This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Research Protocols, is properly cited. The complete bibliographic information, a link to the original publication on http://www.researchprotocols.org, as well as this copyright and license information must be included.
The complementary feeding period is a time of unparalleled dietary change for every human, during which the diet changes from one that is 100% milk to one that resembles the usual diet of the wider family in less than a year. Despite this major dietary shift, we know relatively little about food and nutrient intake in infants worldwide and virtually nothing about the impact of baby food “pouches” and “baby-led weaning” (BLW), which are infant feeding approaches that are becoming increasingly popular. Pouches are squeezable containers with a plastic spout that have great appeal for parents, as evidenced by their extraordinary market share worldwide. BLW is an alternative approach to introducing solids that promotes infant self-feeding of whole foods rather than being fed purées, and is popular and widely advocated on social media. The nutritional and health impacts of these novel methods of infant feeding have not yet been determined.
The aim of the First Foods New Zealand study is to determine the iron status, growth, food and nutrient intakes, breast milk intake, eating and feeding behaviors, dental health, oral motor skills, and choking risk of New Zealand infants in general and those who are using pouches or BLW compared with those who are not.
Dietary intake (two 24-hour recalls supplemented with food photographs), iron status (hemoglobin, plasma ferritin, and soluble transferrin receptor), weight status (BMI), food pouch use and extent of BLW (questionnaire), breast milk intake (deuterium oxide “dose-to-mother” technique), eating and feeding behaviors (questionnaires and video recording of an evening meal), dental health (photographs of upper and lower teeth for counting of caries and developmental defects of enamel), oral motor skills (questionnaires), and choking risk (questionnaire) will be assessed in 625 infants aged 7.0 to 9.9 months. Propensity score matching will be used to address bias caused by differences in demographics between groups so that the results more closely represent a potential causal effect.
This observational study has full ethical approval from the Health and Disability Ethics Committees New Zealand (19/STH/151) and was funded in May 2019 by the Health Research Council (HRC) of New Zealand (grant 19/172). Data collection commenced in July 2020, and the first results are expected to be submitted for publication in 2022.
This large study will provide much needed data on the implications for nutritional intake and health with the use of baby food pouches and BLW in infancy.
Australian New Zealand Clinical Trials Registry ACTRN12620000459921; http://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=379436.
DERR1-10.2196/29048
The biggest dietary change in every human’s life is the change from consuming a 100% milk diet in the first months of life to consuming a diet that is broadly the same as that of the rest of the family by the first birthday. This change is large. In fact, an infant following the current infant feeding guidelines globally [
Baby food pouches are squeezable containers with a plastic spout described as a “mess-free and easy alternative for baby food on the go” [
Although the foods offered in baby food pouches are broadly similar in content to those offered as baby foods in jars and cans, there has been some concern, albeit not universal [
First, these smooth highly processed products with multiple blended ingredients bear little resemblance to intact fruits and vegetables, and they are marketed well beyond the early weeks of complementary feeding when it might be argued a “super smooth” product is appropriate (eg, many are marketed for infants 8 months plus). This raises a number of questions. Do these products increase energy intake because they are so easy to eat (smooth consistency and do not require chewing; an entire 120-g serve can be accessed merely by squeezing the soft pouch and sucking)? Conversely, does the ease of consumption lead to displacement of other more nutrient-rich foods, such as breast milk (or infant formula), from the diet (eating “on the go” is unlikely to be consistent with New Zealand Ministry of Health recommendations that until 8 months of age infants are offered solid foods after milk to avoid displacing nutrient-rich milk [
Second, if these pouches are indeed being consumed “on the go” by infants, what implications does this have for learning about food and eating? For example, do infants who feed themselves by sucking a purée out of a pouch while “on the go” have the same relationship with food as that for infants who sit and eat with their family? The description of pouches “promoting self-feeding and independence” [
Third, what is the impact of prolonged exposure to these often sweet and acidic (and therefore presumably cariogenic) foods on erupting teeth? There is a strong relationship between the frequency of cariogenic food intake and childhood caries [
While the use of food pouches is starting to be investigated internationally [
These are just anecdotal reports, but they underline the urgent need to determine the effect of pouches on infant nutrition and health. Interestingly, given the lack of research in this area, some health professionals in Germany [
The second recent phenomenon in infant feeding is the popular adoption of BLW, an alternative approach to introducing solids to infants. In BLW, infants feed themselves all their foods from the start of the “complementary feeding” period. This means no spoon feeding by a parent, and only “finger foods” are offered [
Infant milk (breastmilk or infant formula) is a substantial component of the diet for infants during the complementary feeding period, providing more than half of their energy intake at 7 months of age [
The FFNZ study will determine the iron status, growth, food and nutrient intakes, breast milk intake, eating and feeding behaviors, dental health, oral motor skills, and choking risk of New Zealand infants, with a particular focus on the use of baby food pouches and BLW.
