Effectiveness of small-quantity versus medium-quantity lipid-based nutrient supplements for prevention of undernutrition in young children: protocol for an individually randomised, parallel-group, non-inferiority trial

Abstract

Introduction

Undernutrition during the first 2 years of life remains a major public health challenge in Pakistan. Lipid-based nutrient supplements (LNS) are widely used to prevent growth faltering in nutritionally at-risk populations. Evidence shows that small-quantity LNS (SQ-LNS, 20 g/day) is effective in preventing undernutrition; however, evidence comparing SQ-LNS with medium-quantity LNS (MQ-LNS, 50 g/day) under programmatic conditions (especially in food-insecure poor households) remains limited. Given that the MQ-LNS daily dose may be too large for older infants to consume fully, comparative effectiveness evidence is needed to inform the national stunting prevention programme. This study compares the effectiveness of SQ-LNS versus MQ-LNS in reducing undernutrition (stunting, wasting and underweight) and anaemia among children aged 6–12 months enrolled in the Benazir Nashonuma Programme (BNP), a government-led nutrition-sensitive social protection programme.

Methods and analysis

This will be an individually randomised, parallel-group, non-inferiority trial conducted within BNP facilitation centres in Dadu and Peshawar districts. Eligible children aged 6–7.5 months will be randomised 1:1 to receive either 20 g/day SQ-LNS or 50 g/day MQ-LNS for 6 months. The primary outcome is stunting assessed using the WHO’s Child Growth Standards. Secondary outcomes are wasting, underweight, anthropometric z-scores, morbidity and anaemia. The required sample size is estimated at 1060 (530 per arm) for Dadu and 508 (254 per arm) for Peshawar. Analyses will follow a non-inferiority framework using a mixed-effects logistic regression model adjusted for relevant covariates. Additionally, we will assess utilisation and acceptability of the intervention among beneficiaries.

Ethics and dissemination

Ethical approval has been obtained from the Aga Khan University Ethical Review Committee (2025–11806-38699) and the National Bioethics Committee of Pakistan (NBCR-1380). We will obtain written informed consent from mothers of eligible children. Findings will be disseminated through engagement with government stakeholders and implementing partners, as well as through peer-reviewed publications and scientific forums.

Trial registration number

NCT07451951.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This is a pragmatic non-inferiority randomised trial from Pakistan comparing a medium-quantity lipid-based nutrient supplement (LNS) with a small-quantity LNS for the prevention of undernutrition in children aged 6–12 months.

• The trial is conducted across two districts with contrasting stunting prevalence, one rural (Dadu, Sindh) and one urban (Peshawar, Khyber Pakhtunkhwa), which enhances the generalisability of findings to Benazir Income Support Programme Kafaalat households.

• As an open-label trial, blinding of caregivers and study team is not feasible given the visible difference in supplement size.

• Self-reported outcomes by caregivers may introduce recall and social desirability bias.

Introduction

Undernutrition in early childhood remains a major public health challenge in Pakistan, particularly among children from food-insecure and socioeconomically disadvantaged communities. Poor nutrition during the first 2 years of life has been associated with impaired physical growth and long-term neurodevelopmental delays. Globally, stunting affects approximately 148.1 million children under the age of 5, with South Asia bearing a disproportionate burden.¹ In Pakistan, nearly 40.2% of children under 5 are stunted, reflecting chronic undernutrition and inadequate maternal and childcare practices.²

Many approaches have been evaluated to prevent and treat malnutrition and to fill nutrient gaps in children during their complementary feeding period from 6 to 23 months of age.³ In particular, lipid-based nutrient supplements (LNS) have been shown to meet essential nutrient requirements and improve growth in children in several studies.^4–7 Small-quantity lipid-based nutrient supplements (SQ-LNS) are designed to provide energy, essential fatty acids and essential micronutrients without replacing breastmilk or complementary local foods. These supplements provide approximately 20 micronutrients in small, palatable daily doses and have shown positive effects on linear growth and anaemia in children.⁸ A strong global evidence base supports SQ-LNS as a preventive intervention to improve child growth and development and reduce mortality. A pooled analysis of SQ-LNS trials (n=14), including data from approximately 37 000 children, demonstrated reductions in child mortality (27%), severe wasting (31%), severe stunting (17%), iron deficiency anaemia (64%) and developmental delay (16–19%).^{9 10} Additionally, cost-effectiveness studies suggest that SQ-LNS is a substantial, cost-effective intervention.^{11 12} SQ-LNS, with a dose of 20 g/day, typically provides 120 kilocalories per day to prevent undernutrition. At the same time, medium-quantity LNS (MQ-LNS, 50 g/day) products deliver more energy and macronutrients. They are primarily intended to prevent moderate acute malnutrition, especially in food-insecure settings and among children with a wider age range, for example, 6–35 or 6–59 months.^13–15

