Infant Honey Feeding and Associating Sociodemographic and Clinical Factors: Are there risks for infant botulism in Oman?

TL;DR

This study in Oman found that nearly 40% of infants were fed honey in their first year, which could increase the risk of infant botulism.

Contribution

The study identifies sociodemographic and clinical factors linked to infant honey feeding in Oman.

Findings

38.7% of infants were fed honey during their first year, often for cough relief or cultural practices.

Higher maternal education and income were associated with lower odds of feeding honey to infants.

Infants fed honey had a higher likelihood of constipation and muscle weakness.

Abstract

This study aimed to assess the risks associated with infant botulism in Oman by examining honey-feeding practices among infants and related sociodemographic and clinical factors.

This cross-sectional study included Omani mothers with children under 5 years and were conveniently recruited from health centres. Mothers completed an infant feeding practices and beliefs survey between October 2023 and March 2024. Data were represented as frequencies and percentages and analysed via Chi-square test and logistic regression.

A total of 3,222 participants were included of which 58.8% were aged 26–35 years. The study found that 38.7% of children were fed honey during their first year. Local Omani honey was used by 34.6%, often for reasons including cough relief (15.2%), boosting immunity (11.6%) and Tahneek (8.6%). Additionally, 16.7% of mothers felt pressured to introduce honey and 45.9% viewed early honey feeding as acceptable. Multivariate logistic regression analysis revealed that infant honey feeding was negatively and significantly associated with maternal education (odds ratio [OR] = 0.77, 95% confidence interval [CI]: 0.65–0.91; P <0.001), family income (OR = 0.61, 95% CI: 0.51–0.74; P <0.001) and receipt of infant feeding counseling (OR = 0.83, 95% CI: 0.72–0.97; P = 0.007). Additionally, results indicated that infants who were fed honey had a higher likelihood of experiencing constipation (OR = 1.29, 95% CI: 1.10–1.51; P = 0.001) and muscle weakness (OR = 1.84, 95% CI: 1.15–2.95; P = 0.011).

This study highlights the potential risks of infant botulism in Oman, emphasising the importance of following updated clinical guidelines for optimal infant care and effective reporting.

Tables

Table 1.. Characteristics of participants (N = 3,222).

Characteristic	n (%)
Governorate
Muscat	247 (7.7)
Dakiliya	637 (19.8)
Dhofar	93 (2.9)
North Batinah	454 (14.1)
South Batinah	602 (18.7)
North Sharqia	374 (11.6)
South Sharqia	343 (10.6)
Dahera	288 (8.9)
Buraimi	184 (0.7)
Infant age in months *
Less than 6	97 (3.0)
6–11	98 (3.0)
12–17	1,027 (31.9)
18–23	872 (27.1)
24–29	842 (26.1)
30–36	218 (6.8)
Unknown but under 36 months	68 (2.1)
Infant gender
Male	1,671 (51.9)
Female	1,506 (46.7)
Unknown	45 (1.4)
Mother's age in years
18–25	451 (14.0)
26–35	1,894 (58.8)
36–45	817 (25.4)
46–55	23 (0.7)
Unknown	37 (1.1)
Mother's education level
Less than high school	198 (6.1)
High school	1,189 (36.9)
Diploma/high diploma	736 (22.8)
Baccalaureate	1,029 (31.9)
Master/PhD	63 (2)
Unknown	7 (0.2)
Father's education level
Less than high school	308 (9.6)
High school	1,529 (47.5)
Diploma/high diploma	530 (16.4)
Baccalaureate	734 (22.8)
Master/PhD	96 (3)
Unknown	25 (0.8)
Number of children
1–3	2,048 (63.6)
4–6	964 (29.9)
7–9	125 (3.9)
>9	35 (1.1)
Unknown	50 (1.5)
Family income in OMR
<500	710 (22.0)
500–999	1,382 (42.9)
1,000–1,999	909 (28.2)
2,000–3,999	134 (4.2)
≥4,000	43 (1.3)
Unknown	44 (1.4)

Table 2.. Infant honey feeding practice (N = 3,222).

