Calculating Exclusive Breastfeeding Rates: Comparing Dietary “24-Hour Recall” with Recall “Since Birth” Methods

Abstract

Background:

Calculating exclusive breastfeeding (EBF) rates based on the previous-day recall has been recommended by the World Health Organization to avoid the recall bias but it also may not accurately reflect the feeding pattern since birth and leads to overestimate of the proportion of exclusively breastfed infants. The objective of this study was to compare the (EBF) rates calculated by the 24-hour recall and since birth recall and their association with different sociodemographic and maternal data.

Subjects and Methods:

Prospective descriptive study in Mansoura District including 1,102 mother-infant dyad attending primary healthcare centers for vaccination. One thousand ninety-one and 1,029 were followed up at 4 and 6 months during a period from January to October 2015. Sociodemographic data, maternal, antenatal, birth, and some infant related data were collected through interview. Questions about EBF using the 24-hour recall and since birth recall definitions were asked.

Results:

This study shows consistent difference in breastfeeding pattern reported by 24-hour recall with recall since birth at all age intervals. At the age of 6 months 13.6% of infants were EBF as reported by 24-hour recall method versus 5.2% for recall since birth method. Different factors were associated with EBF practice reported using these different methods.

Conclusions:

The two recall methods describe the reality in different and incomplete ways. It is better to measure and report EBF rated using both methods so as to give a full picture of breastfeeding practice. And it is very important to distinguish between both methods and not to be used interchangeably with each other.

Introduction

According to the World Health Organization (WHO), exclusive breastfeeding (EBF) is defined as infant consumption of human milk with no supplements of any type except for vitamins, minerals, and medications, with no other food or fluids (including plain water and juices).¹ It recommends EBF during the first 6 months of an infant's life in both developed and developing countries.² Although highly recommended, still less than 40% of infants under 6 months of age in the developing world are exclusively breastfed.³

Studies describing the pattern and factors affecting the duration of EBF are mostly cross-sectional and based on maternal recall.⁴ The validity of data generated by such cross-sectional studies is questionable since long time and the data collected retrospectively specially with long intervals between data collection and the actual practice of breastfeeding may be subjected to maternal recall bias.^5–7 Thus, and to avoid this limitation, the previous-day recall has been found appropriate and has been recommended by the WHO in breastfeeding surveys describing infant feeding practices.⁸

However, the indicators derived from the previous-day recall has its own limitations too and may not accurately reflect the feeding pattern since birth⁹ as practices vary widely from day to day.⁸

To the best of the authors’ knowledge there are no studies that compare the two different methods of maternal recall for EBF (on the previous 24 hours and since birth) in Egypt. This study aims to compare the EBF rates calculated by the 24-hour recall and since birth recall and their variation with different sociodemographic and maternal data.

Population and Methods

This was a prospective descriptive study done in Mansoura District, about 150 km North to Cairo, during a period from January to October 2015. The target population was healthy infants and their mothers. Mothers were interviewed at 2, 4, and 6 months after birth during the vaccination sessions at the health facilities.

A pilot study was done on 100 infants, not included in the full-scale study (50 urban and 50 rural) to test questionnaire, train nurses on data collection, and estimate the expected EBF rate at age of 6 months.

Sample size was calculated online (www.dssresearch.com/KnowledgeCenter/toolkitcalculators/samplesizecalculators.aspx). From the pilot study EBF rate was about 15% based on a 24-hour recall at age of 6 months, with alpha error = 5%, Given the study power = 80% and a precision of 3%, the sample size needed was calculated to be at least 823.

Twenty percent was added to compensate for drop outs. Thus, the final sample size was 988.

There are 50 primary healthcare facilities in Mansoura District, 11 urban and 39 rural. A cluster of 20–25 infants was selected from each health facility. A total of 1,102 mother-infant dyad were recruited in the study and completed the questionnaire at the initial visit. However, 1,091 and 1,029 were followed up at 4 and 6 months (response rates of 99% and 93.4%; respectively).

Trained nurses interviewed mothers during the vaccination sessions and completed the questionnaire. The questionnaire covered the sociodemographic data of the mother and her family, number of antenatal care visits along with place and mode of delivery. Infant's birth order, sex, birth weight, and gestational age in addition to history of infant hospitalization and giving prelacteal feed were recorded. This questionnaire was completed at the initial visit at age of 2 months. At the three visits mothers were asked two questions for measuring EBF using the 24 hour-recall and since birth recall definitions; respectively: (1) if any fluids or foods items (other than drugs and vitamins) had been given to the infant, during the previous 24 hours (from the previous morning to the morning of the interview), and (2) if any fluids or foods items (other than drugs and vitamins) had been given to the infant since birth up to the time of interview.

