Development and Validation of Breastfeeding Knowledge Test for Women with Gestational Diabetes Mellitus

Abstract

Background/Objectives: Gestational diabetes mellitus (GDM) affects approximately 12.7% of pregnant women in South Korea. While breastfeeding provides critical health benefits for mothers with GDM and their infants, including improved insulin resistance and reduced Type 2 diabetes risk, no validated GDM-specific breastfeeding knowledge instrument exists. This study aimed to develop and validate a breastfeeding knowledge instrument for women with GDM. Methods: This methodological study employed systematic procedures for the development and validation of knowledge test. Initial item generation yielded 30 items across three domains: postpartum physical characteristics, breastfeeding barriers, and breastfeeding benefits. Content validity was evaluated by six clinical experts and ten experiential experts (women with GDM). An online survey was conducted in October 2022 with 220 women diagnosed with GDM who were either pregnant or within six months postpartum. Item analysis, exploratory factor analysis, and reliability testing were performed. Convergent validity was assessed by calculating the Pearson correlation coefficient with an established breastfeeding knowledge scale. Results: Following expert review and psychometric analysis, the final instrument comprised 14 items across three factors: postpartum physical characteristics (3 items), breastfeeding barriers (2 items), and breastfeeding benefits (9 items). The Kaiser–Meyer–Olkin measure was 0.884, Bartlett’s test was significant (χ² = 838.835, p < 0.001), and factor loadings were satisfactory. The KR-20 reliability coefficient was 0.826, and criterion validity was confirmed. Conclusions: This first validated GDM-specific breastfeeding knowledge instrument enables the identification of knowledge gaps and the development of targeted educational interventions to improve maternal-child health outcomes.

1. Introduction

Gestational diabetes mellitus (GDM) is a prevalent metabolic disorder complicating pregnancy, with a global prevalence of approximately 14.0% as reported by the International Association of Diabetes in Pregnancy Study Group [1]. In South Korea, the prevalence has reached 12.7%, with an annual increase of 1–2% [2]. GDM poses considerable short- and long-term health risks for both mothers and their offspring. In the short term, it is associated with gestational hypertension, preeclampsia, fetal macrosomia, cesarean delivery, and neonatal intensive care unit admission [3]. More critically, women with a history of GDM have an 8.3-fold higher risk of developing type 2 diabetes mellitus (T2DM) after childbirth compared with women without GDM [4]. Furthermore, children born to mothers with GDM face elevated risks of T2DM, obesity, and neurodevelopmental impairment [5].

Breastfeeding is an evidence-based, well-established intervention known to mitigate the long-term metabolic consequences of GDM. Among women with GDM, breastfeeding improves glucose tolerance, enhances insulin sensitivity, and significantly reduces the subsequent incidence of T2DM [6]. For offspring, it confers protection against obesity and T2DM development [6]. Despite these well-documented benefits, women with prior GDM tend to exhibit lower exclusive breastfeeding rates and shorter breastfeeding duration than the general population [7]. This disparity arises from multiple barriers, including physiological complications (e.g., maternal obesity, neonatal hypoglycemia), insufficient professional support, and diminished maternal self-efficacy [6]. Moreover, persistent misconceptions—such as the belief that maternal milk may elevate infant blood glucose or lack sufficient nutritional quality—further discourage breastfeeding practices among women with GDM [8].

A notable knowledge gap exists regarding the specific benefits of breastfeeding in preventing long-term GDM-related complications. Many women with GDM remain unaware that breastfeeding can substantially reduce the risk of future metabolic disorders [8]. This lack of awareness is compounded by a generally low perception of long-term health risk within this population [9]. Prior studies have demonstrated that a higher perceived risk of T2DM is associated with greater engagement in preventive health behaviors [10]. Consequently, improving targeted knowledge about the health benefits of breastfeeding in the context of GDM may enhance both breastfeeding intentions and actual practices.

