**Background:** Breastfeeding provides optimal infant nutrition and is linked to reduced child mortality, yet global rates remain suboptimal. In Jordan, only 25% of infants 0–5 months were breastfed (any breastfeeding) as of 2022, lower than many neighboring countries. Maternal attitude toward breastfeeding is a known predictor of intention, initiation, and continuation. However, research on determinants of breastfeeding attitudes in Jordan is scarce. This study aimed to assess postnatal mothers' attitudes toward breastfeeding and identify their sociodemographic, pregnancy, and delivery-related determinants.
**Methods:** A hospital-based cross-sectional study was conducted at a Baby-friendly accredited tertiary teaching hospital in Amman, Jordan, from February 2021 to January 2022. A convenience sample of 301 postnatal mothers was recruited. Data were collected via interviewer-administered questionnaires using the validated Arabic version of the Iowa Infant Feeding Attitude Scale (IIFAS), a 17-item, 5-point Likert scale (total score range 17–85; higher scores = more positive toward breastfeeding). Sociodemographic data, pregnancy/delivery outcomes, and breastfeeding practices were also collected. Cronbach's alpha for the IIFAS was 0.72. Statistical analysis included descriptive statistics, independent t-tests, one-way ANOVA, and binary logistic regression. P-values <0.05 were considered significant.
**Key Results:** The mean total IIFAS score was 65.0 ± 7.2 (out of 85), falling within the upper limit of the neutral range (49–69). Only 24.3% of mothers had a positive attitude toward breastfeeding (score ≥70). The highest-scoring items were "Breastfeeding increases mother-infant bonding" (mean 4.9 ± 0.6) and "Breast milk is more easily digested than formula" (mean 4.7 ± 0.8). The lowest-scoring items were "Mothers who occasionally drink alcohol should not breastfeed" (mean 1.4 ± 1.1) and "Women should not breastfeed in public places" (mean 2.0 ± 1.5).
Bivariate analysis showed significantly higher total attitude scores among mothers with higher income (p = 0.05), those who had pregnancy complications (p = 0.05), delivery complications (p = 0.01), preterm infants (p = 0.04), those who intended to breastfeed (p = 0.002), and those willing to breastfeed exclusively (p = 0.01). No significant associations were found for age, parity, education, employment, mode of delivery, NICU admission, breastfeeding counseling, or time of initiation.
Binary logistic regression (χ²(9) = 21.7, p = 0.01; Nagelkerke R² = 0.284; 82% correctly classified) identified two independent predictors of positive breastfeeding attitude: highest family income (>651 JD/month) had 14.77 times the odds of positive attitude compared to lowest income (0–300 JD) (95% CI: 2.25–99.64, p = 0.004), and willingness to breastfeed exclusively was associated with 3.41 times higher odds (95% CI: 1.35–8.63, p = 0.01). The middle-income group (301–450 JD) also showed significantly higher odds (OR = 4.88, 95% CI: 1.12–23.8, p = 0.05).
**Clinical Implications:** The neutral attitude toward breastfeeding among Jordanian mothers, with only one-quarter showing positive attitudes, underscores a critical gap. The strong association between higher income and positive attitude suggests that low-income mothers may perceive formula as an ideal food they cannot afford, rather than preferring breastfeeding. The lack of association with education or employment indicates that knowledge alone is insufficient. The finding that willingness to exclusively breastfeed predicts positive attitude highlights the importance of strengthening prenatal intention-building. The low rate of early initiation (43% on Day 1) and high rate of no initiation before discharge (28.6%) in a Baby-friendly hospital suggest that institutional policies are not fully translated into practice. Interventions should target low-income communities, involve family members, enforce workplace breastfeeding support, and strengthen hospital breastfeeding protocols. Limitations include a smaller-than-planned sample due to COVID-19, potential social desirability bias from interviewer-administered surveys, and limited generalizability due to the single-site, teaching hospital setting with a highly educated sample.