The outcome variable is the EBF rate as measured by both the 24-hour recall method and the recall since birth method.

Ethical consideration

The study was approved by the IRB of Faculty of Medicine, Mansoura University. Mothers gave informed verbal consent to participate in the study, before the interview.

Data analysis

Data were analyzed using SPSS version 20. Variables were described as number and percent. In categorical variables χ² test and unadjusted hazard risk (HR) were used for comparison between groups.

Significant predictors of EBF in bivariate analysis were entered into a logistic regression using the forwards Wald methods and adjusted HR was calculated. p ≤ 0.05 was considered statistically significant.

Results

Table 1 shows the EBF rate by the 24-hour recall were 97.5%, 69.3%, and 13.6% at 2, 4, and 6 months; respectively. The corresponding rates for the recall since birth were 56.3%, 30.4%, and 5.2%; respectively.

Table 1.

Exclusive Breastfeeding at Different Ages of Infants

		Exclusive breastfeeding
Age (months)	Number	24-hour recall N (%)	Recall since birth N (%)
2	1,102	1,074 (97.5)	620 (56.3)
4	1,091	756 (69.3)	332 (30.4)
6	1,029	140 (13.6)	53 (5.2)

Table 2 reveals that educational level of both parents are very important factors determining the 24-hour recall for exclusive breastfeeding, where mothers at the level of secondary and less than secondary education recalled EBF 3.1and 2.3 times more than those with higher level of education. Infants born to less than secondary and secondary educated fathers were 3 and 1.7 times exclusively breastfed more than those born to highly educated fathers as detected by 24-hour recall. Previous-day recall was inversely associated with the socioeconomic class, mothers from very low, low, and middle social classes recalled EBF 9.9, 4.8, and 4.3-folds compared to mothers from high social class.

Table 2.

Association of Family Sociodemographics with Exclusive Breastfeeding at Age of 6 Months

		24-hour recall		Recall since birth
	Total	N (%)	HR (95% CI)	N (%)	HR (95% CI)
Overall	1,029	140 (13.6)	—	53 (5.2)	—
Residence
Rural	574	78 (13.6)	0.99 (0.7–1.4)	37 (6.4)	1.8 (1.03–3.3)^a
Urban	455	62 (13.6)	1 (r)	16 (3.5)	1 (r)
Mother's work
Housewife	807	140 (17.3)	Undefined	53 (6.6)	Undefined
Working	222	0		0
Mother's education
<Secondary	138	20 (14.5)	2.3 (1.3–4.0)^a	14 (10.1)	3.1 (1.5–6.4)^a
Secondary	493	95 (19.3)	3.1 (2.0–4.7)^a	26 (5.8)	1.6 (0.8–3.1)
>Secondary	398	25 (6.3)	1 (r)	13 (3.3)	1 (r)
Father's work
Farmer/MW	467	79 (16.9)	1.1 (0.7–1.5)	28 (6.0)	1.9 (0.9–4.3)
Prof/Semi-prof.	337	25 (7.4)	0.4 (0.3–0.8)	18 (5.3)	1.7 (0.7–4.0)
Others^b	225	36 (6.0)	1 (r)	7 (3.1)	1 (r)
Father's education
<Secondary	162	41 (25.3)	3.0 (2.0–4.6)^a	9 (5.6)	1.1 (0.5–2.4)
Secondary	446	64 (14.3)	1.7 (1.2–2.5)^a	25 (5.2)	1.0 (0.6–1.8)
> Secondary	421	35 (8.3)	1 (r)	21 (5.0)	1 (r)
Socioeconomic class
Very low	80	26 (32.5)	9.9 (4.6–21.0)^a	3 (3.8)	0.9 (0.3–3.2)
Low	372	59 (15.9)	4.8 (2.4–9.9)^a	26 (7.0)	1.7 (0.8–3.4)
Middle	333	47 (14.1)	4.3 (2.1–8.9)^a	14 (4.2)	1.0 (0.5–2.3)
High	244	8 (3.3)	1 (r)	10 (4.1)	1 (r)

Significant compared to the reference group.