Accurately assessing and improving breastfeeding knowledge necessitates the use of validated, disease-specific measurement tools. However, to date, no validated breastfeeding knowledge instrument specifically tailored to women with GDM has been published. Existing assessments were developed for the general population [11] and are inadequate for GDM populations for several reasons. First, general instruments do not address GDM-specific physiological challenges such as delayed lactogenesis II, which affects approximately one-third of women with GDM due to insulin dysregulation impairing mammary gland development [12,13]. Second, these tools fail to capture unique barriers faced by women with GDM, including neonatal hypoglycemia management requiring formula supplementation, higher cesarean delivery rates, and misconceptions that breast milk may adversely affect infant blood glucose [6,14]. Third, general instruments do not assess knowledge about the disease-specific preventive benefits of breastfeeding, such as reducing maternal risk of type 2 diabetes and improving long-term metabolic outcomes for offspring [15,16]. Likewise, instruments such as the Breastfeeding Self-Efficacy Scale [17] and the Iowa Infant Feeding Attitude Scale [18] provide valuable insights into maternal confidence and attitudes but do not measure factual, GDM-specific knowledge. Recent evidence demonstrates that patients with chronic conditions present multifaceted levels of care complexity, reinforcing the importance of developing validated, condition-specific assessment tools rather than relying on general instruments [19]. Without validated GDM-specific tools, healthcare professionals cannot effectively identify knowledge deficits unique to this population, and researchers are unable to accurately evaluate the impact of targeted educational interventions [11].

Therefore, this study aimed to develop and validate a breastfeeding knowledge instrument specifically designed for women with GDM, addressing the three critical knowledge domains: postpartum physical characteristics relevant to breastfeeding, GDM-specific breastfeeding barriers, and disease-related preventive benefits of breastfeeding.

2. Materials and Methods

2.1. Study Design

This methodological study followed systematic procedures for instrument development and psychometric validation as outlined by DeVellis and Thorpe [20], incorporating expert content evaluation and statistical testing (Appendix A).

2.2. Instrument Development Process

2.2.1. Initial Item Generation

Through methodological triangulation [21], three essential knowledge domains were identified by converging evidence from multiple sources. The literature review identified GDM-specific physiological challenges (e.g., delayed lactogenesis II, altered insulin requirements), unique barriers (e.g., neonatal hypoglycemia management, misconceptions about breast milk), and disease-specific metabolic benefits (e.g., reduced T2DM risk) [6,12,13,14,15,16,19]. Analysis of FGI transcripts [8] corroborated these findings, revealing that women with GDM expressed concerns about postpartum physical changes, reported obstacles including fear of hypoglycemia, and demonstrated limited awareness of long-term breastfeeding benefits. Clinical expert consultation during content validation confirmed that these three domains comprehensively captured essential GDM-specific knowledge. This convergence of empirical literature, qualitative findings, and clinical expertise provided triangulated evidence supporting the three-domain framework: postpartum physical characteristics, breastfeeding barriers, and breastfeeding benefits (

Table 1. Major Domain and Sub-domain for the Breastfeeding Knowledge Test.

Major Domain	Sub-Domain	Example Topics
Postpartum physical characteristics	Physiological aspects	Postpartum metabolic regulation, insulin demand, glucose tolerance
	Lactogenesis	Initiation and maintenance of milk production
	Lactogenesis	Hormonal regulation of milk secretion
Breastfeeding Barriers	Pathological	Maternal obesity, neonatal hypoglycemia, delayed lactation
	Lifestyle	Diet, exercise, and sleep affecting production of milk.
	Perception	Misconceptions, self-efficacy, perceived risk of diabetes
	Infant biological	Infant feeding tolerance, neonatal glucose response
Breastfeeding Benefits	Metabolic effects	Improved insulin sensitivity, glucose control
	Long-term effects (>1 year postpartum)	Prevention of T2DM, cardiovascular benefits
	High-intensity breastfeeding	Dose–response effects of extended lactation
	Prevention of recurrence	Reduced recurrence of GDM
	Prevention of complication	Lower risk of obesity and metabolic syndrome
	Offspring health	Reduced childhood obesity, lowered diabetes risk.

).

This identified three essential knowledge domains: postpartum physical characteristics, breastfeeding barriers, and breastfeeding benefits (Table 1). A multidisciplinary team (two maternal-child health nursing professors, one postdoctoral researcher in instrument development, and two doctoral students) systematically generated 30 preliminary items addressing all three domains. The initial distribution comprised eight items on postpartum physical characteristics, eight on breastfeeding barriers, and fourteen on breastfeeding benefits, reflecting the scope and complexity of each domain.

