This study uses a longitudinal convergent parallel mixed methods design with both quantitative and qualitative components. The longitudinal framework captures developmental changes and the cumulative effects of diet over time. Quantitative measures include dietary intake, growth trajectories, bone health and body composition parameters, cognitive assessments, and a structured knowledge, attitudes, and practices (KAP) questionnaire, allowing statistical evaluation of how dairy intake and dairy perceptions relate to developmental outcomes. The qualitative component explores facilitators and barriers to dairy consumption among caregivers. Participants will be followed for 2.5 years, with assessments at 4 points approximately 10 months apart. Figure 1 presents the study timeline. This study protocol is reported in accordance with the ObsQual Protocol Checklists, a structured guidance tool for observational study protocols [26]. The checklist is available in Multimedia Appendix 1.

Children aged 5‐11 years were selected as the target population to capture a developmental stage characterized by steady linear growth, rapid bone mineral accrual, and meaningful cognitive advancement before the onset of puberty [2,27]. During this middle childhood period, biological maturation is variable but less pronounced than in adolescence, allowing clearer modeling of nutritional influences without the confounding effects of pubertal growth surges [28 ]. This age range also represents a stage in which caregivers continue to exert substantial control over children’s dietary behaviors, making it an optimal window for assessing both dietary exposures and caregiver-related influences [29 ].

The target population for this study consists of children aged 5‐11 years residing in Saskatoon, Canada. According to the 2016 Canadian census, Saskatoon had a total of 44,305 children aged 0 to 14 years, with 14,975 in the 5- to 9-year age group and 13,705 in the 10- to 14-year age group. The total child population consisted of 14,765 (51.5%) boys and 13,915 (48.5%) girls and included 6160 immigrant children [30].

The study sample size was determined to ensure sufficient statistical power to detect a clinically meaningful relationship between dairy intake and linear growth in children aged 5 to 11 years. Linear growth, defined as the change in height over time, served as the primary outcome. The study used a repeated-measures design with 4 data collection points at baseline, 10, 20, and 30 months. Based on prior literature and expert input, the analysis assumed a 2-sided α of .05% and 80% power. The residual SD of height was estimated at 0.5 cm, and the intrasubject correlation was set at 0.8, reflecting expected consistency in individual height measurements over time [31]. The SD for daily dairy intake was approximated at 1 serving per day [32,33]. A clinically meaningful effect was defined as an additional 0.2 cm of height gain over 30 months per extra daily serving of dairy, corresponding to approximately 0.0067 cm per month per serving. This estimate was informed by meta-analytic findings, suggesting a 0.4 cm per year increase in height associated with an additional 245 mL of milk consumed daily [34]. Using a power analysis approximation for a linear mixed-effects model that incorporates the effect of a continuous covariate (dairy intake) on growth rate, the required sample size was estimated at 87 participants. To account for an anticipated 25% attrition over the 30-month study period, the recruitment target was increased to 116 participants.

A combination of convenience and snowball sampling methods was used to recruit participants, reflecting the demographic characteristics of children in Saskatoon. Recruitment was monitored to maintain a balance between boys and girls, and demographic variables such as age and newcomer status were tracked to ensure proportional representation approximating the actual population. Inclusion criteria are (1) children aged 5 to 11 years, (2) healthy status (as determined by parental report and study screening), and (3) no current treatment for malnutrition or other serious medical conditions. Exclusion criteria include (1) chronic diseases or medical conditions affecting growth and development, (2) chronic medication use, and (3) medications known to affect bone metabolism or calcium balance. Retention efforts will rely on flexible scheduling, reminders, and small incentives.

The study is being conducted in Saskatoon, Canada, leveraging various recruiting channels, including libraries, recreational and community centers, social media, and online platforms, to ensure a diverse and representative sample. Assessments are being conducted in research facilities at the University of Saskatchewan.

Ethics approval has been obtained from the University of Saskatchewan Ethics Board (Bio-3339). Written informed consent was obtained from the parents or legal guardians of all participating children. During the consent process, parents or guardians received detailed information about the study’s purpose, procedures, potential risks, and benefits and had the opportunity to ask questions before signing the consent form. Age-appropriate assent was also obtained from each child to ensure understanding and voluntary participation. Children were informed that they could withdraw from the study at any time without consequence. To protect participant privacy and confidentiality, all data are stored on secure, access‑restricted servers housed at the University of Saskatchewan. Identifiable information is maintained separately from study data and replaced with unique coded identifiers. Only authorized members of the research team have access to the linkage file. Any published results will present aggregated findings only, ensuring that no individual participant can be identified. Parents and participating children will receive 1 gift card per follow-up visit as compensation, provided in accordance with institutional guidelines. Compensation is distributed uniformly to all participants and is not dependent on full study completion, which minimizes the risk of coercion or undue influence.

