Epidemiology

At the time of the 1992 report, foodborne, infant and wound botulism were reported worldwide but foodborne botulism was by far the most common form reported in the UK. In contrast, currently in the UK, wound and infant botulism are more common forms than that associated with food.

There has been no significant change in the disease symptoms reported since 1992^25–27. It remains the case that foodborne botulism is such a rare disease in the UK that misdiagnosis is a possibility due to the association of symptoms, by General Practitioners, with other more common diseases^27–29. Misdiagnosis may result in delayed response to cases and outbreaks and may warrant increased awareness amongst the medical profession. The US Centers for Disease Control and Prevention (CDC) has issued clinical guidelines for the diagnosis and treatment of botulism²⁷.

Investigations into many outbreaks of botulism have focused on the determination of the toxin type, rather than on the identification of the organism and its phenotype, and this has led to the loss of valuable data for subsequent risk assessment.

In this chapter, cases and outbreaks of botulism are reviewed firstly with respect to their geography and then with respect to the type of organism and its associations and therefore some repetition of events is evident.

4.1 Botulism – the global picture

Botulism is regularly reported throughout the world and Table 2 summarises the cases of all types of botulism from selected countries where data are available. Foodborne botulism is responsible for most reports in many countries, but it is not universally the most prevalent form of botulism.  For example, in the USA, infant botulism is the most common form of botulism with 1862 reported cases between 2001 and 2017, vastly exceeding 326 foodborne cases. In the UK wound botulism causes significantly more cases than infant and foodborne botulism combined but this may partly reflect the nature of clinical investigations and their capability.

4.1.1 UK

The UK has an extremely low number of outbreaks of foodborne botulism with only 10 reported outbreaks involving 13 cases since 1992 (data to 2019⁸ are included in Table 3). There has been no increase in foodborne botulism in the UK in recent decades. There has been a noticeable change in the origin of foods associated with outbreaks; eight of the ten most recent outbreaks involved foods produced or illness acquired abroad. This highlights effectiveness of the safety controls applied by the UK food industry in relation to this organism. Food produced in the home caused the greatest number of outbreaks since 1992 and five of six outbreaks, where details of food storage were known, involved ambient storage. A single outbreak implicating chilled food was caused by temperature abuse. Where the cause or potential cause of an outbreak was identified the evidence indicates that established controls for C. botulinum, if applied correctly, would have prevented the incident.

4.1.2 Worldwide

A selection of outbreaks of foodborne botulism that have occurred globally since 1992 are detailed in Table 4³⁰. There has been no significant change in the nature of foodborne botulism in recent decades with the exception of the identification of rare cases caused by neurotoxigenic C. butyricum and C. baratii. A few outbreak strains originally thought to be proteolytic C. botulinum are now recognised as C. sporogenes (details are included later in this chapter). The proportion of botulism outbreaks implicating commercial foods is weighted towards chilled foods rather than ambient foods (details are included later in this chapter and in Table 4) and this may constitute a trend in relation to the food types causing outbreaks. However, it is possible that bias may have been introduced by supplementing the literature sources with review papers specifically on chilled foods. 42 of the 90 outbreaks reported involved non-commercially produced foods i.e. home produced, 40 involved commercially produced food and 8 were unclear. Temperature abuse was a contributory factor in 30 of 36 outbreaks where a cause was known or suspected. For commercially produced chilled foods (and one frozen food) where the cause was known or suspected 23 of 24 products were subject to some form of temperature abuse prior to the outbreak; the remaining product was consumed 3 days beyond the “Use By” date. Outbreaks implicating commercially produced products that were destined for ambient storage and where a cause was known or suspected occurred due to a variety of control failures including inadequate sterilisation, post-process contamination, inadequate formulation to prevent growth and toxin production and/or temperature abuse. In the case of both chilled and ambient storage of commercially produced foods, the established controls for C. botulinum, if appropriately applied, would have prevented these outbreaks. Novel food technologies are not a feature of reported botulism outbreaks which tend to implicate traditional technologies which have been widely employed for food production over many years e.g. canning.