2.2.2. Content Validity

Content validity was established through three steps:

Clinical Expert Panel: Six clinical experts (all female; mean age 45.3 ± 6.8 years; two lactation specialists, two maternal-child health nursing professors with breastfeeding research experience, and two instrument development specialists) evaluated item appropriateness and relevance. Their feedback refined the domains to postpartum physical characteristics, breastfeeding barriers, and breastfeeding benefits, reducing the instrument to 24 items (6, 6, and 12 items, respectively).

Experiential Expert Panel: Ten women with GDM (five pregnant and five postpartum within six months) reviewed items for clarity and bias. Their feedback further refined the instrument to 19 items (5, 5, and 9 items, respectively).

Linguistic Review: A Korean language specialist reviewed the revised instrument for linguistic accuracy, readability, and cultural appropriateness, finalizing the preliminary 19-item version.

2.3. Participants and Sample Size

Eligible participants were women aged ≥18 years clinically diagnosed with GDM who were either pregnant or within six months postpartum. Exclusion criteria included hospitalization in tertiary obstetric/endocrinology wards and impaired cognitive capacity. Based on factor analysis guidelines, a minimum sample of 200 was required [22]; accounting for 20% attrition, 240 participants were targeted.

2.4. Data Collection

Data were collected in October 2022 using a secure web-based platform (Moaform). Participants reviewed an information sheet, confirmed eligibility, and provided electronic informed consent before accessing the questionnaire. During the online survey, the platform automatically recorded completion time for each participant. The system prevented duplicate submissions.

2.5. Instruments

Demographic and clinical characteristics, including age, education, employment, and obstetric/breastfeeding history, were collected.

General Breastfeeding Knowledge Scale (criterion measure) was the Korean version adapted by Ra and Chae’s [23] from Ahmed et al.’s [24], who revised Brodribb et al.’s [25] original 36-item instrument by excluding 12 clinical decision-making items, yielding 24 items. Scoring was identical (0–24), with reported reliability Kuder–Richardson Formula 20 (KR-20) = 0.75.

2.6. Statistical Analysis

Data were analyzed using IBM SPSS Statistics (version 26.0). Participant demographics were summarized using frequencies and percentages for categorical variables, and means with standard deviations for continuous variables.

For item analysis, item means, standard deviations, and item-total correlations were examined. Normality was assessed using skewness and kurtosis criteria [26]. Items with corrected item-total correlations < 0.20 were flagged for removal [27], except conceptually essential misconception items were retained for content validity.

Exploratory factor analysis employed principal components analysis with direct oblimin rotation. The Kaiser–Meyer–Olkin measure and Bartlett’s test of sphericity assessed factorability. Communality analysis identified items with h² > 0.30 for review [28]. Factor retention used the eigenvalue criterion (>1.0) with consideration for interpretability and theoretical meaningfulness [29].

Convergent validity was assessed using Pearson correlation with the established breastfeeding knowledge scale for general women. Internal consistency was evaluated using KR-20. All tests used α = 0.05 (two-tailed).

2.7. Ethical Considerations

This study was approved by the Institutional Review Board of Chungbuk National University (Approval No. CBNU-202210-HR-0229). The online survey provided information about the study’s purpose, confidentiality, and withdrawal rights. Only participants who confirmed eligibility and consented electronically accessed the questionnaire. Upon completion, participants received a small incentive.

3. Results

3.1. Participant Characteristics

Through convenience sampling via GDM-related blogs and breastfeeding clinic websites, 224 women accessed the survey. After excluding incomplete data and careless responding patterns, 220 participants comprised the final sample. Participants were predominantly in their thirties (80.5%, mean age 32.90 ± 3.56), with 54.1% pregnant and 45.9% postpartum. Most had a college education (85.5%), perceived themselves as middle-class (75.5%), and were unemployed (48.6%) or employed full-time (40.5%). The majority (82.3%) were nulliparous, and most managed blood glucose through diet and exercise (76.4%), with 13.2% using medication and 10.4% using no management (

Table 2. Characteristics of participants (n = 220).