Others as trades, business, etc.

CI, confidence interval; HR, hazard ratio; MW, manual workers; Prof., professional; r, reference group.

On the other hand, rural residence and mother's education less than secondary level were significantly associated with increased since birth recall compared to other groups with HR (95% CI) = 1.8 (1.03–3.3) and 3.1 (1.5–6.4), respectively.

The number of antenatal visits was detected as a significant predictor where antenatal care visits less than 10 (<5 and 5–9) were associated with less frequent 24-hours recall HR (95% CI) = 0.9 (0.6–1.4) and 0.5 (0.4–0.8) respectively (Table 3). Also, home delivery was reported as a significant predictor for exclusive breastfeeding in both methods of recall as it was reported 1.6 times more in 24-hour recall and by 2.9 times more in since birth recall compared to private clinic or hospital delivery. Also, nonhospitalization was a strong predictor for 24-hour recall with HR (95% CI) = 4.4 (1.1–17.4).

Table 3.

Association of Maternal and Infant Characteristics with Exclusive Breastfeeding at Age of 6 Months

		24-hour recall		Recall since birth
	Total	N (%)	HR (95% CI)	N (%)	HR (95% CI)
Mother's age (years)
<20	422	56 (13.3)	0.96 (0.5–1.7)	23 (5.5)	1.0 (0.4–2.6)
20–	513	71 (13.8)	1.0 (0.6–1.7)	25 (4.9)	0.9 (0.4–2.3)
35 and more	94	13 (8.1)	1 (r)	5 (5.3)	1 (r)
Antenatal visits
<5	188	33 (17.6)	0.9 (0.6–1.4)	10 (5.3)	0.7 (0.4–1.6)
5–9	588	59 (10.0)	0.5 (0.4–0.8)^a	25 (4.3)	0.6 (0.3–1.1)
10 and more	253	48 (19.0)	1 (r)	18 (7.1)	1 (r)
Place of delivery
Home	123	27 (22.0)	1.6 (1.1–2.4)^a	14 (11.4)	2.9 (1.5–5.4)^a
Governmental hospital	274	28 (10.2)	0.8 (0.5–1.1)	14 (5.1)	1.3 (0.7–2.4)
Private clinic/hospital	632	85 (13.4)	1 (r)	25 (4.0)	1 (r)
Mode of delivery
Caesarean section	332	35 (10.5)	0.7 (0.5–1.0)	18 (5.4)	1.1 (0.6–1.9)
Vaginal delivery	697	105 (15.1)	1 (r)	35 (5.0)	1 (r)
Birth order
First-born	392	52 (13.3)	0.9 (0.6–1.6)	24 (6.1)	0.7 (0.4–1.5)
2 and 3	516	71 (13.8)	1.0 (0.6–1.6)	19 (3.7)	0.4 (0.2–0.9)
4 and more	121	17 (14.0)	1 (r)	10 (8.3)	1 (r)
Sex
Male	552	73 (13.2)	0.9 (0.7–1.3)	26 (4.7)	0.8 (0.5–1.4)
Female	477	67 (14.0)	1 (r)	27 (5.7)	1 (r)
Low birth weight
No	956	136 (14.2)	2.6 (1.0–6.8)	48 (5.0)	0.7 (0.3–1.8)
Yes	73	4 (5.5)	1 (r)	5 (6.8)	1 (r)
Preterm
No	896	128 (14.3)	1.5 (0.9–2.8)	43 (4.8)	0.6 (0.3–1.3)
Yes	133	12 (9.0)	1 (r)	10 (7.5)	1 (r)
Hospitalization
No	967	138 (14.3)	4.4 (1.1–17.4)^a	50 (5.2)	1.1 (0.3–3.3)
Yes	62	2 (3.2)	1 (r)	3 (4.84.6)	1 (r)
Prelacteal feed
No	581	70 (12.5)	0.8 (0.6–1.0)	53 (9.4)	Undefined
Yes	448	70 (15.6)	1 (r)	0

Significant compared to the reference group.

NIC, neonatal intensive care.

The logistic regression revealed independent predictors of 24-hour recall and recall since birth (Table 4). Regarding 24-hour recall, mother's education at secondary level encounters for about double recall (HR [95% CI] = 2.0 [1.1–3.5]), lower socioeconomic class positively affect recall where mothers from very low, low, and middle classes reported exclusive breastfeeding in the last 24 hours significantly 15.4, 3.7, and 3.6 times respectively more than mothers from high socioeconomic class. Utilizing antenatal care visits five to nine times during pregnancy was found to be the significant independent factor for lower recall of exclusive breastfeeding (HR [95% CI] = 0.5 [0.3–0.8]).