Practice	n (%)
Times fed bee honey
NA	1,977 (61.4)
Once or twice	473 (14.7)
More than twice	772 (24.0)
Person introduced bee honey
NA	1,952 (60.6)
Mother	755 (23.4)
Other than the mother	474 (14.7)
Unknown	41 (1.3)
Way bee honey introduced
NA	1,931 (59.9)
By itself	474 (14.7)
Mixed with milk	143 (4.4)
Mixed with herbs	71 (2.2)
Mixed with food	338 (10.5)
Dipped pacifier	118 (3.7)
Other	76 (2.4)
Unknown	71 (2.2)
Indication for feeding bee honey *
NA	1,892 (58.7)
Diarrhoea	92 (2.9)
Boost Immunity	374 (11.6)
Tahneek	276 (8.6)
Colic	131 (4.1)
Cough	491 (15.2)
To make food palatable	81 (2.5)
No reason	264 (8.2)
Kind of honey introduced
NA	1,921 (59.6)
Local (Omani)	1,115 (34.6)
Foreign	6 (0.2)
Can't be identified	143 (4.4)
Unknown	37 (1.1)
Pressure to introduce honey
No	2,632 (81.7)
Yes	537 (16.7)
Unknown	53 (1.6)
Belief feeding bee honey cause no harm
Strongly disagree	755 (23.4)
Disagree	972 (30.2)
Agree	1,300 (40.3)
Strongly agree	179 (5.6)
Unknown	16 (0.5)

Table 3.. Infant honey feeding logistic regression analysis.

Characteristic	X2	P value	Crude OR (95% CI)	Adjusted OR (95% CI)
Mother's age in years	4.65	0.03	1.20 (1.02–1.41)	1.07 (0.90–1.27)
18–35
36–55
Mother's education	34.87	<0.001	0.65 (0.56–0.75)	0.77 (0.65–0.91)*
High school or less
More than high school
Father's education	8.19	0.004	0.81 (0.70–0.94)	1.09 (0.92–1.29)
High school or less
More than high school
Family income in OMR	55.89	<0.001	0.55 (0.47–0.65)	0.61 (0.51–0.74)*
<1,000
≥1,000
Receipt of infant feeding counselling	7.27	0.007	0.82 (0.71–0.95)	.83 (0.72–0.97)*
No
Yes
Constipation	11.16	0.001	1.30 (1.11–1.51)	1.29 (1.10–1.51)*
No/rarely
Sometimes/frequently
Muscle weakness	6.54	0.011	1.79 (1.14–2.80)	1.84 (1.15–2.95)*
No
Yes

Full text

1. Introduction

Infant botulism (IB) is a rare neuromuscular illness caused by Clostridium botulinum, an anaerobic bacterium that colonises the large intestine, adhering to the intestinal epithelium and producing neurotoxins that transcytose across the epithelium.^1^ The neurotoxins bind at the neuromuscular junction and block the release of acetylcholine. Infants with botulism present initially with constipation without fever, then symptoms progress to lethargy, poor feeding, hypoglycaemia, hypotonia, paralysis, cough reflex loss and respiratory failure. Stool specimens contaminated with C. botulinum spores or toxins are the confirmative diagnostic evaluation for IB.^23^

From 2007 to 2021, a total of 2,943 cases of IB were recognised globally, with the majority being type A (54%) and type B (43%).^4^ Surveillance conducted in California, USA, reported 1,345 cases of IB between 1976–2016, where almost 90% of the cases were younger than 6 months old.^5^ In Canada, 63 laboratory-confirmed cases of IB were confirmed from 1979 to 2019 with an average age of four months.^6^ The incidence of IB in the Gulf Cooperation Council region is sporadic. Based on literature, 2 case were reported, 1 in Kuwait and another in the United Arab of Emirate (UAE) in the past 20 years, with no reported cases in Oman, Saudi Arabia, Bahrain or Qatar.^78^

Case reports identified a number of potential sources to IB, including honey ingestion, honey-containing pacifiers, dust-related to home remodelling and chamomile tea.^29^ A global study examining the prevalence of IB from 2007 to 2021 found that 20% of cases were associated with honey consumption.^4^ The link between honey ingestion and IB was explained by the honey's likelihood to host C. botulinum spores, which can germinate and develop in the infant's immature digestive tract.^10^