4.1.2.1 France

France reported 402 outbreaks of human botulism between 1987 – 2016 consisting of 731 cases and 9 deaths³¹. Cooked ham from home-made preparation or from small scale producers was involved in 73.5% of the botulism outbreaks where the food source was identified. These outbreaks were mostly type B botulism; more specifically B4 (and non-proteolytic C. botulinum). The other sources of botulism were home-made canned vegetables or fruits (beans, asparagus, eggplant, spinach, pumpkin, chestnut), home-made meat or fish preparations and a small number of industrial foods (fish soup, chicken/beef sausage, chicken/enchiladas, ground meat, olives/dried tomatoes, fresh pasta carbonara). Most of the outbreaks with non-pork meat were type A botulism (and proteolytic C. botulinum). Two outbreaks of rare C. baratii neurotoxin type F (F7) botulism were observed in 2014 and 2015. Industrial ground meat prepared in a restaurant was identified as the source of one outbreak³². Previous cases of botulism attributed to C. baratii were generally associated with intestinal colonisation i.e. infant botulism or due to other predisposing conditions³³. The origin of the other C. baratii type F outbreak was not identified. Clostridium butyricum was implicated in a case of botulism in a 10-year-old boy with history of Meckel’s diverticulum and chronic constipation, presenting dysarthria, dry mouth, hypotony, respiratory failure, and cardiac arrest in 2011. Stool analyses were positive for C. butyricum neurotoxin E5 by PCR and DNA sequencing up to 2 months after discharge. Botulism by intestinal colonisation with neurotoxigenic C. butyricum from undetermined origin was strongly suspected³⁴.

4.1.2.2 Italy

Italy has one of the highest numbers of foodborne botulism cases in Europe with 1173 suspected foodborne cases between 1986 – 2015, of which 421 were laboratory-confirmed²⁵. Homemade canned foods were implicated in 80.5% (95/118 incidents, involving 143 persons) of confirmed outbreaks including an outbreak caused by restaurant canned green olives. Vegetables canned in oil and in brine/water were associated with 43.2% and 28.8%, respectively, of laboratory-confirmed outbreaks. Other types of food implicated in confirmed outbreaks were home-bottled tuna (7.6%), ham (5.9%), home-bottled meat (5.9%), salami/sausages (4.2%), cheese (2.5%) and tofu and seitan (1.7%). Among vegetables, the most frequent products involved in cases or outbreaks were mushrooms in oil, olives and turnip tops. Regarding fish products, home-canned tuna was the most common food linked to confirmed incidents. Cheese or dairy products were seldom associated with confirmed incidents despite a large outbreak caused by mascarpone cheese³⁵. Although not reported for foodborne botulism alone, 96% (316/330) of the laboratory confirmed incidents were due to toxin produced by proteolytic C. botulinum (neurotoxins Type B 79.1%, Type A 9.7%, Type F 0.3%, Type Ab 1.5% and Type Bf 0.6%). Of 36 cases of infant botulism in Italy from 1986 to 2015, C. butyricum neurotoxin type E was implicated in three cases. Botulism involving intestinal colonisation with C. butyricum neurotoxin type E was also found in two boys having a Meckel’s diverticulum³⁶.

4.1.2.3 USA

There were 326 laboratory confirmed foodborne botulism cases in the USA between 2001 and 2017 of which 277 implicated a food or beverage (a food or beverage was a laboratory confirmed source of botulinum neurotoxin in 156 cases)³⁷. 47% of cases implicated food prepared in the home excluding canned foods, 29% home canned foods, 10% commercially canned foods and 6% other commercially prepared foods (no food preparation method was available for 8% of cases). Neurotoxin types A, E and B were responsible for most of the cases (65%, 25% and 7%, respectively). Outbreaks implicating commercial foods included a chilli meal, chilli sauce and nacho cheese sauce. A case of botulism implicating spaghetti with sauce/meat was attributed to C. baratii type F³⁸.