United Nations Children’s Fund (UNICEF) guidelines suggest that the maximum impact of SQ-LNS occurs within 6–11 months.¹⁶ Moreover, a recent review comparing the effects of MQ-LNS to SQ-LNS in children aged 6–23 months posits that MQ-LNS may not offer additional benefits over SQ-LNS.¹⁷ The younger children (6–12 months) who may not fully consume the larger MQ-LNS ration will have a lower additional intake of micronutrients than with full consumption of SQ-LNS, because SQ-LNS and MQ-LNS provide the same amount of micronutrients per daily dose. For older children, who can consume the full dose of MQ-LNS per day, the added benefit of MQ-LNS over SQ-LNS could be the higher amount of energy, protein and fat.

The composition of MQ-LNS is comparable to that of SQ-LNS in terms of the content of vitamins and minerals, essential fatty acids and dairy protein that is provided by a daily dose, as SQ-LNS is more concentrated. A recent study identified Pakistan among the top 20 priority countries with the highest burden of stunting and wasting, and where SQ-LNS will be most impactful for children 6–23 months.¹⁸ Given the strong evidence base for SQ-LNS but questions about its comparative advantage over MQ-LNS, operational research is warranted. Across Pakistan, the Benazir Nashonuma Programme (BNP), a government-led multisectoral nutrition programme under the social safety net programme (Benazir Income Support Programme (BISP)), partners with United Nation organisations.

Beyond nutritional effectiveness, the impact of LNS at scale also depends on several operational factors. Programme uptake may be influenced by caregiver acceptance of a change in supplement, particularly given that beneficiaries under BNP previously received Maamta (75 g/day), a lipid-based nutritional supplement for pregnant and lactating women, before transitioning to MQ-LNS and may now receive SQ-LNS. Fidelity to the intervention is equally critical, as sharing of supplements with other children or family members can substantially reduce the intended dose received by the targeted child.^{19 20} Moreover, the logistical feasibility and acceptability of introducing SQ-LNS among programme implementers at the facilitation centre (FC) level are essential to assess before scale-up, as inadequate counselling and weak supervision have been shown to undermine programme effectiveness.²⁰

In Pakistan, while trials have evaluated various combinations of cash transfers, behaviour change communication and specialised nutritious food supplementation, little local evidence directly compares MQ-LNS to SQ-LNS in real-world programmatic settings.^{14 20} Notably, MQ-LNS in combination with maternal LNS has been locally tested and shown to be effective for stunting prevention in food-insecure households, but comparable evidence for SQ-LNS in this context remains limited. Therefore, operational research is warranted to examine the nutritional effectiveness of introducing SQ-LNS as an alternative for children 6–12 months.

The present study thus aims to compare the effectiveness of MQ-LNS and SQ-LNS under the BNP programmatic circumstances in reducing the prevalence of undernutrition (ie, stunting, wasting and underweight) and anaemia among children 6–12 months of age in the districts of Dadu and Peshawar, Pakistan.

Primary objectives

1. To evaluate the effectiveness of SQ-LNS compared with MQ-LNS in reducing the prevalence of undernutrition (stunting, wasting and underweight) in children aged 6–12 months.

2. To evaluate the effectiveness of SQ-LNS compared with MQ-LNS in reducing the prevalence of anaemia in children aged 6–12 months.

Secondary objectives

1. To assess the acceptability of SQ-LNS and MQ-LNS among caregivers of the enrolled children.

2. To assess the logistical feasibility and acceptability of introducing SQ-LNS in place of MQ-LNS among programme implementers.

Hypothesis

We hypothesise that SQ-LNS will be non-inferior (equivalent) to MQ-LNS in reducing the prevalence of stunting in children aged 6–12 months after 6 months of supplementation. Non-inferiority would be declared if the prevalence of stunting in the SQ-LNS group is not worse than that in the MQ-LNS group, within statistical variability, by a margin of 0.10.