Despite efforts to purify honey for the treatment of various inflammatory conditions in neonates and infants, such as diarrhoea, cholestasis and various wound and respiratory infections, there remains a recognised risk of contamination from C. botulinum spores.^1112^ However despite this, the literature reports a recent case involving a 1-month-old infant who was treated in a hospital with medical-grade honey for an umbilical stump infection who subsequently experienced wound botulism.^13^

When the causative relationship between IB and honey ingestion was established in the 70s, the American Academy of Pediatrics published a general recommendation warning parents from introducing honey to their infants.^1^ Despite health practitioners' efforts to educate parents about the risks associated with honey, introducing honey to infants is a common practice in the Middle East and North Africa (MENA) region, including Oman. For example, Tahneek is a tradition where dates are carefully softened and rubbed against the newborn's palate shortly after birth to prevent hypoglycaemia. However, some communities choose to use honey as an alternative to dates.^14^

In Oman, there is a community belief that the long-standing tradition of giving ‘local’ honey to infants has not caused any harm that they have witnessed or heard about. This viewpoint may be supported by the lack of reported cases of IB, despite the frequent practice of feeding honey to infants. Concurrently, it is well-established that honey is an immunity booster, and an effective remedy for coughs and diarrhoea, even in infants when exposed to C. botulinum spore-free samples.^151617^ Meanwhile, the safety of cough and diarrhoea medications for young children has not been confirmed,^1819^ leaving families with home supportive care. As a result, parents may opt to introduce honey, believing it to be a natural and powerful anti-inflammatory supplement.

Due to the scarcity of reported cases of IB in Oman and the other countries in the region, the diagnosis merely relies on infant's clinical presentation and the treatment is primarily symptomatic. Thus, many cases may go undetected or inefficiently managed, necessitating uncovering essential data relevant to infant honey feeding practices and associated risk of IB in Oman.

Therefore, this study conducted a national surveillance to assess the risks associated with IB in Oman by examining honey-feeding practices among Omani infants and related familial sociodemographic factors. Additionally, the study investigated the likelihood of clinical manifestations associated with IB in infants who were fed honey, specifically constipation and muscle weakness.

2. Methods

This descriptive cross-sectional included Omani mothers with children under 5 years old who were literate in Arabic and consented to participate. The data collection process was conducted from October 2023 through March 2024. This study design type was used to enable the identification of common practices while exploring their relationship with various factors.^20^ Exclusion criteria included mothers with cognitive disabilities, non-Omani mothers and those who had previously completed the survey. Assuming that there were 7 predictors in the multivariable logistic model and using the 10-events-per-variable criterion, a minimum of 70 cases with identified IB risks was deemed necessary.^21^ Given plausible prevalence rates of 1–5%, a total sample size of 1,400–7,000 infants was calculated.^22^ This range ensured an adequate events-per-variable ratio for reliable regression estimates.

Study sample was conveniently recruited from a total of 120 health centres located in the 9 largest governorates in Oman. Health centres were the ideal setting for obtaining the required sample size as vaccinations are given at health centres at a rate of 99%.^23^ The authors trained 28 research assistants to collect data, where they screened participants and provided consent forms. Data were collected in-person using paper and pen.

This study utilised secondary data from the Infant Feeding Practices Study, a nationally representative survey conducted by this study's first author (unpublished). The survey gathered comprehensive information on the infant feeding practices and beliefs, beside general family sociodemographic and infant birth and clinical history characteristics. The item content validity index varied between 0.67 and 1, suggesting a good content validity. Details about the study instrument are published in another manuscript.^24^ Analysis for this study focused on descriptive data related to infant honey feeding practices, familial sociodemographic characteristics and clinical history during infancy.