Characteristics	Categories	n (%) or M ± SD
Pregnancy status	Pregnant	119 (54.1)
	Delivered within six months	101 (45.9)
Age (years)		32.90 ± 3.56
	20≤ and <30	32 (14.5)
	30≤ and <40	177 (80.5)
	40≤ and <50	11 (5.0)
Education	≤High school	15 (6.8)
	College	188 (85.5)
	≥Graduate school	17 (7.7)
Perceived economic status	Low	32 (14.5)
	Middle	166 (75.5)
	High	22 (10.0)
Employment	Full-time	89 (40.5)
	Part-time	24 (10.9)
	Unemployed	107 (48.6)
Parity	None	181 (82.3)
	Yes	39 (17.7)
Blood sugar control methods	Medication	29 (13.2)
	Non-medication (diet and/or exercise)	168 (76.4)
	No management	23 (10.4)

Note: M = Mean; SD = Standard Deviation.

).

3.2. Content Validity Evaluation

Clinical Expert Evaluation: Six clinical experts evaluated 30 preliminary items using a four-point Likert scale. The I-CVI values reported represent the average of appropriateness and relevance ratings for each item, ranging from 0.67 to 1.00, with 75% of items achieving an I-CVI ≥ 0.83. Four items were deleted and 16 revised based on expert feedback. The S-CVI/Ave, calculated as the mean of all I-CVI values, was 0.94. The instrument was reduced to 24 items (6 postpartum physical characteristics, 6 breastfeeding barriers, 12 breastfeeding benefits).

Experiential Expert Evaluation: Ten women with GDM (five pregnant, five postpartum) reviewed the 24-item instrument. All agreed items were clear and understandable; 90% rated the difficulty as moderate. Five items were refined, and two were deleted due to redundancy, resulting in a 19-item instrument (5 postpartum physical characteristics, 5 breastfeeding barriers, 9 breastfeeding benefits).

3.3. Item Analysis

All 19 items demonstrated acceptable statistical properties. Item difficulty, represented by item means, ranged from 0.20 to 0.80, indicating appropriate difficulty distribution without floor or ceiling effects. Items with means closer to 0.50 represent moderate difficulty, while those approaching 0 or 1 indicate very difficult or very easy items, respectively [26]. Correct answer rates for the final 14 items ranged from 21.4% (Q3) to 75.0% (Q10), with most items showing moderate difficulty levels (39.5–71.4%). All items met normality criteria (skewness ±2.0, kurtosis ±7.0) [27]. Three items (Q6, Q14, Q16) with item-total correlations < 0.20 were flagged for removal [26].

3.4. Exploratory Factor Analysis

Initial EFA on 16 items: Kaiser–Meyer–Olkin = 0.862, Bartlett’s test χ² = 883.854, df = 120, p < 0.001. Two items with communalities <0.30 were removed. Subsequent EFA on the refined 14-item set: KMO = 0.884; Bartlett’s test χ² = 838.835, df = 91, p < 0.001. Three factors with eigenvalues >1.0 (4.81, 1.23, 1.18) explained 51.63% of variance (34.38%, 8.80%, 8.45%, respectively) (

Table 3. Summary of Exploratory Factor Analysis Results for the GDM-Specific Breastfeeding Knowledge Test (n = 220).

Item	Factor 1: Breastfeeding Benefits	Factor 2: Postpartum Physical Changes	Factor 3: Breastfeeding Barriers	Communality	Correct Answer Rate (%)
Q2	0.742			0.588	65.0
Q5	0.681			0.512	71.4
Q7	0.625			0.465	50.9
Q9	0.718			0.601	52.7
Q11	0.659			0.512	65.0
Q13	0.745			0.638	66.8
Q15	0.702			0.548	57.3
Q17	0.689			0.512	61.8
Q19	0.715			0.618	56.4
Q1		−0.814		0.512	61.4
Q8		−0.756		0.448	54.5
Q12		−0.698		0.588	39.5
Q3			0.825	0.612	21.4
Q10			0.798	0.625	75.0
Eigenvalue	4.814	1.232	1.183
Variance explained (%)	34.38	8.80	8.45
Cumulative variance (%)	34.38	43.18	51.63

Note: GDM = Gestational Diabetes Mellitus; Negative loadings observed for Factor 2 reflect the arbitrary direction of the rotated factor solution and do not affect the interpretation of the factor structure.