In infants aged 7.0 to 9.9 months, we will determine whether iron status and BMI z-score differ according to the extent of food pouch use and complementary feeding approach (BLW compared with TSF).
In infants aged 7.0 to 9.9 months, we will estimate the following: (1) Nutrient intake, nutrient adequacy, and foods of cultural importance in New Zealand infants and in infants fed using baby food pouches regularly or those following BLW; (2) Breast milk intake in New Zealand infants and in infants fed using baby food pouches regularly or those following BLW; (3) Prevalence and nature of food pouch use; (4) Prevalence of BLW; (5) How eating behaviors (ability to eat with appetite, speed of eating, and picky eating) and feeding behaviors (parental responsiveness to infant hunger and satiety cues) differ according to the extent of food pouch use and complementary feeding approach (BLW compared with TSF); (6) How dental health differs according to the extent of food pouch use and complementary feeding approach (BLW compared with TSF); (7) How oral motor skills differ according to the extent of food pouch use and complementary feeding approach (BLW compared with TSF); and (8) How the risk of choking differs according to the extent of food pouch use and complementary feeding approach (BLW compared with TSF).
The FFNZ study is an observational cross-sectional study of food and health in infants aged 7.0 to 9.9 months. The study will compare infants using baby food pouches with those not using these pouches, and those following BLW with those following TSF, while collecting data on nutrient intake and nutritional status in this age group in general. The age range has been chosen because it is close enough to when complementary feeding starts (usually 4-6 months of age) that we can expect to see large variations in both baby food pouch use and BLW rates, while also giving enough time from the start of complementary feeding for eating patterns to have had an impact on iron status and growth. A narrow age range has specifically been chosen because diet changes rapidly in infancy. Observational study designs are appropriate for identifying associations between behaviors as they are carried out in the “real world.” While a randomized controlled trial is required to determine causality, it is not ethical to randomize infants to follow BLW or pouch use because that would require randomization of participants to eating patterns that health professionals have concerns about [
In total, 625 parents/guardians who have an infant less than 9.9 months of age will be recruited from two regions of New Zealand (Dunedin and Auckland) to participate in the study when their infant is aged 7.0 to 9.9 months. Recruitment will occur by advertisement and word of mouth and will target all infants rather than those adopting BLW, TSF, or food pouch use. We aim to recruit a sample that is broadly representative of the ethnicity and socioeconomic status of New Zealand children. It is not feasible to recruit a truly representative sample using typical methods, such as electoral roll and door knocking, because they would identify very few infants in the narrow age band that is necessary for this study (because diet changes so rapidly in infancy). We will, however, collect data from a diverse range of ethnic and socioeconomic groups by (1) engaging with Māori and Pasifika community health organizations to assist with recruitment, (2) targeting recruitment in suburbs with a high proportion of Māori, Pasifika, and Asian populations, and (3) having research team members who have experience working with or who culturally identify with Māori, Pasifika, and Asian communities. We will also statistically weight the estimates to account for demographic disparities if appropriate. The study has ethical approval from the Health and Disability Ethics Committees New Zealand (19/STH/151), and written informed consent will be obtained at the first appointment. The study is registered with the Australian New Zealand Clinical Trials Registry (registration number: ACTRN12620000459921).