Physical activity is measured using both subjective and objective methods. For subjective assessment, the Physical Activity of Children and Youth module from the CCHS is administered [48]. This module forms part of the health components within the CCHS and collects information on physical activity behaviors among children and youths. It is used nationally to monitor movement behaviors and complements other health and nutrition data collected through the survey. To complement self-reported data, accelerometers are used to provide objective measures of physical activity. Worn on the hip for 7 consecutive days, these devices record continuous data on movement patterns, including time spent in sedentary, light, moderate, and vigorous activity levels.

Using the related Canadian Health Measures Survey, outdoor activity and duration of sun exposure are measured based on the number of hours and the period of the day. Further, because the study includes children from different ethnic backgrounds with varying skin pigmentation (important for UV absorption and vitamin D synthesis), a compact portable spectrophotometer (CM-600d, Konica Minolta) is used with skin analysis software to measure skin pigmentation [49].

In this study, salivary IGF-1 is examined as a noninvasive biomarker of growth and development in children [50]. Saliva samples are collected using the Oasis Pure•SAL-C Kit between 1 PM and 4 PM to maintain uniform conditions and minimize diurnal variation in IGF-1 levels. The collected saliva samples are analyzed using enzyme-linked immunosorbent assay techniques. Enzyme-linked immunosorbent assay is a highly sensitive and specific method that allows for the quantification of IGF-1 concentrations in biological samples [51].

Cognitive functioning in children is assessed using the Kaufman Assessment Battery for Children, Second Edition (KABC-II), a standardized, individually administered instrument designed to evaluate a broad range of cognitive abilities in children aged 3 to 18 years [52]. The KABC-II is grounded in contemporary theories of intelligence, including the Cattell-Horn-Carroll model and Luria’s neuropsychological framework, making it particularly appropriate for use in diverse populations [53]. It provides composite scores across key cognitive domains, including sequential processing, simultaneous processing, learning, planning, and knowledge, as well as an overall Fluid-Crystallized Index [53]. In this study, the KABC-II is administered to children at 2 time points (months 10 and 30). The assessment typically requires 90 to 150 minutes to complete and is conducted in a quiet, child-friendly environment by trained research personnel. Administration and scoring follow standardized procedures outlined in the KABC-II manual [53].

Two separate KAP questionnaires on dairy foods were developed for caregivers and children following the Food and Agriculture Organization guidelines for assessing nutrition-related KAP [54] and grounded in the health belief model [55]. The caregiver questionnaire is administered once, at month 10, to capture baseline KAP regarding dairy consumption, while the child questionnaire is administered at 2 time points (months 10 and 30) to assess changes over time. Both questionnaires were developed by the research team and reviewed for content relevance, clarity, and age appropriateness. In combination with children’s dietary intake data, these questionnaires capture KAP regarding dairy foods within participating families.

The qualitative component of the study consists of semistructured interviews with a subset of primary caregivers involved in the study. The interviews explore facilitators and barriers to milk and dairy product consumption. Participants are selected from the overall study sample using purposive sampling to ensure representation across ethnic and socioeconomic groups, with interviews continuing until data saturation is achieved [56,57]. Interviews are conducted in private, face-to-face settings convenient for caregivers, and are audio-recorded. Each interview lasts about 30 minutes with informed consent.

All data are managed in accordance with the project’s Data Management Plan and institutional ethics protocols, ensuring confidentiality, data integrity, and controlled access for authorized personnel only. Researchers may request access to nonrestricted anonymized data by submitting a proposal to the principal investigator (HV). As this is a protocol, no data currently exist.

Quantitative data are captured using REDCap (Research Electronic Data Capture; Vanderbilt University), a secure web-based system. Quantitative data from field and laboratory measurements (eg, dietary recalls, anthropometry, sociodemographics, food security, salivary IGF-1, DXA, and HR-pQCT) are entered directly into REDCap. Additional quantitative datasets, such as KABC-II results and accelerometry outputs, are exported from specialized software as digital or CSV files and stored securely in the institutional repository. Dietary recalls data are coded with the ESHA Food Processor software and exported as CSV files. Qualitative data, including audio-recorded interviews and REDCap interview notes, are transcribed, anonymized, and structured, then imported into NVivo (QSR International) for analysis, with the original audio files archived in secure formats.

The study incorporates a comprehensive quality assurance and monitoring strategy to ensure the collection of high-quality and reliable data. A foundational element of this approach is the rigorous training of research assistants (RAs). All RAs participate in training sessions that include an overview of the study protocol, ethical procedures, data collection methods, as well as instruction in using REDCap for data entry and NVivo for qualitative data analysis. The training combines theoretical instruction with hands-on demonstrations to build competency, with refresher sessions held periodically to reinforce protocol adherence and address emerging challenges.

Ongoing supervision and regular data audits are conducted to uphold data integrity throughout the study. Supervisors perform routine checks on field activities to ensure compliance with research protocols and provide immediate, constructive feedback to RAs. Randomly selected data entries are audited for accuracy, completeness, and consistency. These audits include a thorough examination of missing values, outlier handling, and the calculation of derived variables. Any discrepancies identified are promptly corrected and documented, ensuring transparency and continued quality control.