Methods and analysis

Trial design

This will be an individually randomised, parallel-group, non-inferiority trial.

Trial setting and population

The study will be conducted within the existing BNP FCs in the districts of Dadu and Peshawar. FCs serve as the operational hub of BNP and are located within secondary-level public health facilities, including district and tehsil headquarters hospitals.²¹

At FCs, enrolled mothers receive a package of services that includes child growth monitoring, infant and young child feeding counselling and awareness sessions. Conditional cash transfers are then provided to mothers based on compliance with these services. The programme currently operates through more than 500 FCs across 157 districts of Pakistan.²¹

Study staff will be based at the FCs and will work in collaboration with FC personnel. The study team will conduct monthly follow-ups at the household level. The study team will identify all pregnant and breastfeeding women enrolled in BNP from the BISP beneficiary registry and prospectively track them until their infant reaches 6 months of age. From 6 to 7.5 months, mothers will be approached at the FC and infants meeting the inclusion criteria will be enrolled after written informed consent is obtained. The infant will then be individually randomised to receive either SQ-LNS or MQ-LNS (participant flow is shown in figure 1).

Figure 1. Flow of participants through the trial.

Study duration

The total study duration will be 12 months, with a planned recruitment period of 6 months. Each enrolled child will be followed for a minimum of 6 months, which is considered an adequate exposure duration to observe meaningful changes in nutritional and anthropometric outcomes based on existing literature and previous implementation experience.

Trial status

Recruitment started on 26 February 2026, and is expected to be completed by July 2027.

Eligibility criteria

Inclusion criteria

1. Children aged 6–7.5 months with a weight-for-length z-score (WLZ) ≥−2 SD and mid-upper arm circumference (MUAC) ≥12.5 cm, born to BISP-recipient mothers registered with BNP at any point during their pregnancy.

2. Permanent residents of the selected districts, with no plan to relocate or migrate within the next 6 months.

3. The child’s caregivers provide written informed consent.

Exclusion criteria

1. Children with moderate acute malnutrition, defined as WLZ ≥−3 to <−2 SD or MUAC ≥11.5 cm to <12.5 cm.

2. Children with severe acute malnutrition (SAM), defined as WLZ <−3 SD, MUAC 11.5 cm or bilateral pitting oedema, require therapeutic feeding.

3. Children with known congenital anomalies, chronic illness or conditions affecting growth or feeding (eg, cerebral palsy, congenital heart disease).

4. Children already enrolled in any nutritional supplementation other than BNP or another clinical trial, and whose families will be unwilling to provide informed consent, or are likely to be unavailable for follow-up.

Intervention

Participants will receive either 20 g/day SQ-LNS or 50 g/day MQ-LNS sachets, to be consumed directly or mixed with local complementary foods, according to their random assignment. Caregivers will administer supplements daily until the child is 12 months old. However, all caregivers of the participants of both groups will receive counselling on the appropriate use of the assigned LNS product, the importance of continued breastfeeding and age-appropriate complementary feeding practices to ensure dietary diversity. The supplement sachets will be pre-packaged into boxes (MQ-LNS) or bags (SQ-LNS) of 90 each, distributed quarterly at the FC by the FC staff.

Outcomes

Primary outcome

Stunting assessed at 12 months of age, defined as length-for-age z-score (LAZ) <−2 SD

Secondary outcomes

1. Wasting is defined as WLZ <−2 SD.

2. Underweight defined as weight-for-age z-score (WAZ) <−2 SD.

3. Child LAZ, WLZ and WAZ.

4. Morbidity will include diarrhoea (liquid stools defined as ≥3 loose or watery stools within 24 hours), respiratory infection (cough plus rapid, difficult breathing), vomiting, fever, rash/sores, coughing, wheezing and pneumonia.