The infant honey feeding data were obtained based on 6 multiple-choice questions and 1 Likert scale item. The questions were; “1) How many times did your child eat bee honey before he/she was one year old?”, “2) Who introduced bee honey to your child before he/she was one year old?”, “3) How bee honey was introduced to your child before he/she was one year old?”, “4) What kind of bee honey did your child eat before he/she was one year old?”, “5) What is the reason(s) for feeding bee honey to your child before he/she was one year old?” and “6) Have you felt pressure from people around you to feed your child bee honey before he/she was one year old?”. The Likert scale item derived from the infant feeding beliefs section stated, "I believe there is no harm in my child consuming honey during the first year of life". The item was answered on a 4-point Likert scale from strongly disagree to strongly agree.

Familial characteristics used in this study were geographical area, mother's age, mother's education, father's education, family income and mother's recipient of infant feeding counselling.

Data from two questions under the birth and clinical history section were retracted for this study: (1) “Did the child suffer from constipation during his/her first year? (not having a bowel movement at least once a day)?” with the answers being no, rarely (every one or more months), sometimes (every 2–3week) and frequently (every week) and (2) “Did the child develop flaccidity or muscle weakness during his/her first year?” with answers on a no or yes basis.

Data were cleaned, checked and analysed using Statistical Package for Social Sciences (SPSS) software, Version 23 (IBM Corp., Armonk, New York, USA). Data cleaning included checking completeness, resolving inconsistencies and removing duplicates and outliers. Participants' characteristics, infant honey feeding practices, sociodemographic and clinical variables were described as frequencies and percentages. Studied variables for relationships were dichotomised before inferential analysis, except the geographical area. Chi-square tests was used to analyse associations between infant honey feeding and sociodemographic factors, in addition to clinical factors. Binary and multivariate logistic regression were conducted to examine the predictors of infant honey feeding. The significance level was set at P <0.05.

3. Results

A total of 3,564 respondents were identified. After conducting initial data cleaning, 60 respondents who did not provide an answer to the study's primary outcome (the number of times honey was fed during the first year of life) were excluded. Furthermore, 282 respondents were removed due to response inconsistencies; some participants reported not feeding honey in one item, but they acknowledged feeding honey in another. This thorough data cleaning process resulted in a final sample size of 3,222 respondents.

Among participating mothers, 58.8% were 26–35 years old. The majority of participants lived in Dakiliya (19.8%) and South Batinah (18.7%). Concerning the children, 51.9% were male and 31.9% were 12–17 months old at the time of data collection [Table 1].

A total of 24.0% of children were fed honey more than twice and 14.7% once to twice during their first year; the cumulative infant honey feeding rate was 38.7%. Among participating mothers, 23.4% reported being responsible for introducing honey to their infants. Honey was mostly offered by itself (14.7%) or mixed with food (10.5%). Several reasons were cited for feeding honey to infants including cough (15.2%), boosting immunity (11.6%), Tahneek (8.6%), colic (4.1%), diarrhoea (2.9%) and making food more palatable (2.5%). Among the mothers surveyed, 34.6% indicated using local Omani honey for feeding their infants honey. Furthermore, 16.7% of mothers felt pressured to introduce honey and 45.9% agreed or strongly agreed that feeding honey to their child before their first birthday was acceptable [Table 2].