The primary outcome of this study is the association between dairy consumption and bone and body composition outcomes in children aged 5 to 11 years. Bone outcomes are assessed using DXA, including BMC and BMD at the lumbar spine, femur, and total body, as well as lumbar spine bone mineral apparent density. Body composition outcomes are also derived from DXA and include total body muscle mass, fat mass, and percent body fat. These measures are used to describe how dairy intake may be linked to skeletal development and overall body composition during middle childhood.

Secondary outcomes include broader indicators of growth and development. The study evaluates how milk and dairy foods contribute to total dietary intake and overall diet quality. It also examines associations between dairy consumption and cognitive development, as well as hormones and biomarkers connected to growth, including IGF-1. These outcomes help clarify how dairy intake may relate to developmental processes beyond bone health.

The study also investigates KAP related to dairy consumption among children and their caregivers. This includes an assessment of factors that influence dairy intake over time, such as household routines, access to foods, and perceived supports or challenges. Quantitative measures are used to identify facilitators and barriers to dairy consumption among Canadian children. Together, these outcomes provide a detailed picture of how dairy intake fits within patterns of growth, development, dietary behavior, and family context. By addressing these primary and secondary outcomes, the study aims to provide comprehensive insights into the role of dairy consumption in the growth, development, and overall health of Canadian children. The findings will contribute to evidence-based dietary recommendations and interventions aimed at improving child health outcomes.

The research uses a mixed methods approach, using both quantitative and qualitative data. Quantitative data analyses follow the internal data processing protocol, beginning with data cleaning and preparation. This process includes checking for data entry errors, examining distributions, and identifying missing data and outliers. Data entry issues are evaluated by inspecting frequencies, identifying misspelled entries for categorical variables, and examining minimum and maximum values, means, histograms, and ranges for continuous variables. Loss to follow-up will be addressed by comparing baseline characteristics of participants who complete the study with those who withdraw. These comparisons will draw on standardized mean differences and logistic regression to identify differences in demographics, dairy intake, and growth measures [58]. Missing data are addressed using multiple imputation or maximum likelihood methods, depending on the pattern of missingness [59].

Once cleaned, the dataset is used to generate derived variables relevant to the study objectives. These include anthropometric indicators such as BMI, BMI z scores, height-for-age, and weight-for-age z scores; dietary measures such as total dairy intake, calcium intake, and diet quality indices; and consumption groupings based on children’s dairy intake relative to national dietary guidelines. Descriptive statistics summarize participant characteristics and key outcomes. Continuous variables are reported as means with SDs or as medians with IQRs, and categorical variables are summarized as counts and percentages.

Longitudinal associations between dairy intake and outcomes such as somatic growth, body composition, bone health, and cognitive development are evaluated using mixed-effects models and generalized estimating equations. Mixed-effects models account for repeated measures within participants and include random intercepts and slopes where appropriate [60]. Generalized estimating equation estimates population-averaged effects while accounting for within-subject correlation over time [60]. Covariates, including age, sex, socioeconomic status, physical activity, and total energy intake, are incorporated to control for potential confounding. Subgroup analyses will be conducted for male and female participants separately to explore sex-specific associations and developmental trajectories. Interaction effects by sex and baseline diet quality are also examined. Sensitivity analyses assess the robustness of findings, including alternative categorizations of dairy intake and different approaches to handling missing data. Exploratory analyses may be used to examine associations between diet quality indices and secondary outcomes such as bone microarchitecture using regression or correlation models. All quantitative analyses are performed using R (R Foundation for Statistical Computing) or SPSS (IBM Corp) software, in accordance with internal protocols to ensure reproducibility and transparency.

Qualitative data from caregiver interviews and child surveys are analyzed using inductive thematic content analysis. The analysis begins with familiarization through transcript review, followed by iterative coding using NVivo software. Codes are grouped into preliminary themes, which are then reviewed and refined. To ensure consistency and reduce bias, a second coder independently codes approximately 15% of the data using a predefined codebook. Intercoder agreement is quantified using Cohen κ, and any discrepancies are resolved through discussion. Thematic saturation is monitored throughout, and analysis continues until saturation is achieved.

The study uses a comprehensive knowledge translation strategy involving strong collaboration with key partners such as Osteoporosis Canada, public health agencies, community-based child health organizations, and immigrant and cultural health advocacy groups. A dedicated working group will coordinate stakeholder engagement and tailor dissemination strategies for specific audiences, including the public, service providers, policymakers, and academics. Efforts will prioritize cultural sensitivity, particularly for immigrant and Indigenous communities. Dissemination will include policy briefs, workshops, peer-reviewed publications, and conference presentations. Continuous evaluation will guide adaptive improvements to ensure broad impact and sustained use.