Study profile
Figure 1 presents a summary of the study profile. Our data shows that 610 mother-baby pairs were enrolled to the EMTCT program between June 2014 and June 2018. Of those enrolled, 77 were excluded because they had transferred to other health facilities. Of the remaining 533 mother-baby pairs, 183 were excluded for the following reasons: 24 infants had died, 142 mother-baby pairs were lost to follow-up, and 17 infants had seroconverted and were started on ART. Of 350 mother-baby pairs, another 115 were excluded: 28 infants were below one year of age and 87 had gross missing data. The final number of records analysed was for 235 mother-baby pairs.
Sociodemographic characteristics of mothers living with HIV and their infants
Table 1 shows the sociodemographic characteristics of mother-baby pairs. Of 235 mothers living with HIV in the cohort, 138 (58.7 %) were aged 16 to 25 years, 69 (29.4 %) were multiparous, 119 (50.6 %) had not attended antenatal care during the recent pregnancy, 220 (96.7 %) had delivered in a health facility, and 116 (49.4 %) gave birth to a male infant. The mean age of all the participants was 25.2 ± 4.8 years but ranged from 16 to 39 years. The median age was 24 years (IQR: 22-28 years).
Bivariate analysis of cessation of breastfeeding at one year by maternal and infant factors
Table 2 summarized differences in cessation of breastfeeding at one year with respect to maternal and infant factors. Overall, our data showed that 150 (63.8 %) mothers living with HIV had ceased breastfeeding at one year. A higher proportion of cessation of breastfeeding at one year was observed among mothers living with HIV who were in the age category of 16 to 25 years (54.0 %), those who had not initiated breastfeeding on same-day as birth (98.7 %), those who had not attended antenatal care during the recent pregnancy (54.0 %), and those who had delivered in a health facility (96.6 %). Similarly, a higher proportion was observed among those who had given birth to a male infant (56.0 %) and among infants who had received cotrimoxazole prophylaxis before six weeks of age (52.7 %).
We observed statistically significant differences in cessation of breastfeeding with respect to initiation of breastfeeding on same-day as birth (p < 0.01), antenatal care attendance during recent pregnancy (p = 0.02), time of initiation of cotrimoxazole prophylaxis (p = 0.04) and infant sex (p = 0.002). Maternal age category demonstrated borderline statistical significance (p = 0.051). On average, the mothers who had ceased breastfeeding at one year were older than those who had not ceased breastfeeding: 25.7 ± 0.39 versus 24.3 ± 0.52 years respectively, (p = 0.027).
There was no statistically significant difference in cessation of breastfeeding with regard to parity, place of delivery, and ART regimen (all p > 0.05). Sensitivity analysis showed similar results of statistical significance for the variables infant sex and maternal ART regimen with cessation of breastfeeding in the presence and absence of missing data, suggesting that the results are robust to missing data.
Factors associated with cessation of breastfeeding at one year
In unadjusted analysis (Table 3), our data showed that cessation of breastfeeding at one year was more likely when the infant was male than female (Unadjusted RR [URR] 1.35; 95 % CI 1.10, 1.66), and when the mother was ≥ 25 years of age compared to those below 24 years of age (URR 1.21; 95 % CI 1.00, 1.46). Conversely, cessation of breastfeeding was less likely when cotrimoxazole prophylaxis was initiated at or after six weeks of birth relative to initiation of cotrimoxazole before six weeks of birth (URR 0.87; 95 % CI 0.79, 0.95), and when breastfeeding was initiated on same-day as birth compared to when it was on another day (URR 0.10; 95 % CI 0.03, 040). Antenatal care attendance during recent pregnancy was not associated with cessation of breastfeeding at one year (URR 1.05; 95 % CI 0.87, 1.28).
After adjusting for all statistically significant and clinically relevant factors, our results showed that cessation of breastfeeding at one year was more likely when the infant was male than female (aRR 1.25; 95 % CI 1.04, 1.50) and when the mother was multiparous than primiparous (aRR 1.26; 95 % CI 1.04, 1.53). Conversely, cessation of breastfeeding at one year was less likely when the infant was initiated on breastfeeding on same-day as birth compared to when the initiation of breastfeeding took place on another day (aRR 0.06; 95 % CI 0.01-0.41).