In the logistic regression analysis, the analysed predictors included the mother's age, mother's education, father's education, infant feeding counselling, family income, constipation and muscle weakness. The model showed good fit following Hosmer and Lemeshow (χ^2^ = 3.43; P = 0.91) and was statistically significant (χ^2^ = 83.19; P <0.001), accounting for 4% of the variance in the outcome. A significant association was observed between infant honey feeding and the mother's age (χ^2^ = 4.65; P = 0.03). Older mothers (36–55 years) were more likely to feed honey to their infants, with a crude odds ratio (OR) of 1.20 (95% confidence interval [CI]: 1.02–1.41). However, this significance was lost after adjusting for the other variables examined (adjusted OR = 1.07, 95% CI: 0.90–1.27). Significant associations were found with the mother's education (χ^2^ = 34.87, crude OR = 0.65, 95% CI: 0.56–0.75; P <0.001) and the father's education (χ^2^ = 8.19, crude OR = 0.81, 95% CI: 0.70–0.94; P = 0.004), suggesting that parents with lower education levels (high school or less) were more likely to feed honey to their infants. The association for the mother's education remained significant in the multivariable model (adjusted OR = 0.77, 95% CI: 0.65–0.91) but not for the father's education (adjusted OR = 1.09, 95% CI: 0.92–1.29). Family income also showed a significant correlation (χ^2^ = 55.89, crude OR = 0.55, 95% CI: 0.47–0.65; P <0.001) as did infant feeding counselling (χ^2^ = 7.27, crude OR = 0.82, 95% CI: 0.71–0.95; P = 0.007), suggesting that infants of families with higher monthly income and mothers who received infant feeding counselling had lower odds for being fed honey. Both of these relationships persisted after adjustment. In addition, the analysis revealed significant associations between infant honey feeding and both constipation (χ^2^ = 11.16; P = 0.001) and muscle weakness (χ^2^ = 6.54; P = 0.011). Infants with a history of sometimes or frequent constipation were more likely to be fed honey during infancy (crude OR = 1.30, 95% CI: 1.11–1.51) and this association remained significant in the multivariable model (adjusted OR = 1.29, 95% CI: 1.10–1.51). Similarly, infants with a history of muscle weakness had higher odds of being fed honey during infancy (crude OR = 1.79, 95% CI: 1.14–2.80) even after controlling for other variables (adjusted crude OR = 1.84, 95% CI: 1.15–2.95) [Table 3].

4. Discussion

The current study aimed to examine the risk of IB in Oman by conducting a national survey on the practices of honey feeding to infants, alongside associated sociodemographic and clinical factors. Findings revealed a significant prevalence of honey feeding among infants (38.7%) corresponding with the high rate of belief that introducing honey to infants under 1 year is harmless. In a distinct study from Chile focused on mothers who practiced baby-led weaning with children under 24 months, approximately 7.7% reported offering honey within the first 2 years of life.^25^ However, the specific timing of honey introduction before the age of 1 was not elucidated.

The existing literature concerning infant honey feeding is notably sparse, particularly within the context of the MENA region.^26^ Comparative data from Saudi Arabia indicates that nearly 52% of mothers provide honey to their infants before the age of 12 months.^26^ In another cross-sectional study that evaluated the awareness of the public in Saudi Arabia regarding the risks of feeding honey to infants, 35.5% of the participants admitted administering honey to infants younger than 1 year for traditional and health purposes including cough and constipation.^27^ A study in Palestine indicated that 15.8% of participants provided honey to infants under 12 months, with many believing it aids digestion and offers nutritional benefits.^28^

The current study's findings enhance the understanding of the most common reasons for feeding honey to infants, which include cough relief, immune system support and Tahneek. Moreover, this study offers novel insights, revealing that the majority of honey used for infant feeding is locally sourced and is introduced to infants in a myriad of ways including plain, mixed with other foods, milk or herbs. These insights provide health practitioners with comprehensive information regarding honey feeding practices, facilitating the delivery of culturally-directed health education.

The consensus among recent research indicates that mothers predominantly serve as the primary caregivers responsible for administering honey to infants.^2526^ This assertion is further validated by this study, which also identified a significant proportion of mothers feeling external pressure to provide honey to their infants. Simultaneously, this study's findings demonstrate a notable correlation between the absence of infant feeding counselling and an increased prevalence of honey feeding practices. This underscores the critical need for continuous infant feeding guidance within primary care services. Such support is particularly essential for mothers who may feel overwhelmed by societal pressures to conform to certain feeding practices, including the introduction of honey in their infants' diets.

Moreover, infant honey feeding practices has been significantly influenced by geographic residence.^26^ This suggests that cultural backgrounds linked to various geographic areas are instrumental in shaping the tendencies surrounding these feeding practices. This is particularly evident in the country's interior community, as observed in this study, where local traditions impact the acceptance of honey feeding among infants.

Additional maternal factors that correlate with honey feeding to infants include advanced maternal age, lower educational attainment and reduced family income.^26^ This data underscores that socioeconomically disadvantaged mothers may be least likely to engage with adequate child health promotion initiatives, thereby necessitating focused interventions to address these disparities.

