When analyzing Table 1, the results for adolescent primiparous women (n = 163) are as follows: The mean BFSES score is 56.02 ± 15.05, with a median of 63 and a mode of 70. The mean LATCH score is 8.3 ± 2.17, with both the median and mode at 10. The mean MSPSS score is 53.5 ± 19.6, with a median of 58 and a mode of 54.

When examining Table 2, it is noted that 4.3%, 21.5%, 38%, 18%, and 36.2% of adolescent primiparous women were 16, 17, 18, and 19 years old, respectively. No statistically significant differences were found in BFSES (KW = 2.07, p > 0.05), LATCH (F = 1.083, p > 0.05), and MSPSS (KW = 1.68, p > 0.05) values among these age groups. Although not statistically significant, the mean values of BFSES, LATCH, and MSPSS were lower for 16-year-old adolescent primiparous women.

Regarding educational background, 62.6% of the adolescents were primary school graduates, 20.2% were middle school graduates, and 17.2% were high school graduates. There were no statistically significant differences between education level and BFSES (F = 0.607, p > 0.05), LATCH (F = 0.093, p > 0.05), and MSPSS (F = 0.580, p > 0.05).

In terms of residence, 60.7% of primiparous adolescents lived in the province, 23.3% in the district, and 16.0% in the village. No statistically significant differences were observed between place of residence and BFSES (KW = 4.76, p > 0.05), LATCH (KW = 1.39, p > 0.05), and MSPSS (KW = 1.41, p > 0.05).

Regarding family income, 47.9% of participants had income less than expenses, 46.0% had income equal to expenses, and 6.1% had income more than expenses. There were no statistically significant differences between family income level and BFSES (KW = 1.14, p > 0.05), LATCH (F = 0.191, p > 0.05), and MSPSS (F = 1.125, p > 0.05).

Family type was also analyzed, with 36.8% of participants living in nuclear families and 63.2% in extended families. No statistically significant differences were found between family type and BFSES (t(161) = −0.341, p > 0.05), LATCH (t(161) = 0.089, p > 0.05), and MSPSS (t(161) = −0.131, p > 0.05).

Regarding breastfeeding education, 12.3% of participants had received breastfeeding education, while 87.7% had not. There were no statistically significant differences between breastfeeding education status and BFSES (t(161) = 0.217, p > 0.05), LATCH (t(161) = 0.800, p > 0.05), and MSPSS (t(161) = 0.711, p > 0.05).

Pregnancy planning was examined, with 50.9% of participants having planned pregnancies and 49.1% having unplanned pregnancies. No statistically significant differences were found between planned pregnancy status and BFSES (t(161) = 0.654, p > 0.05), LATCH (t(161) = 0.138, p > 0.05), and MSPSS (t(161) = 1.304, p > 0.05).

The mode of delivery was analyzed, with 52.1% of babies delivered via normal delivery and 47.9% via cesarean section. No statistically significant differences were found between the mode of delivery and BFSES (t(161) = 0.405, p > 0.05), LATCH (t(161) = 0.002, p > 0.05), and MSPSS (t(161) = 0.034, p > 0.05).

Regarding the gender of the babies, 48.5% were female and 51.5% were male. No statistically significant differences were observed between the baby’s gender and BFSES (t(161) = 0.957, p > 0.05), LATCH (t(161) = 0.471, p > 0.05), and MSPSS (t(161) = 0.241, p > 0.05).

When analyzing the timing of the first breastfeeding after birth, 68.1% of adolescents started breastfeeding within the first hour (60 min), 28.2% started within the first two or more hours (60 min later), and 3.7% did not breastfeed their baby at all. There was a statistically significant difference in BFSES based on the timing of the first breastfeeding (KW = 9.73, p < 0.05). However, no statistically significant differences were found for LATCH (KW = 3.62, p > 0.05) and MSPSS (KW = 2.28, p > 0.05). Adolescents who began breastfeeding within the first hour had higher BFSES levels compared to those who started later.

Regarding exclusive breastfeeding, 41.1% of participants reported exclusively breastfeeding their newborns, while 58.9% did not. Statistically significant differences were found between exclusive breastfeeding status and BFSES (t(161) = −5.017, p < 0.01), LATCH (U = 2054.0, p < 0.01), and MSPSS (t(161) = −3.442, p < 0.01). Mothers who exclusively breastfed their newborns had significantly higher breastfeeding self-efficacy, breastfeeding success, and perceived social support compared to those who did not exclusively breastfeed.