5. Prevalence of anaemia is defined as haemoglobin <10.5 g/L.²²

   1. Mild: Haemoglobin 95–104 g/L.

   2. Moderate: Haemoglobin 70–94 g/L.

   3. Severe: Haemoglobin <70 g/L.

Measurement of outcomes

All anthropometric outcomes will be measured using Seca anthropometry kits and calculated using the WHO Anthro software. SECA 874 portable electronic weighing scale and SECA 416 for length will be used. MUAC will be measured using a tricoloured MUAC tape to the nearest 0.1 cm. Further, to ensure accuracy, the scales will be calibrated every morning before field activities to minimise calibration error. All measurements will be taken two times. A third MUAC reading will be taken if the difference between the first two readings is greater than 0.5 cm. Moreover, if the discrepancy between two length measurements exceeds 0.5 cm, a third measurement will be taken. Similarly, weight will be taken two times; if the discrepancy exceeds 100 g in children, a third measurement will be recorded.

Weight will be measured using the mother–child tare function of the weighing scale. The caregiver’s weight will be recorded first, after which the tare function will be activated. The caregiver will then hold the child while standing on the scale, and the scale will automatically display the child’s weight. Emphasis will be placed on hygiene and on respondents’ respect throughout the measurement process. The weighing scale and length board will be cleaned before and after each participant’s measurements to foster a trusting environment between enumerators and participants.

Additionally, caregiver-reported symptoms will be collected using a structured questionnaire administered during monthly home visits to track and register morbidity. For anaemia, haemoglobin levels will be measured using a capillary blood sample with a HemoCue Hb 301 point-of-care device. Standardised measurement protocols will be applied consistently across all study groups to ensure comparability.

Data collection

Data will be collected at the facility level during the recruitment phase and at home visits during the follow-up period by trained data collectors using a structured questionnaire. The questionnaire will be translated from English into Urdu and Sindhi, and then back-translated into English to ensure validity. A designated field supervisor, trained specifically for this role, will supervise all data collection activities. The supervisor will conduct a daily morning briefing to review tasks, discuss issues and ensure consistency in data collection practices.

At the end of each day, data collection teams will return to the field office, where completed forms will be reviewed for completeness and accuracy. Teams will check the entire filled questionnaire for completeness before handing it over to the field supervisor. The field supervisor will then review the questionnaire for errors and inconsistencies. In case any discrepancies are identified, the forms will be returned to the respective data collectors for correction. Once verified, digital data entry will be conducted for all completed forms using tablets via a customised Java application with a SQLite backend for data storage. At the end of each day, the collected data will be transferred immediately from the devices to the database server via the internet.

Data will be collected every month, from 6 months to 12 months of children’s age:

For baseline data, data will be captured using structured questionnaires and direct assessments on socio-demographic information including parental education and occupation, mother’s height, weight and MUAC; household size and food security status using the Household Food Insecurity Access Scale (HFIAS) by the Food and Agriculture Organisation²³; handwashing practices; health and nutrition history and Infant and Young Child Feeding (IYCF) practices; child vaccination status; anthropometric measurements including weight, length and MUAC; and screening for anaemia (table 1).

Table 1. Schedule of enrolment, interventions and assessments.

	Trial period
	Enrolment	Post-randomisation
Time point (±3 days)	Month 0	Month 1	Month 2	Month 3	Month 4	Month 5	Month 6
Enrolment
Eligibility screening	✓
Informed consent	✓
Randomisation	✓
Intervention/comparator
MQ-LNS (50 g/day) – one sachet per day*		✓					✓
SQ-LNS (20 g/day) – one sachet per day*		✓					✓
Assessments
Haemoglobin assessment	✓						✓
Socio-demographic (parental education and occupation, mother's height, weight and MUAC); household data collection (household size and food security status using HFIAS by FAO; handwashing practices)	✓
Child vaccination status	✓				✓
IYCF practices	✓			✓			✓
Anthropometry – weight, length, MUAC	✓	✓	✓	✓	✓	✓	✓
Caregiver-reported morbidity data	✓	✓	✓	✓	✓	✓	✓
Compliance assessment (used sachets count, caregiver-reported sachet consumption, sharing and partial use)			✓	✓	✓	✓	✓

^*Arrow indicates continuous delivery of intervention.

FAO, Food and Agriculture Organization ; HFIAS, Household Food Insecurity Access Scale; IYCF, Infant and Young Child Feeding; MQ-LNS, medium-quality lipid-based nutrient supplements; MUAC, mid-upper arm circumference ; SQ-LNS, small-quantity lipid-based nutrient supplements.

