Article Text


Respiratory infections in the military
  1. Matthew K O'Shea1,2 and
  2. D Wilson2
  1. 1The University of Oxford, The Jenner Institute, Oxford, UK
  2. 2Department of Military Medicine, Royal Centre for Defence Medicine, Birmingham, UK
  1. Correspondence to Surg Lt Cdr M K O'Shea, The University of Oxford, The Jenner Institute, Old Road Campus Research Building, Roosevelt Drive, Oxford OX3 7DQ, UK; matthew.oshea{at}


Military training facilities and operational theatres, and the stressful activities undertaken in such settings, are unique. Military personnel living and working in these environments are at considerable risk of the acquisition and onward transmission of a variety of respiratory infections. While these generally cause mild illness, severe disease may occur with significant associated morbidity and, occasionally, mortality. Epidemic outbreaks among military personnel may have a significant detrimental impact on training schedules and operational effectiveness. The recognition of the burden of such illness among British military personnel, and the development of strategies required to prevent or limit negative impacts, can only be achieved through the use of comprehensive laboratory-based surveillance programmes.

  • Respiratory
  • Infections
  • Military

Statistics from

Key messages

  • Military personnel undergoing initial recruit training or deployed to an operational setting are at risk of the acquisition and onward transmission of respiratory infections.

  • Whilst typically causing mild illness, severe disease may occur with significant associated morbidity and occasionally mortality.

  • Epidemic outbreaks among the military are well recognised and may impact detrimentally on training schedules and operational effectiveness.

  • Comprehensive laboratory-based surveillance programmes are required to guide decisions concerning appropriate interventions such as vaccine strategies and the role of prophylactic antibiotics.


The role of respiratory infections as a cause of morbidity and mortality among military personnel in training and operational environments is well recognised. When compared with an equivalent population of young healthy adults, military recruits undergoing initial training appear to be at a greater risk of acquiring and transmitting respiratory infections (Figure 1).1 ,2 The increased vulnerability to acute respiratory disease (ARD) and febrile respiratory illness (FRI) observed in military recruits has historically been attributed to demanding physical training schedules and crowded habitation.3–5 Furthermore, personnel newly deployed to areas highly endemic for certain pathogens to which they may be exposed for the first time are also at risk of developing respiratory infections, especially in the context of stressors such as poor nutrition, high person density and resource-limited settings.6 Numerous infective agents including adenoviruses, influenza A and B viruses, Streptococcus pneumoniae, Streptococcus pyogenes, Chlamydia pneumoniae, Mycoplasma pneumoniae, Epstein–Barr virus, coronavirus and rhinoviruses have previously been identified as causes of respiratory disease among military populations.1 This review begins with a general discussion of the epidemiology, microbiology, diagnosis and management of lower respiratory tract infections acquired in the community, followed by a discussion of the three of the most significant and currently relevant causes of respiratory infections in military populations: S pneumoniae, adenoviruses and Bordetella pertussis.

Figure 1

Rates of hospitalisation for acute respiratory disease among US Army recruits and civilian adults, 1991–1994.1

Community-acquired pneumonia

Community-acquired pneumonia (CAP), defined as an acute illness characterised by signs and symptoms of a lower respiratory infection accompanied by new radiographic changes which is acquired outside hospital or a long-term nursing facility, is a common cause of morbidity and mortality accounting for 1% of all medical admissions in the UK.7–11 Between 22% and 42% of adults with CAP require hospital admission among whom associated mortality is 5%–14%.9 While the aetiological agents responsible for community CAP are diverse, specific pathogens are confirmed in fewer than 30% of cases, which may reflect false negatives tests for known organisms and/or the contribution of novel pathogens.12 Among established diagnoses, bacterial infection is the most common cause of CAP and a significant proportion of cases result from S pneumoniae, Haemophilus influenzae and Moraxella catarrhalis (Table 1). S pneumoniae is by far the most commonly isolated infectious agent. Influenza A and rhinoviruses are the commonest viral causes of CAP, accounting for up to 30% of cases, and may precede super-added bacterial infection.13 The relative frequency of particular pathogens is affected by age, environment and co-morbidities. The geographical setting and the environment are particularly pertinent to physicians caring for military personnel who may present during or following deployment on operations or training exercises overseas. In addition, the epidemiological patterns characteristic of certain pathogens may be of use to the clinician (Table 1).