Qualitative findings
In qualitative data analysis, two major themes emerged regarding cessation of breastfeeding at one year: 1) reasons for ceasing breastfeeding at one year and this has four sub-themes, and 2) reasons for not ceasing to breastfeed at one year, with two sub-themes. These themes and sub-themes are described below.
Major theme 1: Reasons for ceasing to breastfeed at one year
Sub-theme 1: Male infants over breastfeed than females
Mothers living with HIV reported that the male infants breastfeed more frequently than female infants. Accordingly, the male infants were stopped from breastfeeding earlier than the female infants as illustrated in the excerpts below.
“Those boys can feed, they want to breastfeed every second and I feared my breast may get torn. Personally, I stopped breastfeeding boys at 9 months because feeding them needs too much” (FGD 1 with mothers living with HIV).
“They [meaning male infants] feed so much. We [meaning mothers living with HIV] don’t get peace at all (FGD 2 with mothers living with HIV).
Sub-theme 2: Maternal literacy and knowledge adequacy about breastfeeding
Participants indicated that the mother’s level of education and knowledge about breastfeeding an infant were important in influencing cessation of breastfeeding at one year. Cessation of breastfeeding was mostly observed among educated than non-educated mothers, and among mothers who had knowledge of breastfeeding an infant exposed to HIV compared to those who had no knowledge of breastfeeding.
“I cannot afford to buy those manufactured milk for my baby because I have to buy food, pay fees and treat the sisters and brothers, so I preferred to breastfeed until the time Musawo (meaning nurse) told me to stop breastfeeding. I was told to stop at 1 year sharp” (FGD 4 with mothers living with HIV).
“When mother is educated, she is able to understand what I say, she can even go ahead and read and understand more, she can see the pictures and charts we have here and is able to read compared to those who never went to school, we have to repeat several times before she can understand” (KII 1 with healthcare provider).
“Of course, those who are more educated do not disturb us, they follow whatever we say and even and want their babies healthy” (KII 2 with healthcare provider).
Sub-theme 3: Support and reminders from the partner
Participants mentioned that for married mothers, the partners played an important role in influencing cessation of breastfeeding through regular reminders on when to stop breastfeeding. In addition, the partners supported them in providing basic family needs. It arose that the partner support and reminders enabled mothers to breastfeed until one year.
“If you are not married you can breast feed your child to whatever time you want or stop whenever you want. But for the married women, your husband will ask you why the baby is not breastfeeding, your husband can also remind you of the recommended time of stopping to breastfeed” (FGD 3 with mothers living with HIV).
“A married woman is different from single one, a single mother has no time to breastfeed due to the responsibility of looking for basic needs, but married women get the support from the husband” ((FGD 2 with mothers living with HIV).
Sub-theme 4: Boys can bite the breast once they get teeth
Mothers reported that the male infants bite their breasts frequently than the female infants, especially when they develop teeth. Therefore, male infants were discontinued from breastfeeding much earlier than female infants.
“The boys, once they get teeth, they can bite so hard and it’s so painful and so I had to stop him from breast feeding” (FGD 1 with mothers living with HIV).
Major theme 2: Reasons for not ceasing to breastfeed at one year
Sub-theme 1: Insufficient knowledge about breastfeeding
It emerged that mothers who had missed routine health education sessions conducted at the health facility had inadequate knowledge about breastfeeding and this has resulted into not ceasing to breastfeed at one year. Second, key informants mentioned that primiparous and secundiparous mothers were less interested in breastfeeding compared to multiparous mothers.
“When mother is far, she comes late, some times when I have finished health education and she ends up missing. Somehow she may fail to follow the guideline and even breastfeed more than one year or believe that not breastfeeding when the teeth are out is the better” (KII 3 with healthcare provider).
“The girls who have one or two children do not want to breastfeed, they say that they do not want their breasts to fall” (KII 1 with healthcare provider).
Sub-theme 2: Girls feed a bit less
Mothers were concerned about the feeding habits of male versus female infants. Many of the mothers reported that the female infants breastfeed comparatively less than the male infants. Consequently, female infants were not discontinued from breastfeeding before one year compared to male infants.
“They [meaning male infants] feed so much. We [meaning mothers living with HIV] don’t get peace at all, the girls feed a bit less and we can manage our daily activates while breastfeeding” (FGD 2 with mothers living with HIV).
“The girls feed a bit less and we can manage our daily activates while breastfeeding” (FGD 2 with mothers living with HIV).