Our sample size calculation is based on comparing the BMI z-score and plasma ferritin concentration in infants following BLW and TSF, as there are currently no data internationally on these measures in pouch users. Our recent studies suggest that 29% of infants will meet the definition of BLW [
Participation in the study will involve three (participants in the main study) or five (participants in a consecutive subsample of breastfed infants) contacts over 2 weeks following recruitment. For the main study (n=625), the first main appointment will generally be held in the participant’s home and involve a 24-hour diet recall, completion of two questionnaires, and anthropometric measurements of the child. The second main appointment will generally take place at university research rooms, and involve a second 24-hour diet recall and photography of the infant’s teeth to assess dental health. A third main appointment will take place at our university rooms (Dunedin) or a local blood testing facility (Auckland) to collect a blood sample to measure the iron status. Finally, a self-administered questionnaire will be completed by the participants in their homes. For the subsample (n=150) involved in the measurement of breast milk intake, a stable isotope will be given at the first main appointment, with three additional saliva samples collected over the ensuing fortnight (the third sample being collected at the second main appointment).
At the initial appointment, ethnicity, maternal education, maternal work status, household deprivation (New Zealand deprivation index 18 [
This is related to primary objective 1 and secondary objectives 1, 2, 3, 5, 6, 7, and 8. We will measure the frequency of pouch use in the past month by a questionnaire. We intend to define infants as being frequent baby food pouch users if their parents state that they are currently being given food from a food pouch “5 to 6 times a week,” “once a day,” or “more than once a day,” although this may need to be modified when the distribution of intakes is determined (there are currently no published data on the frequency of baby food pouch use to base this cutoff on). We will collect data on the frequency of pouch use, use of “ready-to-eat” pouches versus “home-filled” pouches, extent to which the infant feeds directly from the pouch rather than being fed by spoon, types of foods given in “pouches,” contribution of pouch foods to total intake of solid foods, proportion of the “pouch” consumed on a typical eating occasion, duration of a typical eating occasion, physical situations in which pouches are used, proximity of an adult when the pouch is being used, reasons for using pouches rather than other methods of food delivery, and anything not liked about using baby food pouches. Key pouch questions will be asked referring to when the infant first started eating solids, when the infant was around 6 months of age, and “now.”
This is related to primary objective 1 and secondary objectives 1, 2, 4, 5, 6, 7, and 8. Parents will be asked to describe the way their infants were fed when they first started eating solids, when they were around 6 months of age, and “now,” using five answer options. Parents who choose “spoon fed by an adult” or “mostly spoon fed by an adult, some baby feeding themselves” will be classified as TSF. Parents who select “about half spoon feeding by an adult and half baby feeding themselves” will be classified as partial BLW [
This is related to primary objective 1. A nonfasting venipuncture blood sample will be collected at the third main appointment (3-mL EDTA anticoagulated vacutainer blood collection tube; Becton Dickinson and Company) to determine the plasma ferritin concentration and iron status defined using the body iron concentration (calculated using plasma ferritin and soluble transferrin receptor concentrations [
Iron status categories.
Category | Body iron value | Hemoglobin value | Plasma ferritin value |
Iron sufficient | ≥0 mg/kg | ≥105 g/L | ≥15 μg/L |
Iron depleted | ≥0 mg/kg | ≥105 g/L | <15 μg/L |
Early functional iron deficiency | <0 mg/kg | ≥105 g/L | N/Aa |
Iron deficiency anemia | <0 mg/kg | <105 g/L | N/A |
aN/A: not applicable.