Table 3 shows the correlations between breastfeeding self-efficacy (BFSES) and various factors in adolescent primiparous women. BFSES levels were not significantly correlated with the mother’s age (r = 0.108, p > 0.05), education level (r = 0.083, p > 0.05), place of residence (r = −0.140, p > 0.05), family income level (r = 0.022, p > 0.05), family type (r = 0.027, p > 0.05), planned pregnancy status (r = −0.111, p > 0.05), breastfeeding education status (r = 0.001, p > 0.05), mode of delivery (r = −0.005, p > 0.05), or baby’s gender. However, there was a statistically significant positive correlation between BFSES and the following variables: MSPSS (r = 0.755, p < 0.05), LATCH (r = 0.832, p < 0.05), and exclusive breastfeeding status (r = 0.333, p < 0.05). Additionally, a strong negative correlation was found between BFSES and the first breastfeeding hour after birth (r = −0.350, p < 0.05).

Despite these significant correlations, multiple linear regression analysis was conducted to identify which factors most strongly influenced BFSES in adolescent primiparous women.

The regression model established using the enter method in Table 4 was statistically significant, indicating that at least one of the independent variables had a significant effect on the dependent variable (F = 137.109, p < 0.05). The independent variables accounted for 76.7% of the variance in the dependent variable, BFSES. Among the independent variables in the model, exclusive breastfeeding status did not have a statistically significant effect on BFSES (t = 1.791, p > 0.05). However, the first hour of breastfeeding after birth (t = −6.258, p < 0.05), MSPSS (t = 3.017, p < 0.05), and LATCH (t = 9.171, p < 0.05) were found to have statistically significant effects on BFSES.

The analysis revealed that increases in breastfeeding rates, MSPSS scores, and LATCH scores within the first hour after birth were associated with higher breastfeeding self-efficacy among adolescent primiparous women. A negative moderate correlation was found between BFSES and the first hour of breastfeeding after birth (r = −0.371 and r = −0.446), indicating that as the time spent breastfeeding within the first hour decreased, BFSES scores increased. A positive correlation was observed between BFSES and MSPSS (r = 0.755 and r = 0.233), suggesting that increased perceived social support was associated with higher breastfeeding self-efficacy. Additionally, a strong positive relationship was found between BFSES and LATCH (r = 0.832 and r = 0.589), indicating that improved breastfeeding success was associated with higher breastfeeding self-efficacy.

In summary, the model suggests that increased perceived social support (MSPSS) and breastfeeding success (LATCH), along with decreased time spent breastfeeding within the first hour, are associated with higher mean BFSES scores among primiparous women.

A regression tree analysis was conducted using the BFSES, with CHAID employed as the decision tree method. In the regression analysis in Table 5, all independent variables influencing the BFSES were incorporated into the regression tree model, and all variables were statistically significant except for the status of exclusive breastfeeding. A strong positive correlation was observed between the breastfeeding self-efficacy scale (BFSES) of the current sample and the model’s predicted values (r = 0.866, p < 0.001), with the model explaining 75% of the variance. This indicates that the breastfeeding self-efficacy level of an adolescent primiparous woman can be predicted with 75% accuracy based on the decision tree produced.

According to the analysis in Fig. 1, breastfeeding self-efficacy (BFSES) among adolescent primiparous women was categorized into two groups: those with a breastfeeding success score of less than 9.0 (n = 73, 44.8%) and those with a score of 9.0 or higher (n = 90, 55.2%). The mean BFSES for women with a breastfeeding success score below 9.0 was 42.79 ± 12.8, while those with a score of 9.0 or above had a mean BFSES of 66.73 ± 4.4. This difference was statistically significant (F = 273.133, p < 0.001), indicating that the BFSES scores for women with breastfeeding success below 9.0 were significantly lower.

Decision tree model for BFSES

Among those with a breastfeeding success score higher than 9.0 (n = 90, 55.2%), the sample was further divided based on the timing of breastfeeding initiation. Women who initiated breastfeeding within the first hour (n = 71, 43.6%) had a mean BFSES of 67.38 ± 3.4, whereas those who began breastfeeding after the first hour (n = 19, 11.7%) had a mean BFSES of 64.3 ± 6.6. The timing of breastfeeding initiation was found to be statistically significant for those with high breastfeeding success (greater than 9.0) (F = 7.661, p < 0.05), suggesting that initiating breastfeeding within the first hour positively affected both breastfeeding success and BFSES.

For those with a breastfeeding success score of less than 9.0 (n = 73, 44.8%), the sample was categorized based on perceived social support (MSPSS). The mean BFSES decreased as perceived MSPSS levels declined, which was statistically significant (F = 20.09, p < 0.001). Specifically, women with an MSPSS score below 23.00 (n = 15, 9.2%) had a mean BFSES of 32.26 ± 7.6; those with an MSPSS score between 23.00 and 53.00 (n = 36, 22.1%) had a mean BFSES of 40.66 ± 11.8; and those with an MSPSS score above 53.00 (n = 22, 13.5%) had a mean BFSES of 53.45 ± 9.3. Thus, perceived social support was crucial for adolescent primiparous women with a breastfeeding success score of less than 9.0.