Table 1

The microbiology, epidemiology and investigation of community-acquired pneumonia (CAP)7 ,9

Symptoms suggestive of pneumonia include fever (≥38°C), cough, sputum production, dyspnoea, rigors and pleuritic chest pain and certain clinical features may be indicative of specific causative pathogens (Table 1). Clinical examination reveals a fever, tachypnoea and rhonchi in over 80% of patients, with signs of lobar consolidation in up to 30%.14 The differential diagnosis for CAP is broad and alternative diagnoses should be considered in those patients who fail to respond to appropriate treatment for CAP.11

Hospital admission and investigation are indicated for patients who are sufficiently unwell or in cases where the diagnosis is uncertain, response to treatment is poor or there is recurrence. Investigations include chest radiography, arterial blood gas analysis, full blood count, electrolyte concentrations, renal and liver function tests, inflammatory markers and HIV testing.7–11 ,15 Microbiological investigations include microscopy, culture and sensitivity of organisms from clinical specimens (blood, sputum and pleural fluid) together with pathogen-specific tests (Table 1).9

The CURB-65 score is used to assess disease severity and directs the most favourable setting for the safe treatment of patients and also optimises empirical antibiotic therapy. Patients are stratified into one of three groups of disease severity in relation to 30-day mortality on the basis of the presence of confusion, uraemia, tachypnoea, hypotension and age ≥65 years (Tables 2 and 3). Additional clinical markers of disease severity (eg, the extent of radiographic changes, acidosis and hypoxia) and the use of certain biomarkers (eg, C-reactive protein (CRP) and pro-calcitonin) may improve risk stratification and prognostic scores.9 A high CRP (>100 mg/L) suggests bacterial infection and is independently associated with the development of complicated parapneumonic effusions.

Table 2

Assessment of severity of community-acquired pneumonia using the CURB-65 score9

Table 3

Severity scoring, investigation and management of community-acquired pneumonia9

The management of patients with CAP include general supportive care (hydration, appropriate oxygen supplementation, nutrition, venous thromboembolism prophylaxis, early mobilisation and cessation of smoking) and antimicrobial therapy. Empirical therapy should be commenced within 4 h of presentation and continued for between 7 and 10 days in low-moderate and high severity CAP, respectively (Table 3). Positive microbiological investigations allow for pathogen-specific antibiotics to be used. The failure of a patient to respond to therapy raises the possibility of the presence of resistant organisms, the development of complications, inadequate treatment due to altered pharmacokinetics or an incorrect diagnosis. Complications of CAP include sepsis and acute respiratory distress syndrome, parapneumonic effusion and empyema, lung abscess formation and metastatic infection (requiring prolonged courses of antibiotics and occasionally surgical drainage).11

Streptococcus pneumoniae

S pneumoniae is the most common cause of CAP, accounting for up to 75% of cases in the UK.7–10 Outbreaks of pneumococcal pneumonia may occur in crowded institutional settings, such as nursing homes, day care centres, shelters for the homeless and prisons.16–19

Historically, large epidemics of pneumonia occurred in the military in the pre-antibiotic era, particularly among crowded populations during the winter and following outbreaks of influenza. S pneumonia was associated with significant morbidity and mortality and rates of death of up to 50% have been reported.20 With the development of effective antibiotics, the frequency of such epidemics declined sharply although today S pneumoniae is still the leading cause of pneumonia hospitalisations among the military in the USA.1 ,21–24 Furthermore, outbreaks of pneumococcal pneumonia among military personnel have been increasingly recognised in recent years and S pneumoniae has been identified as the primary pathogen in several outbreaks among military trainees and deployed units.20 ,24–26 The increasing frequency of such epidemics may reflect an epidemiological shift and re-emergence of the threat of pneumococcal disease to military populations.