Table 4.

Logistic Regression Analysis of the Independent Significant Predictors of Exclusive Breastfeeding

	24-hour recall		Recall since birth
	β	HR (95% CI)	β	HR (95% CI)
Mother's education
<Secondary	0.14	0.9 (0.4–2.0)	0.99	2.7 (1.4–5.2)^a
Secondary	0.7	2.0 (1.1–3.5)^a	1.4	3.9 (2.4–6.2)^a
>Secondary	—	1 (r)	—	1 (r)
Socioeconomic class
Very low	2.7	15.4 (5.5–43.5)^a
Low	1.3	3.7 (1.5–9.0)^a
Middle	1.3	3.6 (1.6–8.3)^a
High	—	1 (r)
Antenatal visits
<5	−0.1	0.9 (0.6–1.6)
5–9	−0.7	0.5 (0.3–0.8)^a
10 and more	—	1 (r)
Place of delivery
Home			0.3	1.4 (1.1–2.4)^a
Governmental hospital			−0.6	0.5 (0.3–1.9)
Private clinic/hospital			—	1 (r)
Constant	−5.01			−2.66
Model χ²	97.8, p ≤ 0.001			46.2, p ≤ 0.001
Percent correctly predicted	86.1			86.4

Significant compared to the reference group.

On the other hand, both secondary and less than secondary levels of maternal education found to be significant independent predictors for since birth recall (3.9 and 2.7) respectively times more compared to the reference group. Also, delivery at home found to be significant independent factor for more frequent since birth recall of exclusive breastfeeding (HR [95%CI] = 1.4 [1.1–2.4]).

Discussion

This study compares lifelong EBF practice with EBF from the previous morning to the morning of the interview in the same cohort of women. It shows consistent difference at all age intervals that can be attributed to either overestimating the EBF practice by current status recall or underestimating it by since birth recall or both.

This considerable degree of disagreement was detected in other studies. For example, in a Swedish study the rates of previous day and since birth recall decreased from 11% to 1.8%, respectively.⁹ According to Engebretsen et al., after the age of 6 months 50% of infants were exclusively breastfed as reported by 24-hour recall compared to 0% as reported by recall since birth.¹⁰ In Sri Lanka, the 24-hour recall of EBF was compared with since birth EBF recall and it was 77.4% versus 49.1%; respectively.¹¹ Also, many other studies showed that 24-hour recall method overestimates the EBF rate, compared to the “recall since birth” method.^12,13

This overestimation can be explained simply with the complexity of infant feeding patterns. According to UNICEF,¹⁴ even under optimum conditions it is not possible to achieve 100% of EBF for many reasons like postpartum complications, medication taken by the mother or child sickness leading to moving in and out of feeding categories.

Also, over-reporting associated with face to face interview and the desire to give preferred and socially accepted answer regarding current breastfeeding practice was suggested in an earlier study.¹⁵

In addition, potential maternal recall bias related to since birth recall may be a reason for the difference between both rates. However, recall bias was limited in this study by the short duration of recall and we consider data detected by since birth recall to be reliable.

Initiation and duration of EBF are known to be affected by many factors related to the infant, the mother, and their sociodemographic.

In this study the independent factors associated with EBF at the age of 6 months differed according to the recall method. This brings attention to the great difference between recall methods in rates and the associated factors.

For example, EBF detected by since birth recall was significantly and independently associated with lower maternal educational levels, which can be related to higher employment rated among mothers with higher levels of education and the cessation of exclusive breastfeeding with their return to work after delivery. Also, well-educated mothers are more anxious about adequacy of breastmilk, resulting in a lower threshold to stop breastfeeding.

Also, home delivery was found to be significant predictor for since birth EBF recall when compared to delivery in private sector. Home delivery in Egypt can be considered as an indirect indicator for lower socioeconomic class and its relation to practice exclusive breastfeeding for longer duration.

On the other hand, higher rates of EBF by 24-hour recall showed independent and dose-related association with lower socioeconomic class in comparison with high class. This can be explained by the inconsistency of infants feeding pattern especially in lower social classes because in Egypt it is not unusual for families to run out of money at the end of the month and thus a mixed fed baby might, therefore, be exclusively breastfed at some days of the month because of inadequate finances.

