Introduction

Individuals in the meat packing, produce processing, and farm working industries, referred to collectively as food processing industries are exposed to some of the greatest occupational hazards in the United States [1–4]. These risks are related to physically strenuous work, long hours, dangerous machinery, chemical exposures, and poor work protections [2,3]. Furthermore, workers – many of whom are migrants - often have limited access to health care resources and little autonomy to advocate for improved working conditions. Health disparities among food processing workers were recently highlighted by the disproportionate burden in this population during the COVID-19 pandemic, particularly among racial and ethnic minorities [5]. Yet, little is known about the potential occupational risk of tickborne diseases in food processing workers.

Over the last twenty years, the annual incidence of tickborne diseases has more than doubled in the U.S., and workers in certain industries appear to be more susceptible, particularly those requiring work outdoors [6–8]. In North Carolina, two of the most common tickborne diseases are Spotted Fever rickettsiosis (SFR), a category which includes Rocky Mountain spotted fever (RMSF, caused by Rickettsia rickettsii), and ehrlichiosis. Transmitted by the American dog tick (Dermacentor variabilis) and lone star tick (Amblyomma americanum), respectively, both infections can be fatal if not promptly identified and treated [9]. According to the North Carolina Department of Health and Human Services, the average incidence rate in North Carolina from 2015-2019 was 0.65 cases per 100,000 residents for ehrlichiosis, and 5.15 cases per 100,000 residents for SFR, over twice the national average [10].

Our study group previously conducted an observational cohort study among Latino workers employed in the food processing industry and their family members in central North Carolina, which has one of the largest food processing industries in the U.S [11]. Although the primary aim was to investigate the epidemiology of COVID-19, we also examined associations between other infections, including tickborne diseases, with potential occupational risk factors. By leveraging the existing database of food processing workers and their family members, we investigated sociodemographic and occupational risk factors for Ehrlichia and Spotted Fever Group Rickettsiosis (SFGR) infection among Latino food processing communities in North Carolina.

Methods

Results

Baseline characteristics

Sufficient plasma were available for testing for 201 of 224 (89.7%) individuals enrolled in COFF-NC, including 112 index workers and 89 family members. Participants excluded due to inadequate sample volume did not significantly differ from the remaining participants by age or sex. Participants had a median age of 34 years (interquartile range [IQR] 15–47), 43.8% were male, and 80.6% spoke Spanish as their primary language. Slightly less than half of participants were employed full-time, and about a quarter attended school full-time (Table 1). Among index workers, 42 worked in meat packing, 10 worked in produce processing, and 60 worked in farming. Of the farm workers, 9 were self-reported as migrant farm workers.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Baseline characteristics.

https://doi.org/10.1371/journal.pgph.0005116.t001

SFGR seroprevalence

Among 201 total participants, 30 (14.9%) were seropositive for SFGR (Table 2). SFGR seropositive participants were significantly older than seronegative participants (median 45 [IQR 35–55] vs. 33 [14–45] years, p < 0.001) (Fig 1). Older participants had 1.05 times the odds of SFGR positivity (95% confidence interval [CI] 1.02-1.07). Participants with a smoking history (odds ratio [OR] 3.19, 95% confidence interval [CI]: 1.09-9.30) or with current school attendance (OR 0.15, 95% CI: 0.03-0.66) had higher and lower odds of SFGR seropositivity, respectively; these associations lost statistical significance when controlled for age. No other participant factors were significantly associated with SFGR seropositivity, including sex, primary language, employment status, medical comorbidities, or recent travel. Among adults aged 18 years and older, SFGR seropositivity rates did not significantly differ between index workers and household members.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Participant characteristics associated with SFGR seropositivity among index workers and household members.

https://doi.org/10.1371/journal.pgph.0005116.t002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. SFGR seroprevalence by age category.

https://doi.org/10.1371/journal.pgph.0005116.g001

Among 112 index workers, 22 participants (19.6%) were seropositive for SFGR (Table 3). SFGR seropositive participants were significantly older than seronegative participants (median 45.5 [IQR 37–52] vs. 41 [31–49] years, p = 0.05). SFGR seropositivity rates were highest among farm workers (15, 25.0%), followed by meat packing workers (7, 16.7%) and produce processing workers (0, 0.0%), although no statistically significantly differences were seen between groups. There was no significant association with indoor vs. outdoor work, hours worked per week, home type, or household size. In a sensitivity analysis in which equivocal SFGR results (n = 6) were considered either positive or excluded, no significant change in results was found (S1 and S2 Tables).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Participant characteristics associated with SFGR seropositivity among index workers.

https://doi.org/10.1371/journal.pgph.0005116.t003

Ehrlichia seroprevalence

Among 201 total participants, 40 (19.9%) were seropositive for Ehrlichia (Table 4). Ehrlichia seropositivity appeared to have a bimodal association with age, with peaks in children under age 10 and adults in their forties and fifties (Fig 2). Participants with a history of diabetes mellitus or hypertension had higher odds of Ehrlichia seropositivity (OR 2.67, 95% CI: 1.24-5.77), and those who currently attended school had lower odds of Ehrlichia seropositivity (OR 0.37, 95% CI: 0.14-0.93); these associations lost statistical significance when controlled for age. No other participant factors were significantly associated with Ehrlichia seropositivity, including sex, primary language, employment status, or recent travel.