As ferritin is an acute phase reactant and can be artificially elevated by inflammation, we will also analyze two inflammatory markers (C-reactive protein and α-1-acid glycoprotein) as recommended by the World Health Organization (WHO) [
Participants will be given verbal instructions on how to apply a local anesthetic (Ametop gel; Perstorp Pharma) and will be given access to an instruction video. The gel is to be applied to the insides of both of the infant’s elbows (this enables phlebotomy to be attempted on the second arm if necessary) and covered with an occlusive dressing at least 1 hour (no more than 4-6 hours) before the blood test appointment. It is removed after 30 to 45 minutes. The blood sample will be taken by a pediatric phlebotomist. The research team has extensive experience overseeing research projects involving the collection of venipuncture blood samples from infants and toddlers [
Commercial laboratories (Southern Community Laboratories, Dunedin, New Zealand and Labtests NZ, Auckland, New Zealand) will determine complete blood count (requires fresh blood) and plasma ferritin, so that the iron status can be immediately communicated to the infant’s general practitioner if the infant is identified as having anemia. The remaining plasma will be frozen at −80°C for batch analysis of soluble transferrin receptor, C-reactive protein, and α-1-acid glycoprotein concentrations at the University of Otago Department of Human Nutrition Laboratory at the end of the study [
This is related to primary objective 1. Infant weight will be measured at the initial appointment on an electronic scale (model 354; Seca) and length will be measured on a 99-cm measuring mat (model SE210; Seca) in duplicate following WHO protocols [
This is related to secondary objective 1. Information on infant nutrient intake and adequacy, food group intake, dietary patterns, and culturally important foods will be obtained using interviewer-administered multiple pass 24-hour recalls collected at the first and second appointments. The two 24-hour recalls take place on different days of the week to capture variation in intake between days. Collecting two 24-hour recalls will enable us to calculate “usual intake” using the multiple source method (MSM) for estimating usual dietary intake of individuals [
Particular focus will be on free sugars and added sugars given a recent small study [
Questionnaire and 24-hour recall data will also be used to determine the extent to which key indicators of diet quality are being met, as guided by the New Zealand Ministry of Health Eating and Activity Guidelines for New Zealand Infants and Toddlers [
This is related to secondary objectives 1 and 2. We will obtain accurate data on the amount of breast milk infants consume using the stable isotope (deuterium oxide) “dose-to-mother” technique [
This is related to secondary objective 5. Eating behaviors will be assessed using the following four subscales from the Children’s Eating Behavior Questionnaire (CEBQ) [
This is related to secondary objective 6. Photographs of the infant’s upper and lower teeth will be taken by trained interviewers using a dedicated study Oppo Reno2 Z (Oppo) mobile phone with a small portable Smile Light MDP lighting source specifically designed for taking dental pictures [
This is related to secondary objective 7. The questionnaires administered at the final appointment include the validated Child Oral Motor Proficiency Scale (ChOMPS) to identify oral motor and eating skill delay [
This is related to secondary objective 8. The questionnaire administered at the initial appointment will include questions on choking since birth. We developed these retrospective questions for previous work in this age group and have demonstrated that they provide data that are comparable to choking data collected prospectively using a daily choking calendar [
We expect that there will be crossover between BLW/TSF status and pouch use. We will be able to explore whether mean differences in energy and nutrient intake, as well as other measures, between pouch users and nonusers are different for those who use BLW and those who do not. These results will be stratified by BLW/TSF status, and estimated differences will be compared.
Regression models will be used to determine differences between groups. Propensity score matching will be undertaken to reduce the bias caused by differences in demographics between the groups (eg, maternal education and ethnicity), infant age (to the nearest week), and sex. Propensity score matching is not like traditional paired matching, where each individual is matched to another individual in the other group according to covariates. Instead, propensity score matching uses a participant’s propensity score (found using covariates) and estimates what their outcome (eg, energy intake) would have been if they were in the other of two dichotomous groups. By using this method, we expect the estimates will more closely represent a potential causal effect [
Estimates of BLW, frequent pouch use, nutrient intake, status, and adequacy will be calculated for the whole sample along with 95% CIs. If the sample is not demographically representative of the wider population, statistical weighting of these estimates will be undertaken using the survey command in Stata (StataCorp).