For subsequent follow-up data, monthly (±3 days) data will be collected for anthropometric measurements, including weight, length and MUAC; compliance with the assigned intervention (SQ-LNS or MQ-LNS) assessed using caregiver-reported consumption by the child, sharing practices (partial or entire sachets shared) and counting of available sachets (used and unused); caregiver’s acceptability of LNS at third and sixth follow-up visit; morbidity; immunisation status at 10 months. IYCF practices will be recorded at 9 and 12 months of age. Screening for anaemia will be conducted at enrolment and at 12 months of age. Moreover, any child identified with SAM during follow-up will be referred to the BNP staff at the facility for treatment with ready-to-use therapeutic food, as per their protocol and followed-up by the team until normal nutritional status is achieved. Acceptability of introducing SQ-LNS in place of MQ-LNS among programme implementers will also be assessed.

Training of study teams

Before data collection commences, the data collection teams will receive comprehensive 5-day hands-on training from the study investigators. The central training will cover study objectives, methods, data collection techniques, anthropometric measurement techniques and ethical issues. This will be followed by a 1-day field-testing session to pilot the procedures and ensure team readiness for actual fieldwork. An anthropometry standardisation training will also be conducted at the FC. During this session, each data collector will perform anthropometric measurements on five children after obtaining parental consent. The instructor will perform anthropometry on the same children. The measurements taken by the team member and the instructor will be compared with evaluate interobserver reliability. Team members will be retrained if measurements differ. Each team will receive a survey manual with instructions, methodological details, the outcome assessment procedure and troubleshooting tips. The printed questionnaires used by data collectors will be in Sindhi and Urdu for ease of administration.

Assessment of compliance

The study team will assess compliance during monthly monitoring and quarterly procedures aligned with the programme. At monthly follow-up visits, study staff will count both used (empty) and unused sachets retained by caregivers. To facilitate this, caregivers will be provided with labelled zip-lock bags at enrolment for all participating mothers, and will be requested to store the empty sachets in these bags. In addition to sachet counts, caregivers will be asked about the number of sachets consumed by the child, the number of sachets shared and the extent of partial consumption or sharing. These self-reported measures will complement sachet counts to provide a more accurate assessment of adherence. Non-adherence will be documented, and reasons will be explored. Compliance will be defined as the percentage of sachets consumed, calculated as the number of sachets consumed divided by the total number of sachets provided during the consumption period, multiplied by 100. It will be further classified based on its distribution within the study population using percentile cut-offs, with low compliance defined as values below the 25th percentile, moderate compliance as values between the 25th and 75th percentiles and high compliance as values above the 75th percentile.

In line with the programmatic requirements, caregivers are expected to return at least 90% of empty sachets during quarterly visits at FCs. For study participants, compliance will be assessed at these quarterly visits, as per routine programme procedures. To mimic programmatic circumstances and minimise deviations from intended interventions, caregivers will receive clear instructions from FC staff on proper use, dosage and storage of the assigned supplement during the initial enrolment session. These instructions will be reinforced during the quarterly visit to the FC. In particular, this will include the following messaging:

1. SQ-LNS and MQ-LNS are powerful supplements.

2. While SQ-LNS is smaller than MQ-LNS, it is expected to provide the same benefit to the child and this will be tested in the study.

3. Because of the testing in the study, it is very important not to mix these supplements (ie, do not ‘trade’ with friends) and to consume the allocated supplement (MQ-LNS or SQ-LNS) in full (no wide sharing).

In addition to individual-level compliance, delivery system performance will be evaluated monthly. This will include assessments at the FCs of stock availability, adequacy of the food storage system and record maintenance. These assessments will be conducted using a Direct Observation Checklist and a review of information systems and records.

Harms

The study is considered low risk, as both MQ-LNS and SQ-LNS are approved nutrient supplements, with MQ-LNS already in use in BNP. Potential risks to participants may include minor pain or discomfort during finger-prick haemoglobin testing. All anthropometric and haemoglobin assessments will be performed by trained personnel using standardised SECA equipment and a HemoCue device to ensure participant safety.

Sample size

The sample size calculation is based on the prevalence of stunting among children aged 6–12 months from the Multiple Indicator Cluster Survey.^{24 25} Assuming a prevalence of 60.6% for Dadu and 13.2% for Peshawar, no expected difference between the study groups, 90% power, an alpha of 0.05, an anticipated 5% attrition rate and a non-inferiority margin of 10%, the required sample size is estimated at 1060 (530 per arm) for Dadu and 508 (254 per arm) for Peshawar.