).

Factor 1 ‘Breastfeeding Benefits’ comprised 9 items (Q2, Q5, Q7, Q9, Q11, Q13, Q15, Q17, Q19). Factor 2 ‘Postpartum physical characteristics’ comprised 3 items (Q1, Q8, Q12). Factor 3 ‘Breastfeeding Barriers’ comprised 2 items (Q3, Q10).

3.5. Reliability

Internal consistency by Kuder–Richardson Formula 20: initial 19-item instrument KR-20 = 0.785; 16-item version KR-20 = 0.804; final 14-item version KR-20 = 0.826, indicating good internal consistency. Mean scores per item ranged from 0.69 to 0.73 across subscales, suggesting relatively consistent knowledge levels across the three domains (

Table 4. Internal consistency and descriptive statistics for total and subscales.

Subscale	k	KR-20	Mean ± SD	Mean per Item
Breastfeeding Benefits	9	0.798	6.24 ± 2.15	0.69
Postpartum Physical Characteristics	3	0.712	2.18 ± 0.89	0.73
Breastfeeding Barriers	2	0.681	1.42 ± 0.68	0.71
Total Score	14	0.826	9.84 ± 3.28	0.70

Note: k = number of items; KR-20 = Kuder–Richardson 20; SD = Standard Deviation; Mean/k = Mean per item.

).

3.6. Convergent Validity

The Pearson correlation between the developed instrument and an established breastfeeding knowledge scale was r = 0.43 (p < 0.001), indicating a moderate positive relationship. This pattern supports convergent validity and implies that the instrument captures the shared domain of breastfeeding knowledge while maintaining specificity to GDM-related knowledge.

3.7. Final Instrument Description

The GDM-specific Breastfeeding Knowledge Test comprises 14 dichotomously scored items (0–14 range). Items are scored 1 for correct responses, 0 for incorrect/“do not know,” including two reverse-coded misconception items. The three-factor structure (9 breastfeeding benefits items, 3 postpartum physical characteristics items, 2 breastfeeding barriers items) explains 51.63% of variance with KR-20 = 0.826.

Based on recorded completion times, the instrument required a median of 7 min (range: 5–10 min) for completion, making it suitable for clinical and research applications in pregnant or postpartum women with GDM using paper or digital formats. The Korean version and scoring guidelines are in Appendix B.

4. Discussion

This study presents the first systematic development and validation of a GDM-specific breastfeeding knowledge instrument, addressing a notable measurement gap in perinatal care. Through a rigorous multi-stage process—literature review, expert content validation, pilot testing with experiential experts (women with GDM), and psychometric evaluation—we derived a concise, 14-item tool with acceptable statistical properties. Exploratory factor analysis supported a three-factor structure—Breastfeeding Benefits (9 items), Postpartum Physical Changes (3 items), and Breastfeeding Barriers (2 items)—explaining 51.63% of the variance. This variance explanation exceeds the conventional 50% threshold for adequacy in health measurement scales and aligns with comparable knowledge instruments in the literature [30]. Internal consistency was good (KR-20 = 0.826), meeting the recommended threshold of 0.70 for research instruments [26]. These results indicate that the instrument is both efficient and robust enough for use in routine clinical and research settings.

The three-factor structure provides a theoretically meaningful framework for understanding breastfeeding knowledge specific to women with GDM. Factor 1, “Breastfeeding Benefits,” was named to reflect its nine items addressing the protective metabolic effects of breastfeeding, including reduced risk of type 2 diabetes, improved glucose control, and prevention of obesity in both mothers and offspring. This factor accounted for the largest proportion of variance (34.38%), consistent with evidence that lactation intensity and duration are inversely associated with diabetes risk [15,16], and suggesting that knowledge about metabolic benefits should be a primary educational focus. Factor 2, “Postpartum Physical Characteristics,” comprises three items assessing knowledge about GDM-specific physiological changes affecting lactation, including reduced postpartum insulin requirements, the relationship between smoking and breastfeeding success, and the protective effect of breastfeeding against neonatal hypoglycemia. Factor 3, “Breastfeeding Barriers,” includes two items addressing common obstacles faced by women with GDM, specifically the impact of obstetric complications on feeding choices and delayed lactogenesis associated with maternal obesity.