Nasopharyngeal colonisation with S pneumoniae is generally asymptomatic. However, colonisation is a fundamental precursor of pneumococcal disease and also serves as an important source for horizontal spread within the community, which increases in circumstances facilitating transmission such as crowding.17 ,27–29 The rate of pneumococcal nasal colonisation among healthy military recruits ranges between 1% and 7%.21 ,30 ,31 This appears to increase significantly during military training and carriage rates of up to 44% have been noted during pneumococcal outbreaks in military environments.21 ,25 ,30 ,32 Studies of the dynamics of pneumococcal transmission among healthy young adults, not in the context of an outbreak investigation, are scarce.33 A recent study was undertaken to characterise pneumococcal carriage prevalence, acquisition and dynamics among healthy military recruits before and during training following an outbreak of severe pneumococcal disease at an Israeli army training facility.32 ,34 It showed that shortly after mixing of recruits in a confined setting the prevalence of pneumococcal carriage increased significantly, with the acquisition of S pneumoniae occurring in 37% of trainees and the highest risk of acquisition during the first 3 weeks of training.34 Furthermore, higher carriage and acquisition of pneumococci were noted during the winter months and such enhanced bacterial transmission may be attributed to increased crowding, reduced ventilation and concomitant upper respiratory tract viral infections.

There have been very few published data concerning pneumococcal disease among British military personnel and none pertaining to training populations. A small outbreak of CAP among British soldiers deployed to the Gulf between November 1990 and January 1991 has been reported. Radiographic evidence of lobar consolidation was apparent in 78% of individuals and pneumococcus isolated from 11%.35 One review of infectious disease referrals to the Role 4 facility in Birmingham, UK, between 2005 and 2009, revealed that ‘respiratory illness’ accounted for 11% of referrals and a final diagnosis of lower respiratory tract infection occurred in 5% of cases, although causative organisms were unknown.36

An outbreak of pneumonia associated with S pneumoniae among Finnish recruits during a training exercise in 2006 resulted in a hospitalisation rate of 12%, with pneumococcal carriage in 42% of the recruits screened and the outbreak strain isolated in over half of them.25

A recent outbreak of radiologically-confirmed pneumonia was recognised among a training company following two fatal cases of pneumococcal meningitis in US army recruits. A subsequent review of pneumonia and FRI surveillance data at the training facility revealed that 15% of trainees in the battalion were colonised with two predominant serotypes of pneumococcus, suggesting its role in the aetiology of the outbreak. The 23-valent pneumococcal polysaccharide vaccine (PPV23) and adjunctive antibiotic chemoprophylaxis (benzathine penicillin G) were used in an attempt to halt the outbreak. Such a dual approach may provide additional protection until immunity to the vaccine has developed after approximately 2 weeks. Furthermore, antibiotic prophylaxis also reduces nasopharyngeal colonisation with pneumococcus and decreases pneumococcal transmission. Such a regime has previously been employed in other clusters of severe pneumococcal infection including in military settings.21 ,24 ,25 However, this is a controversial approach for numerous reasons and difficulties with the use of prophylactic antibiotics include potential adverse reactions, cost and the promotion of drug resistance.34 The US Military Vaccine Agency does not recommend vaccination with PPV23 for all military trainees but does offer the vaccine to those considered at a high risk of pneumococcal infection.

In the UK, rates of S pneumoniae penicillin non-susceptibility remain low; 94% of bacteraemia isolates and 92% of respiratory isolates remain fully susceptible to penicillin and 85% and 88% of blood and respiratory isolates (respectively) are susceptible to macrolides.9 ,37–39

In contrast, however, the worldwide prevalence of penicillin-resistant S pneumoniae is increasing, occurring in up to 25% of pneumococcal isolates in some areas in the USA and as high as 70% in other countries although patterns of disease are changing with the introduction of conjugate vaccines.40 ,41

The international variability in rates of antibiotic resistance among S pneumoniae isolates should be considered when the physician deployed overseas is treating local patients and/or military personnel from multinational forces, perhaps originating from countries with suboptimal antimicrobial stewardship.