There is no agreed standard for non-inferiority margins in LNS trials.²⁶ Moreover, no prior trials have directly compared SQ-LNS and MQ-LNS for stunting outcomes in Pakistan, and MQ-LNS has not demonstrated a statistically significant effect on stunting prevalence compared with no LNS provision in recent meta-analyses.¹⁷ This makes empirical derivation of the margin from standard treatment-effect data infeasible. In the absence of such data, a 10%-point margin was defined based on clinical and programmatic judgement, consistent with established guidance.²⁷

Sequence generation

The random allocation sequence will be generated using a computer-based random number generator with a 1:1 allocation ratio between the SQ-LNS and MQ-LNS groups. Randomisation will use varying blocks for each FC to ensure balance between intervention arms at both sites.

Allocation concealment mechanism

A statistician not involved in participant enrolment or outcome assessment will prepare the random allocation sequence. Allocation will be concealed using a centralised computer-based assignment system, that is, the allocation sequence will not be accessible to field staff involved in screening or enrolment.

Implementation

Trained study staff will enrol eligible participants at the FCs. Assignment to the intervention arms will follow the pre-generated randomisation sequence. Personnel responsible for enrolment and those assigning participants to the intervention groups will not have access to the random allocation sequence.

Blinding

This will be an open-label trial. Due to the nature of the intervention and differences in the quantity of LNS provided, blinding of caregivers and the data collection team is not feasible. However, standardised protocols for outcome assessment and data collection will be used to minimise potential bias. The statistical analysis team will be blinded to the study group allocation.

Statistical analysis plan

Data will be analysed using Stata V.18 (StataCorp). Anthropometric indices, WAZ, LAZ and WLZ will be calculated based on the WHO Child Growth Standards.²⁸ The primary outcome will be the difference in stunting prevalence among children. Secondary outcomes will include differences in the prevalence of wasting, underweight, LAZ, WLZ and WAZ, morbidity and changes in anaemia status. We will perform intention-to-treat and per-protocol analyses.

Descriptive statistics will include frequencies and percentages for categorical variables. We will assess continuous variables for normality and present results as mean±SD or median with IQR, as appropriate.

For univariate analyses, χ² tests will be used for categorical variables, independent t-tests for normally distributed continuous variables and Mann-Whitney U tests for non-normally distributed continuous variables.

To assess the effect of SQ-LNS compared with MQ-LNS on stunting prevalence, we will fit a mixed-effects logistic regression model. The model will include fixed effects for intervention group, time and the intervention-by-time interaction term. A participant-level random intercept will be included to account for the correlation of repeated measurements within the same child over time. The model will be adjusted for prespecified covariates, including child sex, maternal age, education, maternal body mass index, dietary diversity, HFIAS score and IYCF practices. We will also examine compliance with the assigned LNS as an effect modifier. The intervention effect will be estimated using the interaction term, with statistical significance assessed at a two-sided alpha level of 0.05.

Missing data will be carefully assessed for patterns and mechanisms (eg, missing completely at random, missing at random or not at random). Depending on the nature and extent of missingness, appropriate statistical techniques, such as multiple imputation, weighting adjustments or sensitivity analyses, will be applied to minimise bias and maintain the validity of the results. All decisions and methods for handling missing data will be fully documented and reported in the analysis.

Monitoring

Given the low-risk nature of the intervention, no independent data and safety monitoring committee will be established. The principal investigators and the study coordination team will conduct trial oversight and monitoring weekly. No interim efficacy analyses are planned.

Patient and public involvement

None.

Ethics and dissemination

The study will strictly adhere to ethical research principles. Ethical approval from the Ethical Review Committee of Aga Khan University (2025–11806-38699) and the National Bioethics Committee (NBC), Government of Pakistan (NBCR-1380) has been obtained. The trial has been registered as an individual randomised trial in ClinicalTrials.gov (NCT07451951). All aspects of Good Clinical Practice will be strictly followed.

