We conducted a retrospective multicenter national cohort study in Israel. Included in this study were all infants 1 week to 1 year of age who had IB diagnosed during 2007–2021 by the National Reference Laboratory for Botulism’s Department of Bacteriology at Kimron Veterinary Institute (Bet Dagan, Israel). In addition, we verified with all pediatric infectious disease specialists in Israel that no other patients had IB diagnosed during the study period. Confirmed case-patients in the study were infants (<1 year of age) with clinical signs and symptoms suggesting IB and confirmed by MLB or mass spectrometric–based endopeptidase (EndoPep-MS) assay for detecting and differentiating BoTN serotypes.

We performed laboratory testing for IB by using standard MLB procedures under Kimron Veterinary Institute institutional approval for ethical use of laboratory animals. In parallel, we injected 0.5 mL of serum intraperitoneally into 2 mice and tested the presence of toxigenic C. botulinum bacteria in stool. We inspected mice for symptoms consistent with botulism. When both mice died within 96 hours of inoculation, or showed clear signs consistent with botulism in nonheated samples only, we conducted a neutralization test. We diluted culture supernatant (from stool samples) with A, B, or E antitoxins (provided by the US Centers for Disease Control and Prevention) and injected mice intraperitoneally with 0.5 mL of 1 of the following: native untreated supernatant; supernatant with A, B, or E antitoxin; or native supernatant heated at 80°C for 20 min. Finally, we confirmed botulinum by BoNT typing on the basis of survival and lack of symptoms in mice challenged with 1 of the neutralized samples and the heated samples, and deaths in mice challenged with the remaining native samples.

We used other diagnostic tools, such as an electromyographic test that showed characteristic findings for botulism (e.g., a repetitive stimulation leading to an incremental response), stool PCR, and EndoPep-MS assay (19,20). The EndoPep-MS assay was conducted by the Israeli Institute for Biologic Research (Ness Ziona, Israel), as described by Rosen et al. (21).

We retrieved clinical information about the patients from medical records in 6 hospitals. To evaluate the number of IB cases per live births in Israel, we used data from the Israel Central Bureau of Statistics on live births during the study period (https://www.cbs.gov.il/en/subjects/Pages/Live-Births.aspx). For each case, we collected the following information: demographic data (age, sex, place of residence, hospital where the patient was treated; clinical data (perinatal information, underlying conditions, and breast feeding vs. formula); information on environmental exposures (agricultural occupation of parents or living nearby construction areas), positive toxin test results or clostridial spore presence (as determined by culture or MLB) in honey, formula, soil, and dust from habitant samples; and clinical manifestations. Laboratory results included the toxin type found (A, B, or E), electromyographic features, and blood and cerebrospinal fluid test results. We also collected information on treatment protocols (e.g., use of antibiotics, supportive care only, and administration of antitoxin) and prognosis.

We identified 8 IB patients reported in Israel during 2007–2021; of those, 5 were boys and 3 girls. Three stool samples were positive for BoNT A– and 5 for BoNT B–producing C. botulinum. Seven patients were born at term and healthy with no underlying conditions. One patient was born at late preterm and had previous known hypotonia but was included in the study because of acute clinical deterioration, a positive BoNT B result, and improvement during sequential neurologic follow-ups.

The incidence of IB in Israel during 2007–2021 was 0.3 cases/100,000 live births (8 cases). A possible increase in IB incidence of IB occurred during 2019–2021 (0.9 cases/100,000 live births [5 cases]), compared with incidence during 1996–2018 (0.14 cases/100,000 live births [5 cases]) (Figure).

The median age for diagnosis in our study was 6.5 months (range 2.5–8 months). The median time to diagnosis (from symptom onset to suspected diagnosis) was 9.5 days (range 4–35 days). In 1 case, which was detected by the Endopep-MS (when MLB results were negative), time to diagnosis was 3 days from symptom onset.

The demographic and geographic data indicated that no geographic cluster occurred and that the cases were evenly distributed across Israel (Table 1). Similarly, ethnicity and socioeconomic characteristics of IB patients were diverse. Two of the case-patients resided in a rural area (1 in a temporary shed, 1 in a private house). The others resided in urban areas.

In this study, we describe cases of IB in Israel during 2007–2021. We found a possible increase in incidence in the last 3 years of the study period, especially in 2020, compared to the previous 2 decades. We also noted older ages of the infant patients at illness onset compared with those described previously (4). Honey consumption was reported in only 1 infant, whereas we observed an apparently stronger link to environmental factors as source of infection.

The age of diagnosis in our study (6.5 months) is much older than that found in other studies. Previous studies found that age at illness onset in 90% of IB cases was 1–6 months (4), with a median age of 91 days (22,23). A reasonable assumption is that around this age, the infant undergoes a sharp transition from breastfeeding-only or formula-fed only to exposure to different foods. All those changes can cause a sharp shift of the bowel microbiome and optimize conditions for colonization and proliferation of C. botulinum (24). In addition, around this age, the infant is exposed to other external factors, such as daycare environments, solid foods, and other environmental factors, that may expose them to C. botulinum.