Adenoviruses are non-enveloped DNA viruses and have a long association with the military. They appear to cause three distinct types of pathogenesis in human cells which include lytic infection, latent (chronic) infection and oncogenic transformation.42 ,43 The spectrum of clinical presentations associated with adenovirus infection ranges from pneumonia and bronchitis/bronchiolitis to conjunctivitis, pharyngoconjunctival fever and upper respiratory tract infections, occasionally occurring as epidemic outbreaks.42 ,43

Adenoviruses were first isolated in 1954 from the adenoidal tissue of US soldiers presenting with ARD and were subsequently isolated during an outbreak of ARD at a military recruit training facility in the Netherlands in 1955 and in several other countries thereafter, including the UK.43–48 The significance of adenovirus infection during such outbreaks of ARD and FRI among military populations became increasingly recognised throughout the 1950s and 1960s. Adenoviruses were found to be responsible for 60%–80% of cases of ARD among recruit trainees, accounting for up to 90% of all recruits requiring hospital admission with pneumonia during the winter.49–54 Adenovirus serotypes 4 and 7 were found to be particularly associated with military epidemics and recruits were at significantly higher risk of acquiring infection than comparable civilian adults (prevalence of 2%), especially in the first 3 weeks of training.50 ,51 ,55–57 In the early 1960s, 600–800 hospitalisations per week due to ARD occurred at military training facilities in the USA during the winter, reducing the healthy population by 40%–50%, disrupting military training programmes, potentially impacting operational commitments and causing significant economic loss.49–55 The need for an effective vaccine was established.

In 1971, the US Department of Defense introduced the routine use of a bivalent oral live-attenuated, enteric-coated vaccine against adenovirus serotypes 4 and 7 for all recruits entering basic training facilities.49 ,52–54 ,58 ,59 The effects on adenovirus-associated ARD and FRI were dramatic. Reductions of disease incidence of between 36% and 73% were observed in new recruits and on average a 50% reduction in adenoviral-related hospitalisations.52 ,53 While between 1967 and 1974 five adenovirus-associated deaths were reported in military personnel, there were none reported between 1975 and 1998.52 ,60–63 It was estimated that the vaccine had resulted in a cost saving to the US Army of approximately $5 million a year in 1973.64 In 1984, in response to recurring ARD epidemics, the US Army initiated year-round vaccinations against adenoviruses and between 1984 and 1994 the rate of ARD among trainees was the lowest ever previously recorded, with no documented outbreaks associated with adenovirus types 4 or 7.53

Unfortunately, the sole manufacturer of the vaccines ceased production in 1996 and the US military introduced a programme of phased cessation of remaining stocks until eventual termination of all adenovirus vaccinations in 1999.65–67 This resulted in a resurgence of adenovirus infections to pre-vaccine era levels and the re-emergence of a significant threat to military personnel, especially new recruits undergoing initial training (Figure 2).54 ,65–69 Numerous outbreaks of FRI and ARD associated with adenovirus infections have occurred, including several fatalities. Between 1998 and 2010, following complete cessation of the adenovirus vaccine programme, the US military's Mortality Surveillance Division recorded eight deaths in serving personnel attributable to adenoviral respiratory disease.70–72 Serotypes 4 and 7 were predominant among military recruits and the emergence of two new serotypes, 14 and 21, was also observed.70 ,73–77

Figure 2

Febrile respiratory illness and adenovirus infection rates among US Army trainees at Fort Jackson, Columbia, 1994–1999.66

The case to restore adenovirus vaccine capability was made.78 In 2001, the US Army Medical Research Acquisition Agency commissioned a pharmaceutical company to facilitate this which has resulted in the recent development of a new adenovirus live-attenuated oral vaccine against serotypes 4 and 7. In March 2011, the US Food and Drug Administration approved the new vaccine which was introduced to new trainees at recruit training centres in October 2011.79 ,80 A recent assessment of the initial impact of the vaccine on FRI and adenovirus transmission in military recruits undergoing basic training suggests that the vaccine is highly effective in preventing FRI due to adenovirus type 4, demonstrating a 75% reduction in FRI to a rate of 0.15 cases/100 trainees/week, and a substantial reduction in the proportion of respiratory swabs positive for adenovirus type 4 (from 71% to 0%), consistent with the findings of earlier clinical trials.81