Written consent will be obtained from caregivers of study participants before enrolment, and participation will be voluntary (online supplemental material). They will be clearly informed of the study’s objective and the requirement to participate in the trial for 12 months. Participants will be free to leave the study at any time. Caregivers will be informed that participation is completely voluntary and that they may withdraw at any time without penalty. Refusal or withdrawal from participation will not affect existing BNP or BISP benefits or any other health services. We will engage well-trained field staff members, proficient in the local language, who will comprehensively explain the written consent text in their native language to ensure clarity and understanding before initiating the study procedure.

The data collection application will be password-protected. Only users with the provided credentials can use the application. Once a record is saved, data collection staff will no longer be able to edit it. Data stored on handheld devices and during transfer will be encrypted to prevent breaches of participants’ personal information. All collected data will be anonymised by removing personal identifiers during analysis, report writing and dissemination stages.

The findings from this study will be disseminated through engagement with national and provincial stakeholders, such as government partners and implementing agencies, to support evidence-informed decision-making. Results will also be shared through peer-reviewed publications and scientific forums, contributing to the broader global literature on preventive nutrition interventions in low-resource settings.

Discussion

This trial will generate context-specific evidence on the comparative effectiveness of MQ-LNS and SQ-LNS in preventing undernutrition among infants aged 6–12 months in the poorest BISP Kafaalat households of two districts of Pakistan. Both Dadu and Peshawar experience substantial levels of stunting, with prevalence of 60.6% and 13.2%, respectively.^{24 25} This demonstrates regional disparities in food security. This contrast is valuable because it will allow the findings to be interpreted across diverse settings. If results differ across sites, this will provide important insight into whether the comparative effectiveness of MQ-LNS versus SQ-LNS is context-dependent so that future programmatic recommendations can be tailored to Pakistan’s varying food-insecure regions.

Pakistan’s national nutrition-sensitive social protection and nutrition programmes face the ongoing challenge of delivering effective yet cost-efficient interventions to prevent undernutrition in early childhood. Comparative effectiveness research provides a pragmatic approach to generate evidence that can inform programmatic decisions and ensure optimal use of resources. This trial is designed to test the hypothesis that SQ-LNS is non-inferior to MQ-LNS in reducing the prevalence of undernutrition and anaemia when distributed by BNP to children aged 6–12 months, where this age group has been eligible to receive MQ-LNS for the past 3–6 years (depending on when the programme was introduced in the specific district). This will have significant programmatic implications for BNP.

Integrating the trial into the existing national nutrition-sensitive social protection programme will enhance its external validity, as the interventions will be delivered in real-world programmatic conditions. If SQ-LNS is shown to be non-inferior to MQ-LNS, the findings will be critical in considering a programmatic switch. This will help determine whether the evidence supports a shift to SQ-LNS across sites for younger children aged 6–12 months, as per the study’s results. Additionally, this trial addresses a programmatic transition that has not been previously studied; beneficiaries under BNP have progressed from receiving Maamta (75 g/day) to MQ-LNS and may transition to SQ-LNS. Since prior supplementation experience has been shown to influence caregiver utilisation behaviours,^{19 20} understanding caregiver acceptance of the new supplement will be helpful. Evidence generated from this trial will therefore be well-positioned to inform potential scale-up.

The trial has a few limitations. The open-label design, while unavoidable given the visibly different size and packaging of SQ-LNS and MQ-LNS sachets, introduces the possibility of differential caregiver behaviour between arms. We have sought to mitigate this through standardised counselling at enrolment, monthly sachet counts and caregiver-reported sharing data. Compliance ascertainment itself relies on sachet counts and caregiver recall, both of which are susceptible to social desirability bias. Moreover, the trial population is restricted to BNP beneficiary households in Sindh and Khyber Pakhtunkhwa, which limits the generalisability of findings to non-programme populations and to settings with different dietary contexts or health system structures. Finally, haemoglobin will be measured from single-drop capillary blood using a HemoCue Hb 301 device, selected because extreme heat and humidity at the field centres compromise the chemical stability of other HemoCue devices’ reagent-based microcuvettes. However, recent evidence suggests this approach may affect absolute anaemia prevalence estimates compared with venous blood analysed on an autoanalyser.²⁹ Between-arm comparisons within this trial are expected to remain valid, as the same procedure will be applied uniformly across both groups. The devices will undergo external quality control using the manufacturer’s liquid control solution, in addition to the device’s built-in self-test performed before each measurement session.