The psychometric properties of the instrument support its use in both clinical and research settings. The KR-20 of 0.826 exceeds the 0.70 threshold recommended for research instruments [26] and indicates that the instrument yields consistent measurements of GDM-specific breastfeeding knowledge. The moderate correlation with the general breastfeeding knowledge scale (r = 0.43, p < 0.001) provides evidence of convergent validity while confirming that the instrument captures unique GDM-specific content not assessed by existing tools. These findings suggest that the instrument can reliably differentiate knowledge levels among women with GDM and may serve as an outcome measure for evaluating educational interventions.

Findings suggest that overall knowledge among women with GDM is moderate, with comparatively larger gaps in understanding postpartum physiological adaptations and potential barriers. These deficits mirror prior reports of GDM-specific concerns—such as fear of hypoglycemia and uncertainty about medication effects on milk—that can shape breastfeeding decisions [31,32]. The item difficulty distribution (means ranging from 0.20 to 0.80) indicated adequate spread without floor or ceiling effects, supporting discrimination across a range of knowledge levels [26]. The wide range of correct answer rates (21.4–75.0%) further confirms that the instrument effectively differentiates among participants with varying levels of breastfeeding knowledge. Notably, Q3, which addresses the clinical circumstances under which formula feeding may be recommended following obstetric complications such as preterm birth or postpartum hemorrhage, demonstrated the greatest difficulty level (21.4% correct). This finding suggests that women with GDM may have a limited understanding of how obstetric complications influence feeding decisions, highlighting an important area for targeted prenatal education. The moderate overall mean score (9.84 out of 14, or 70.3% correct) suggests that while participants demonstrated reasonable baseline knowledge, substantial room for improvement exists. Mean scores per item were relatively consistent across the three subscales (0.69–0.73), indicating similar knowledge levels across domains. The instrument provides actionable information for tailoring education: brief administration during prenatal visits can identify individual knowledge gaps, enabling targeted counseling that may reduce premature cessation and improve metabolic outcomes [33,34].

Several limitations should be acknowledged. First, the cross-sectional design and single-country sample limit generalizability. The majority of participants had a college education and perceived themselves as middle-class; item difficulty and factor structure may differ in populations with lower educational or socioeconomic backgrounds. Second, the “Breastfeeding Barriers” subscale comprises only two items. Although three items per factor is generally recommended, both items demonstrated strong factor loadings and captured clinically distinct GDM-specific barriers; deletion would have eliminated essential content. Future research should develop additional barrier items. Third, the reliance on exploratory factor analysis limits definitive conclusions about the measurement model. Future studies should conduct confirmatory factor analysis with diverse independent samples, assess test–retest and predictive validity, and pursue cross-cultural adaptation.

5. Conclusions

This study successfully developed and validated the first GDM-specific breastfeeding knowledge instrument, addressing a significant measurement gap in perinatal healthcare. The 14-item instrument demonstrates robust psychometric properties, with good internal consistency (KR-20 = 0.826), adequate construct validity (51.63% explained variance), and confirmed convergent validity (r = 0.429, p < 0.001).

The three-factor structure—Breastfeeding Benefits, Postpartum Physical Changes, and Breastfeeding Barriers—reflects the multidimensional nature of disease-specific knowledge essential for women with GDM. The instrument can be efficiently administered in approximately 5 min and provides a practical tool for healthcare professionals to systematically identify knowledge gaps and enable targeted educational interventions.

By improving knowledge about the protective effects of breastfeeding against postpartum metabolic complications, this instrument has the potential to enhance breastfeeding practices and improve long-term maternal metabolic health and infant well-being in women with gestational diabetes mellitus. However, future confirmatory factor analysis and cross-cultural validation are essential to establish measurement stability and generalizability.