This study's results indicate a positive correlation between honey consumption in infants and the manifestation of significant symptoms associated with IB, such as constipation and muscle weakness. This observation suggests the potential existence of cases of IB in Oman, despite the absence of reported instances. This uncertainty is further underscored by the lack of documented causative relationships between honey feeding and IB within the regional literature. It is noteworthy that a singular culture-confirmed case of IB, attributed to C. botulinum type B, was diagnosed in a 3-month-old female infant in the UAE in 2008.^7^ However, this case was not associated with honey consumption.

The challenges associated with identifying IB primarily stem from the limited resources available for clinical diagnosis, which often necessitate stool toxin assays for laboratory confirmation. Based on the risks identified in this study, policy-makers should reconsider allocating resources for confirmatory diagnostic evaluations to facilitate better surveillance and response to potential cases of IB. Moreover, it is imperative for healthcare professionals to maintain a heightened vigilance and include IB in their differential diagnoses. Adhering to established and updated clinical guidelines is of paramount importance to ensure optimal infant care and an effective reporting system.

The results of this study should guide future research efforts, such as examining the probabilities for underdiagnosed cases, healthcare providers' preparedness to diagnose and treat IB or re-investigating the IB risks associated with honey ingestion from a broader regional perspective. A review of over 70 international surveys conducted between 1978 and 2023 revealed that only 4% of honey samples taken from apiaries and retail sources tested positive for C. botulinum spores.^3^ This suggests that the likelihood of honey containing these spores can vary based on honey types and sources. Future research should investigate the prevalence of C. botulinum spores in Omani honey. Additionally, future studies could analyse stool samples from healthy infants to check for the presence of C. botulinum and identify their subtypes, if any.

The study has several strengths, including a large and representative sample size which enhances the generalisability of the findings. It provides a comprehensive analysis of infant honey feeding practices, an area that is relatively underrepresented in the literature. In addition, recruitment was conducted across diverse geographic areas and demographic groups to minimise potential selection bias. The findings offer valuable insights into current primary care practices and have the potential to significantly impact and improve relevant practices. However, the study employed a cross-sectional design and relied on self-reported retrospective data, which may compromise the integrity of the findings. Another limitation of this study is the use of convenience sampling, which may have introduced selection bias. However, attempts were made to minimise this effect through a large sample size. The lack of documented causal relationships in the region undermines definitive conclusions regarding the connection between honey consumption and IB.

5. Conclusion

Administering honey to infants remains a contentious issue between community beliefs and healthcare professional advisories. A considerable segment of the community endorses traditional notions regarding the health benefits of honey, contrasting with the medical understanding that honey may harbour C. botulinum spores, posing a risk of paralysis in infants vulnerable to C. botulinum colonisation. Although the occurrence of the disease is not reported, or perhaps under-reported, this study showed that risks exist and warrant further research and urgent consideration.

Authors' Contribution

Basma Al Yazeedi: Project administration, Conceptualization, Methodology, Investigation, Formal analysis, Visualization, Validation, Writing – Original Draft, Review & Editing. Omar Al Omari: Methodology, Formal analysis, Validation, Writing –Review & Editing. Salma Al Yazeedi: Conceptualization, Investigation, Validation, Writing – Original Draft, Review & Editing.

Acknowledgement

We acknowledge the Directorate General for Health Services in Muscat, Oman, for facilitating this research, as well as our partner healthcare agencies and all individuals involved in data collection, including research assistants, health center staff, and participating mothers.

Ethics Statement

Ethical approval for this study was obtained from the Research and Ethics Committees at College of Nursing, Sultan Qaboos University (CON/IG/2022/9) and the Health Studies and Research Approval Committee at the Ministry of Health (MOH/CSR/23/26906). Participation in the primary study was voluntary. Only participants who provided informed consent were included in the analysis. The data collection form included an information page and consent statement, where all participants provided written consent. To ensure anonymity, no identifying information was collected. Data were stored on a password-protected computer and paper forms were secured in a locked cabinet within a locked office.

Conflict of Interest

The authors declare no conflicts of interest.

Funding

This work was supported by internal grant fund from Sultan Qaboos University in Oman (IG/CON/MCHH/23/04).

Data Availability

Data is available upon reasonable request from the corresponding author.