While the combination of sustained transmission and relatively high and predictable attack rates of adenovirus-associated disease appears to be unique to US military basic trainees, those from several other countries have been shown to experience such infections.81

Studies of Canadian recruits and those from several European countries in the 1950s and 1960s showed adenovirus infection accounted for 28%–70% of all ARD among recruits and 19%–62% of ARD-associated hospitalisations. Adenovirus serotypes 3, 4, 7 and 21 predominated.82–87 Evidence from the British military (Royal Air Force and Royal Navy recruits) from a similar period has been conflicting with an incidence of 7%–36%.88–91 More recently, there have been reports of outbreaks of adenovirus-associated ARD in a number of European military training facilities.92–100 In the UK, adenoviruses have been the most common respiratory viruses, isolated from up to 35% of symptomatic new naval recruits undergoing basic training.101

Adenovirus-associated ARD and FRI accounts for a significant proportion of disease burden among military recruits and while the clinical manifestations are typically self-limiting and non-severe, complications may arise. Moreover, sporadic epidemics may result in high rates of absenteeism, disrupting training schedules or interfering with the effectiveness of a unit and considerable financial cost. Further epidemiological studies are required to accurately determine the current burden from adenovirus infections and inform future policy.

Bordetella pertussis

B pertussis is a Gram-negative coccobacillus and the cause of ‘whooping cough’. B pertussis infection was once considered a disease restricted to childhood and resulted in significant morbidity and mortality among infants and young children. The introduction of a whole-cell pertussis vaccine into the universal childhood immunisation schedule in the 1950s was associated with a considerable reduction in the incidence of the disease in this group.102 However, neither natural infection nor vaccination provides lifelong protection and B pertussis continues to circulate in the population. Indeed, over the past 30 years, the number of reported cases of B pertussis has continued to increase in developed countries, predominantly among adolescents and adults103–106 (Figure 3). There are likely multiple reasons for this rise including increased recognition of the disease with the advent of improved diagnostics (eg, serological testing) and limited duration of immunity which is thought to wane 4–12 years after vaccination or 4–20 years after natural infection (ie, in early adulthood).107 ,108

Figure 3

Laboratory-confirmed cases of pertussis infection, England and Wales, 1994–2012, Q1–2 (confirmed by culture, PCR and/or serology) (adapted from Public Health England data).

The clinical manifestations of B pertussis infection vary from mild respiratory symptoms to the classic symptoms of overt disease.109 However, the symptoms of pertussis infection in previously immunised or infected adolescents and adults are variable and often atypical.110–112 While B pertussis infection in adults generally results in mild symptoms, which are often attributed to a ‘flu-type’ illness, likely contributing to under-reporting, the predominant symptom may be nothing more than a prolonged cough. Studies have shown that among this group, 12%–32% of individuals with an illness characterised by a cough of 6 days duration or longer have serological evidence of infection with B pertussis.113–116 Pertussis should therefore be considered in the differential diagnosis of any cough illness lasting more than 1–2 weeks. Complications of infection with B pertussis in adults may include pneumonia, seizures and encephalopathy and, rarely, intervertebral disc herniation, angina and carotid artery dissection have been reported.112 ,117–121

If untreated, B pertussis infection is highly contagious during the first few weeks of symptom onset and a single primary case may cause up to 17 secondary cases in an immune-naive population.122 ,123 Adolescents and adults with unrecognised infection may serve as a reservoir for B pertussis.

Many of these factors are relevant to the epidemiology of B pertussis infection among military populations. Outbreaks of pertussis occur on a cyclical basis, every 3–4 years, and both localised and widespread outbreaks among civilian adults have also been described previously.113 ,114 ,116 ,118 ,124–126 Furthermore, while its role in the epidemiology and the transmission of the disease is unclear, transient carriage of B pertussis has been observed in children and adults. Outbreaks have been reported in US marine recruits and soldiers on exercise, and Israeli infantry troops and recruits, with an incidence of between 6% and 23%.127–130 Sustained transmission of B pertussis may be enhanced by the unique demographics of basic trainees including the closed nature of the community and a high turnover rate providing a frequent influx of susceptible individuals.127

Such infections may also occur in deployed personnel who are generally older and may possess different susceptibilities due to waning vaccine-induced immunity.108

A recent review of pertussis diagnoses among service members and other beneficiaries of the US Military Health System between 2007 and 2012 reported an incidence of 16.2% and 2.7% of confirmed B pertussis infection among active and reserve service personnel (respectively) over the 5-year surveillance period.131

The possible role of pertussis as a cause of respiratory disease in the deployed setting has been a particular concern during recent operations in Afghanistan.