Previous research has shown that the connection between honey consumption and IB in the United States is considerable; honey consumption accounts for ≈15%–20% of all IB cases, and up to 25% of honey products were found to be positive for C. botulinum spores (3,25,26). In Europe, 59% of the IB cases are related to exposure to honey before symptom onset (18). In our study, honey consumption was reported in only 1 case (a honey-cake that was later found to be C. botulinum contaminated). This modest link is notable, given that people in Israel are known to be avid consumers of both local honey and honey imported from Europe and the United States; further, honey consumption increases during some of the Jewish holidays, around September–October. In fact, we found no cases of IB during those months. Many case series show that IB cases linked to honey consumption were in patients who were positive for BoNT B (25). In our study, the infant who consumed honey was also positive for BoNT B. A possible explanation might be an early and mandatory parental education program in outpatient infant clinics in Israel that forbids honey products until the infant is 1 year of age.

We observed a possible link to certain environmental factors, including dust (vacuum cleaner content), having a father who works as a farmer, crowded living spaces, and living near construction sites, all already known as potential risk factors (8,10,20,25,27). Those cases might indicate that environmental factors are major risk factors and should be investigated thoroughly and carefully.

A possible seasonal relationship emerged from our findings. Most (7/8) of the cases occurred during spring and summer in the Northern Hemisphere (March–July). One case originated from the Negev Desert area, characterized by a year-round hot and dry climate. These findings suggest a possible influence of temperature on IB outbreaks. Supporting evidence of this distinct seasonal incidence was also shown in a study conducted in the US state of Utah, where all cases of IB occurred during March–October and no cases occurred during the winter months (12,22,27). Another possible explanation might be that those are the most convenient months for outdoor activities and traveling, and therefore more exposure to soil and other environmental factors occurs.

Our study demonstrates a possible rise in IB incidence throughout Israel. During 1996–2006 (a 10-year period), only 2 cases of IB were diagnosed (15,28–30), whereas in our study, 8 cases were diagnosed during 2007–2021 (a 15-year period). A possible explanation might be the rapidly increasing number of construction projects, higher population density, and increasing awareness of IB among pediatricians. Still, the national incidence of IB in Israel is relatively low (only ≈0.3 cases/100,000 live births, compared with 1.9–2.1 cases/100,000 live births in the United States [4,15] and 2.2 cases/100,000 live births in Argentina [31]) and is more similar to low-incidence countries such as England (32) or Germany (15).

We identified a notable cluster of 5 cases during 2019–2021 (0.9 cases/100,000 live births). This possible surge in incidence raises an interesting question about a possible link to the COVID-19 pandemic. Is the increase in IB incidence observed in Israel during 2020 also observed worldwide? No up-to-date data are available to answer this question. This possible surge in Israel might also be attributed to the plausibility of random variations in the occurrence of a rare disease.

Most of the cases in our study were caused by BoNT B (5/8) and the rest by BoNT A (3/8). A similar ratio exists in the United States (62% BoNT and 37% BoNT A), but the ratio is different in the rest of the world, where BoTN A is the more prevalent type (15).

The number of infants who required intubation and invasive ventilation in our study was notably lower (2/8) that reported in other case-series reviews, such as one conducted in California, USA (60%–82%) (22,33,35). The 2 infants who were intubated were both positive for BoNT B, which is related to a higher incidence of respiratory failure (33). The older age of patients at onset observed in our study led to a milder clinical course and lower intubation rate.

Half of the infants in our study were treated with antitoxin, whereas the rest showed clinical improvement or recovered without antitoxin treatment. Previous reports showed that antitoxin treatment shortened the length of hospitalization from 23 to 13 days (22,33–35). In our study, infants who were treated with antitoxin were hospitalized for 17 days, and those who were not treated were hospitalized for 13.5 days. A possible explanation might be that the infants who received antitoxins had a more severe clinical picture, which necessitated the use of antitoxin but also led to a more complex and prolonged hospitalization. Another possible explanation might be the fact that in Israel the equine-derived botulism antitoxin is still used, whereas in the United States, the common antitoxin administered in recent years is the human botulism immune globulin.

The main strength of our study is its national reach. Because only 1 national laboratory for diagnosis of IB exists, we were able to access all the diagnosed case-patients’ medical records. The main limitation of our study, which derives from the rarity of IB, is the small sample size of only 8 cases. Another limitation is the retrospective nature of our study, which means some details are unrecorded or unknown.

In summary, although IB is still a rare disease, IB incidence may have increased in Israel during our study period, 2007–2021, and especially during the past 3 years of that period. We also found that honey consumption is less prominent as a risk factor compared with environmental exposures. Clinicians should be aware of the possible changing epidemiology and risk factors of this rare but serious disease in Israel.