4.1.2.4 Turkey

A systematic review of botulism cases in Turkey from 1983 to 2017 identified 91 foodborne cases³⁹. Not all cases or suspected foods were tested for botulinum neurotoxin but 10 of 19 tested were positive for type A toxin (proteolytic C. botulinum). The top-ranking food responsible for cases was canned green beans (30% and 28 reports). Other reported foods include strained yoghurt (x10), home-made local cheese (x24), canned purslane (x16), non-specific canned food, canned ferula communis or fennel (x4), canned peppers, scrambled eggs with garlic sausages, canned fried mushrooms (x5) and unknown.

4.1.2.5 Iran

A review of botulism cases in Iran in the period 2007 – 2017 identified 252 confirmed cases of foodborne botulism, 743 suspected foodborne cases and 48 fatalities⁴⁰. The most commonly implicated foods, accounting for 34.1% of events, were home-prepared traditional processed fish (smoked fish, salted fish, ham, bacon, blood pudding, mosaic salami and sausage). Other implicated foods were commercially canned fish (28.6%), fish spawn (10.5%), dairy products (10.1%), vegetables and home-prepared legumes (9.7%), cottage cheese (5.9%) and canned fruits (1.1%).

4.1.2.6 Canada

There were 91 laboratory-confirmed outbreaks of foodborne botulism in Canada between 1985 and 2005 (205 cases and 11 deaths). Seventy-five outbreaks were associated with non-proteolytic C. botulinum type E; seven outbreaks associated with type A neurotoxin and five outbreaks with type B neurotoxin. The non-proteolytic C. botulinum neurotoxin type E outbreaks were attributed to consumption of traditionally prepared marine mammal and fish products by native communities (principally the Inuit of Nunavik in northern Quebec and the First Nations population of the Pacific coast of British Columbia). Two botulism outbreaks were attributed to commercial ready-to-eat meat products (pâté and cooked boneless pork) and three outbreaks to foods served in restaurants (chopped garlic in oil, bottled chanterelle mushrooms and baked potato). All involved type A or type B toxin. A further eight outbreaks were attributed to non-native home-prepared foods⁴¹. A review of foodborne outbreaks in British Columbia in the period 2009 – 2013 identified 3 botulism outbreaks and 1 death⁴². Implicated foods include fruit and vegetables, seafood and sauces/condiments.

4.1.2.7 Poland

Reviews identified that 54 of 109 botulism cases in Poland between 2014 and 2017 were due to home-produced foods and 55 due to commercial foods. Canned foods accounted for the vast majority of cases (84) with canned meat (other than pork) being the most frequently implicated food followed by canned fish, sausages and cured meat, canned pork, canned meat and vegetables and canned mushrooms, fruits and vegetables^43–46.

4.1.2.8 China

A review of botulism in China between 2004 and 2020 identified a total of 80 foodborne outbreaks with 386 illnesses and 55 deaths²⁹. The most common foods implicated were home-prepared traditional processed stinky tofu and dried beef, accounting for 51.2% of outbreaks. Contributory factors causing the outbreaks included improper processing and improper storage (77.5% of outbreaks). Initial misdiagnosis of illness occurred in 27.5% of cases. In an overview of type E botulism⁴⁷, 11 outbreaks between 1965 and 2005 implicated soy bean milk, fermented bean curd, raw dried beef, dried mackerel and blood sausage. The bacteria involved were either non-proteolytic C. botulinum type E or C. butyricum type E.

4.2 Association of clostridia and non-clostridia with foodborne botulism

4.2.1 Proteolytic C. botulinum and C. sporogenes

The vast majority of botulism cases associated with foods are caused by proteolytic C. botulinum neurotoxin types A and B. Rare cases implicating foods have been reported involving toxin type F.

