: 1471-2431

Contact us