Since 2001, there have been anecdotal descriptions of chronic, non-productive, cough occurring among military personnel following multinational deployments to Afghanistan, termed ‘Kabul Cough’. This was initially attributed to environmental pollution but air sampling studies have failed to support this assertion.132 As a result, the role of respiratory tract infections as an aetiological factor has been considered and B pertussis was highlighted following an outbreak among immunised children in British Forces Germany in 2006 who had been in contact with family members recently returned from deployment in Afghanistan.133 Furthermore, a severe case of pertussis infection arose in theatre in late 2006 and two further cases in British service personnel were confirmed serologically at a French military facility in Kabul in early 2007.132 Indeed, a significant increase in ARD, monitored through French Army and NATO surveillance systems, was noted among the multinational force in Kabul with an incidence of pertussis infection of 20% by clinical definition and 22% following serological testing.134 A further study to investigate the prevalence of pertussis infection in a group of symptomatic British soldiers returning from Afghanistan showed that among the 21 personnel who met the case definition, two confirmed cases and one probable case of B pertussis infection were identified, yielding a probable infection rate of 14%.132 The authors noted while the cases occurred among headquarters staff who had travelled throughout Afghanistan during their deployment, different exposures and immunological profiles may occur among British troops in the field.

A surveillance study to determine pathogen-specific seroprevalence prior to and after deployment among US military personnel deployed in support of operations in Afghanistan reported that pre-deployment IgG seropositivity for B pertussis was 14.2% and the rate of seroconversion was 4.3%.135 This is two to four times higher than that observed in comparable US civilian populations.136 ,137 Seroprevalence in the oldest age group was significantly lower than in the youngest age group, possibly indicating waning immunity. Similarly, seropositivity was higher among those joining the military within 2 years prior to deployment compared with individuals who joined more than 5 years earlier. Interestingly, the authors noted an increased risk of B pertussis seroconversion among individuals who deployed in 2007, which was also the period during which several outbreaks of B pertussis among the Afghan civilian population occurred.134 ,138 ,139 Epidemics of B pertussis among military populations, both under training and while on deployed operations, raises numerous questions including how best to minimise transmission under field conditions in an environment of limited resources and high person density; the role of prophylactic antibiotics and what is the optimal regime (eg, a 3 day course of azithromycin which is effective, well tolerated and offers a high degree of compliance); and, finally, whether personnel should receive a booster vaccination against pertussis, especially prior to deployment overseas to developing countries in which B pertussis remains endemic.108 ,140 ,141


The environments of military training facilities and operational theatres, and the stressful activities undertaken in such settings, are unique. It is clear that military personnel, either undergoing initial recruit training or those newly deployed to an operational setting, are at considerable risk of the acquisition and onward transmission of a variety of respiratory infections, a risk typically greater than that experienced by comparable civilian populations. While such infections generally cause mild illness, severe disease may occur with significant associated morbidity and occasionally mortality. Epidemic outbreaks among the military are well recognised and may have a significant detrimental impact on training schedules and operational effectiveness.

This review has focused on three of the most important respiratory infections in the military to illustrate these issues. The key to appreciating the burden of such illness and disease among British military personnel, and to accrue the evidence required to inform decisions concerning appropriate interventions to prevent or limit negative impacts (such as vaccine strategies and the role of prophylactic antibiotics) can only be through the use of comprehensive laboratory-based surveillance programmes. We support the establishment of such programmes and are currently undertaking related research with the aim of facilitating such a development.


View Abstract


  • Contributors DW initial concept. MKO and DW co-wrote the article.

  • Funding This research received no specific funding.

  • Competing interests None.

  • Provenance and peer review Not commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.