It is widely recognised that proteolytic C. botulinum and C. sporogenes are closely related species. Recent genomic studies have highlighted that a number of strains of proteolytic C. botulinum should actually be classed as strains of C. sporogenes and vice versa, and that neurotoxigenic and non-neurotoxigenic strains of proteolytic C. botulinum and C. sporogenes exist⁸ (details of Taxonomy are included in Chapter 2 of this report). Four outbreaks of foodborne botulism have been attributed to C. sporogenes type B, including one case in the UK, along with rare cases of infant and wound botulism. As the historical identification of proteolytic C. botulinum associated with cases or outbreaks of botulism was based on the neurotoxin type this does not strongly affect the data presented in this chapter; neurotoxigenic C. sporogenes would have been incorrectly assigned to proteolytic C. botulinum. Moving forward it may be necessary to redefine historical data on botulism.  

Proteolytic C. botulinum and neurotoxigenic C. sporogenes are far more resistant to adverse conditions than other C. botulinum groups, requiring stronger heat processes to destroy spores and lower pH and water activity to prevent growth in food. Foodborne outbreaks associated with proteolytic C. botulinum have not shown any significant change in pattern, in recent decades, that would indicate an increased risk to foods. The controls that have been in place to manage the survival, population growth and toxin production by proteolytic C. botulinum for many years appear robust and do not need modification (details of the Occurrence, Growth and Survival of C. botulinum are included in Chapter 5 of this report).

4.2.2 Non-proteolytic C. botulinum

Non-proteolytic C. botulinum neurotoxin types B and E (and very occasionally toxin type F) are associated with foodborne outbreaks of botulism, but generally to a lesser extent than proteolytic C. botulinum neurotoxin types A and B, although this does vary by country.

Non-proteolytic C. botulinum is much less resistant to processing used in the food industry than proteolytic C. botulinum. The spores are more readily destroyed by heat and growth is more readily controlled by pH or water activity than for proteolytic C. botulinum. However, non-proteolytic C. botulinum is able to grow at much lower temperatures than proteolytic C. botulinum; this includes growth at refrigeration temperatures (details of the Occurrence, Growth and Survival of C. botulinum are included in Chapter 5 of this report). This presents a risk in chilled, minimally processed, extended shelf-life foods and this led to the provision of FSA guidelines for manufacture and sale of chilled foods in the UK² (details of the FSA guidelines are included in Chapter 1 of this report).

Despite the risk presented by non-proteolytic C. botulinum in chilled foods there have been no reported outbreaks of botulism involving chilled foods from any commercial product where the food has been stored at the recommended chilled temperature and consumed within its designated shelf life. In a review of global botulism outbreaks⁴⁸ caused by chilled foods, in the period 1985 to 2015, 16 of 26 outbreaks identified (Table 5) were caused by proteolytic C. botulinum, one by C. baratii and four by non-proteolytic C. botulinum (all type E in vacuum packed fish). Five outbreaks did not identify the organism and three of these involved toxin type B (it is unknown whether proteolytic C. botulinum or non-proteolytic C. botulinum were causal).  Temperature abuse was identified as the cause or most likely cause in 25 of the 26 outbreaks and consumption of the product beyond its “Use By” date was the reported cause of the remaining outbreak. In a review of botulism outbreaks in the USA between 1994 and 2021 that involved commercially produced foods intended to be stored chilled, 11 events were due to unrefrigerated storage by the consumer at home and the other two were due to unrefrigerated storage by the retailer prior to sale⁴⁹. In three of the events refrigeration instructions were deemed to be inadequate and contributory to the outbreak. In the 10 events where a toxin type was determined, all involved proteolytic C. botulinum (nine type A and one type Bf).