This difference in the EBF rates between both methods highlights the importance of selecting the appropriate survey method based on study objectives. Knowing only the current status EBF rate is inadequate to study the relation between infant feeding practice and certain health outcomes. On the other hand, inaccuracy of breastfeeding information obtained using maternal recall should be considered and it is better to collect data at shorter durations.

Also, it is very important to interpret data depending on the clearly stated survey method. Misinterpreting 24-hour recall results as if since birth recall prevalence dangers the proper understanding of the true status of EBF, and the appropriate design of policies and programs.

To make fair international comparison between countries in relation to rates and indicators of EBF, it is very important to clearly state the method of data collection and to interpret the results accordingly.

Conclusion

In conclusion current EBF status does not accurately represent the feeding pattern since birth, and long-term EBF recall tends to overestimate the duration of EBF.

Recommendation

Feeding data should, ideally, be collected by prospective longitudinal studies to assemble a lifetime feeding history. It is better to measure and report EBF rates using both methods so as to give a full picture of breastfeeding practice. And it is very important to distinguish between both methods and not to be used interchangeably with each other.

Study Limitations

This is a small-scale study, in one region of Egypt, and its finding cannot be generalized to all over Egypt.

Footnotes

Disclosure Statement

No competing financial interests exist.

References

World Health Organization: Indicators for Assessing Infant and Young Child Feeding Practices. Washington DC, USA: WHO, 2008.

Kramer

, Kakuma

. The Optimal Duration of Exclusive Breastfeeding: A Systematic Review. Geneva: World Health Organization, 2002.

UNICEF: Maternal and Newborn Health. The State of The World's Children, 2008 www.unicef.org/sowc08/docs/sowc08.pdf (accessed May 18, 2016).

Santos Neto

, Oliveira

, Zandonade

, et al. Pacifier use as a risk factor for reduction in breastfeeding duration: A systematic review. Rev Bras Saude Mater Infant, 2008; 8:377–389.

Yoon

, Black

, Moulton

, et al. Effect of not breastfeeding on the risk of diarrheal and respiratory mortality in children under 2 years of age in Metro Cebu, the Philippines. Am J Epidemiol, 1996; 143:1142–1148.

Beaudry

, Marcoux

. Relation between infant feeding and infections during the first six months of life. J Pediatr, 1995; 126:191–197.

Faber

, Benade

. Nutritional status and dietary practices of 24-month-old children from a rural South African community. Public Health Nutr, 1999; 2:179–185.

World Health Organization: Indicators for Assessing Infant and Young Child Feeding Practices. Part 2 Measurement, 2010.

Aarts

, Kylberg

, Hornell

, et al. How exclusive is exclusive breastfeeding? A comparison of data since birth with current status data. Int J Epidemiol, 2000; 29:1041–1046.

10.

Engebretsen

IMS

, Wamani

, Charles Karamagi

, et al. Low adherence to exclusive breastfeeding in Eastern Uganda: A community-based cross sectional study comparing dietary recall since birth with 24-hour recall. BMC Pediatr, 2007; 7:10.

11.

Agampodi

, Fernando

, Dharmaratne

. Comparison of definitions used for exclusive breast feeding-the Labbock's strict definition and the WHO definition as methods of estimating exclusive breastfeeding rates in Sri Lanka. Ceylon Med J, 2009; 54 (Suppl 1):20.

12.

Zohoori

, Popkin

, Fernandez

. Breast-feeding patterns in the Philippines: A prospective analysis. J Biosoc Sci, 1993; 25:127–138.

13.

Ssenyonga

, Muwonge

, Nankya

. Towards a better understanding of exclusive breastfeeding in the era of HIV/AIDS: A study of prevalence and factors associated with exclusive breastfeeding from birth, in Rakai, Uganda. J Trop Pediatr, 2004; 50:348–353.

14.

UNICEF: Progress for children, report card on nutrition: number 4. 2006. http://www.unicef.org/progressforchildren/2006n4/index_breastfeeding.html (accessed March 12, 2016).

15.

Waruru

, Nduati

, Tylleskar

. Audio computer-assisted Self interviewing (ACASI) may avert socially desirable responses about infant feeding in the context of HIV. BMC Med Inform Decis Mak, 2005; 5:24.