Nutrient intake will be determined using 24-hour recall data adjusted to provide estimates of usual intake (using the MSM method [
The best-fitting polynomials to predict breast milk intake will be estimated by regression models using fractional polynomial functions of variables, such as age, sex, body weight, and food and beverage intake (eg, kJ/day). This will result in equations that can be used to predict breast milk intake based on a variety of input variables. Ideally one based on data that can be collected in a single clinical appointment, and another that uses data requiring more extensive collection in a research or surveillance setting.
This observational study has full ethical approval from the Health and Disability Ethics Committees New Zealand (19/STH/151) and was funded in May 2019 by the Health Research Council (HRC) of New Zealand (grant 19/172). Data collection commenced in July 2020, and the first results are expected to be submitted for publication in 2022. Data collection will only take place while New Zealand is in Alert Levels 1 or 2 during the COVID-19 pandemic. The Otago and Auckland regions of New Zealand, where the data collection will take place, have been in Level 1 for all but 7 weeks in Otago (all at Level 2) and 11 weeks in Auckland (7 weeks at Level 2 and 4 weeks at Level 3) since July 2020, as overall case numbers in New Zealand remain extremely low (<2000 in a population of more than 5 million). As daily life is essentially normal in Level 1 with the exception of closed international borders and Level 2 just requires some physical distancing and appropriate hygiene recommendations, we feel confident that the pandemic will have relatively little effect on our data.
This large observational study will provide much needed data on nutrient intake (including breast milk intake) and nutritional status (specifically iron status, growth, and dental health) in a large diverse sample of New Zealand infants. However, our data will also have considerable international appeal given the lack of research assessing the implications for nutritional intake and health for those infants who obtain a large proportion of their food via baby food pouches. Similarly, determining how iron status, growth, nutrient intake, and choking risk may differ in infants following BLW compared with TSF is urgently warranted given the widespread interest in this alternative approach to complementary feeding worldwide.
Peer-reviewer reports and grant approval from Health Research Council of New Zealand.
Ethics approval.
adequate intake
baby-led weaning
Biomarkers Reflecting Inflammation and Nutritional Determinants of Anemia
estimated average requirement
First Foods New Zealand study
US Feeding Infants and Toddlers Study
multiple source method
traditional spoon feeding
World Health Organization
We acknowledge the wider members of the research team (Owen Mugridge, Rio Monzales, Hayley Dodd, and Glenna Paterson), the support of Southern Community Laboratories, and the participants. This study is supported by the Health Research Council (HRC) of New Zealand (19/172). RWT is supported by a Fellowship from the Karitane Products Society. NHM and KJB are supported by PhD scholarships from the Health Research Council of New Zealand, AC is supported by a PhD scholarship from the University of Otago, and MC is supported by a PhD scholarship from Massey University. The HRC had no role in study design, writing of the protocol manuscript, or the decision to submit the manuscript for publication.
RWT and ALMH are the co-principal investigators of the First Foods New Zealand study. RWT, CAC, KLB, PRvH, LAT, LD, JJH, AMM, LAH, and ALMH designed the project and applied for funding. RWT and ALMH produced the first and subsequent drafts of the manuscript. JJH advised on study design, sample size calculation, and statistical analysis. JA is the project coordinator. LD, NHM, AC, LT, MC, KB, EJ, IK, EF, and DA developed the study data collection protocols, and NHM, AC, LT, MC, KB, EJ, IK, MR, and JM undertook data collection. All authors made important intellectual contributions to the manuscript, and all have read and approved the final version.
None declared.