The lack of evidence for outbreaks caused by non-proteolytic C. botulinum in correctly stored chilled foods in the UK, both before and after the introduction of recommendations following the 1992 report, together with a similar lack of evidence from other countries throughout the world, provides some useful context regarding the magnitude of the risk and consequent need for controls in chilled foods. However, this review has not extensively examined the potential for under-ascertainment of botulism cases and outbreaks in other countries nor the legislative or industrial controls applied to foods globally. Currently it is not possible to use this insight in revising the risk from non-proteolytic C. botulinum in vacuum and modified atmosphere packaged chilled foods. Further study of these factors may provide more conclusive evidence that could indicate a lower risk to chilled foods from this organism that might, in turn, merit reduced controls or a focus on foods where the risk is greatest due to the known frequency of contamination and the associations with outbreaks i.e. vacuum packaged fish and seafood.

4.2.3 C. botulinum Group III

C. botulinum Group III (toxin types C and D) is often associated with animal botulism but has been implicated in human disease on rare occasions. In a recent review of all nine of the type C and D botulism outbreaks cited in literature, eight implicated foods; one was associated with infant botulism²⁶. Of the food outbreaks five had food vehicles suspected or confirmed as pâté (x2), smoked chicken, diseased chicken and home-made ham. The vast majority (seven) of these outbreaks were reported before 1992 when methods of identification and typing were less well advanced. Sporadic increases in animal botulism especially in farmed animals such as cows, cattle and chicken have raised concern regarding the potential for increased risk of transmission to humans. However, the ACMSF has reviewed such matters on a number of occasions and has considered the risk to humans to be low and supported the adequacy of current controls^50,51. There is no evidence that C. botulinum Group III presents any new or increased risk in relation to human foodborne botulism.

4.3 Other neurotoxin-producing clostridia

4.3.1 Clostridium butyricum

Neurotoxigenic C. butyricum type E has been reported in a small number of cases of infant botulism³³, adult intestinal botulism³⁶ and foodborne botulism in India⁵²,  China⁵³  and Italy⁵⁴. In the Indian incident clinical samples were not analysed and, despite the organism being isolated from a crisp made from gram flour, it was not possible to definitively confirm foodborne botulism. In a Chinese outbreak that occurred in 1994 implicating salted and fermented paste made of soybeans and wax gourds⁵⁵ six cases were clinically diagnosed with neurotoxin type E botulism that was also confirmed in the food although C. butyricum type E was only isolated from the food following further studies several years later⁵³. Retrospective analysis has also indicated an association of C. butyricum type E with other historical outbreaks of foodborne botulism in China^47,56. Type E botulism in China is most commonly associated with fermented grain/beans and raw meat and is frequently reported far from the sea (e.g. in Qinghai-Tibet plateau at an altitude of approximately 4-5 km). Canederli (bread dumplings) were the suspected food in a case of foodborne botulism in Italy in 1999 involving C. butyricum neurotoxin type E. Neurotoxigenic C. butyricum type E is a relatively newly identified hazard. No foodborne outbreaks of this type have been reported in the UK. Low numbers of confirmed outbreaks in other countries indicates that the risk presented by this organism has not substantially changed.

4.3.2 Clostridium baratii

Clostridium baratii producing botulinum neurotoxin type F has been associated with several foodborne outbreaks of illness although it was first isolated in an infant botulism case in the USA⁵⁷ and has subsequently been associated with a number of further cases of infant botulism^38,58 and adult intestinal botulism^59,60. Foodborne outbreaks have been reported in Spain, implicating individual meat pit pies⁶¹ and in France, where the food source was not identified although the only common item consumed by the two cases was an alcopop (which tested negative for growth and toxin⁶²) that reportedly contained 5% alcohol and had pH = 3.5. An outbreak in France implicated frozen and defrosted ground beef used for the production of spaghetti Bolognese at a restaurant where temperature abuse was again suspected as the underlying cause^32,63. Spaghetti and sauce mixture was implicated in a case of botulism caused by C. baratii type F in the USA in 2001³⁸. Raw deer meat was associated with an outbreak of foodborne botulism in Thailand in 2006, and a strain of C. baratii type F was isolated and its genome sequenced^64,65.

Neurotoxigenic C. baratii is a relatively newly identified hazard. No foodborne outbreaks of this type have been reported in the UK. Low numbers of confirmed outbreaks in other countries indicates that the risk presented by this organism has not substantially changed.

4.3.3 Clostridium argentinense

Clostridium argentinense was originally isolated from soil samples in Argentina in the 1960s and despite being capable of producing botulinum neurotoxin type G, there are no cases of botulism that have been associated with this bacterium⁶⁵. C. argentinense is not covered further in this report.

4.3.4 Other bacteria with potential botulinum neurotoxin genes

The advent of genomics has allowed the identification of potential botulinum neurotoxin genes in a number of bacteria outside of the Clostridium genus (more details are included in Chapter 2 of this report). The presence of botulinum neurotoxin genes is of potential concern but none of these bacteria have been shown to be capable of forming botulinum neurotoxin in foods and there are no known cases or outbreaks of botulism associated with bacteria outside the genus Clostridium.

4.4 Conclusions

4.4.1 Outbreak identification and clinical diagnosis

Foodborne botulism in the UK is extremely rare but consequently this may result in delayed responses due to unfamiliarity in the diagnosis of cases and delays in reporting.

4.4.2 Organisms responsible for botulism

There has been no significant change in the nature of foodborne botulism in recent decades with the exception of the identification of rare cases caused by neurotoxigenic C. butyricum, C. baratii and C. sporogenes. Other clostridia and non-clostridia have been identified with botulinum neurotoxin genes but have not been implicated in cases of foodborne botulism in the UK or elsewhere.

4.4.3 Foodborne botulism trends

There has been a noticeable change in the nature of foods that are associated with outbreaks of foodborne botulism in the UK. Eight of the ten outbreaks in the last 30 years involved foods produced or illness acquired abroad and the majority of these involved home-production.

Botulism outbreaks implicating commercially produced chilled foods appear more prevalent than in previous decades although bias in ascertainment of evidence may have affected this conclusion. The vast majority of botulism outbreaks, for both chilled or ambient stored foods, are identified with proteolytic C. botulinum, i.e. those organisms that do not grow under chilled conditions, and temperature abuse is the single most common cause identified for foodborne outbreaks. In relation to outbreaks in the last 30 years, in the UK and worldwide where a cause can be identified, there is evidence to believe that known controls for the organism, combined with the correct storage of the foods, if applied correctly, would have prevented the incident. Novel food technologies are not a feature of botulism outbreaks and most implicate traditional technologies, such as canning, employed for food production.

4.4.4 Vacuum packaged, extended shelf-life chilled foods

There have been no reported outbreaks of botulism globally in chilled foods from any commercial product where the food has been stored at the recommended chilled temperature and consumed within its designated shelf life. There has not been an extensive review of the potential for under-ascertainment of botulism cases and outbreaks in other countries nor the legislative or industrial controls applied to foods globally. However, further study of these factors may provide evidence that could indicate a lower risk to chilled foods that might, in turn, merit reduced or more focused controls. Nevertheless, the evidence indicates that where chilled foods are associated with outbreaks of botulism this is almost exclusively due to temperature abuse and in most cases is caused by proteolytic C. botulinum.

Table 2

Botulism cases in selected countries

Table 3

Summary of botulism cases recorded in the UK since the publication of the 1992 ACMSF Report on Vacuum Packaged Foods and underlying causes (Adapted from Brunt et al., 2020⁸).

^aProt: proteolytic C. botulinum, NP: non-proteolytic C. botulinum, Cspo: C. sporogenes, un, unknown.

^b not tested in the present study. 

Table 4

Reported outbreaks of botulism in foods occurring since the publication of the 1992 ACMSF Report on Vacuum Packaged Foods
(for UK, see Table 3).