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Participant characteristics associated with Ehrlichia seropositivity among index workers and household members.

https://doi.org/10.1371/journal.pgph.0005116.t004

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Ehrlichia seroprevalence by age category.

https://doi.org/10.1371/journal.pgph.0005116.g002

Among 112 index workers, 25 (22.3%) were seropositive for Ehrlichia (Table 5). Ehrlichia seroprevalence did not significantly differ by industry. Ehrlichia seroprevalence was similar between industries (23.8% among meat packing workers, 20.0% among produce processing workers, and 21.7% among farm workers). There was no significant association with indoor vs. outdoor work, hours worked per week, home type, or household size.

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. Participant characteristics associated with Ehrlichia seropositivity among index workers.

https://doi.org/10.1371/journal.pgph.0005116.t005

Discussion

Among 201 food processing workers and their family members living in North Carolina during the fall of 2020, we identified SFGR and Ehrlichia IgG seropositivity in 14.9% and 19.9%, respectively. Seropositivity to at least one tickborne illness was identified in 31.3% of participants. Seroprevalence rates increased with age for both tickborne illnesses, although Ehrlichia was also common among children aged 10 years and younger, leading to a bimodal distribution for that pathogen. No other covariates in our analysis were significantly associated with SFGR or Ehrlichia seropositivity when adjusted for age. Having a seropositive household member was associated with higher risk of seropositivity for the same tickborne illness.

Limited data exist to compare our findings to prior seroprevalence studies in the North Carolina. A recent study of SFGR and Ehrlichia IgG antibodies among soldiers in North Carolina in 2019 showed seroprevalences of 12.0% and 3.8%, respectively [15]. Another cohort of 488 individuals in a population-based cohort from central North Carolina in 2017–2018 found SFGR and Ehrlichia seroprevalences of 17.1% and 8.6%, respectively [16]. In a 2016 study of 159 North Carolina outdoor workers, seroprevalence for Rickettsia rickettsii and Ehrlichia chaffeensis were 19% and 4%, respectively [8]. In our study, seroprevalence rates were similar to those from past studies for SFGR. However, the observed seroprevalence of Ehrlichia was markedly higher in our study at 19.9%. It is unclear whether this higher Ehrlichia seroprevalence is driven by geographic differences, occupational exposures, socioeconomic differences, or other factors. Some studies have reported rising Ehrlichia seroprevalence nationwide over the last few decades, thought to be in part due to expanding range of the lone star tick related to climate change [17,18], which could partially contribute to the higher seroprevalence observed in our study.

We found a marked association between older age and SFGR seropositivity, consistent with country-wide epidemiological data by the Centers for Disease Control and Prevention (CDC), which show a gradual increase in disease incidence by age group [19]. As tickborne antibodies can persist for years, this is an expected finding as older individuals have more time at risk. The CDC’s data show increasing Ehrlichia incidence with older age, although interestingly our study also found a high Ehrlichia seroprevalence in children under 10 years of age. Among the 25 children in our study age 10 years or younger, 6 (24.0%) were seropositive for Ehrlichia. High Ehrlichia seroprevalence in this population is concerning, as higher case fatality rate has been observed in children under 5 years of age [20]. If Ehrlichia incidence has indeed increased in the last decade, the bimodal distribution could represent a cohort effect [17,18]. As our study occurred during fall of 2020 when schools were closed due to the COVID-19 pandemic, it is possible that children were spending more time outside. This finding could also be explained by artifact in a small sample size.

We found that for both SFGR and Ehrlichia, individuals with a seropositive household member had a higher risk of seropositivity themselves for the same infection, even after accounting for household clustering of cases. Other studies have also shown clustering of symptomatic tickborne illnesses among family members and other close contacts, presumably due to shared environmental and behavioral exposures [21–24]. Workers may also carry ticks into the home on clothing. Thus, physicians should have a high degree of suspicion for tickborne disease among household contacts of recently diagnosed individuals who subsequently develop nonspecific symptoms.

Although the difference was not statistically significant, we find it notable that farm workers had higher SFGR seroprevalence (25.0%) than meat packing plant workers (16.7%) or produce processing workers (0.0%). Farm workers are highly exposed to ticks working outdoors, and inconsistent use of protective clothing and insect repellants as well as often substandard housing, particularly among migrant workers, may increase risk of tickborne illness in this group [2]. These individuals are also more vulnerable to severe disease from tickborne infections because of limited healthcare access, lack of insurance, language barriers, and poverty [4]. Although it is unclear why Ehrlichia seroprevalence did not vary by industry when SFGR seroprevalence did, differences in local tick population size and distribution or tick infection rates could contribute. Larger surveillance studies are needed to better assess tickborne infection risk among farm workers.

Our study did not demonstrate increased risk associated with outdoor work or specific occupation within food processing industries. However, our study does not exclude the possibility of occupational risk factors, primarily because, as it was conceived as a study of risk factors for SARS-COV-2 infection, it was not designed to address that question and does not include appropriate comparison groups such as workers in other industries to make such a conclusion. As we showed that having a seropositive household member increased an individual’s risk of seropositivity, household members of workers could have elevated risk compared to the general population. We also did not have data about whether family members had ever previously worked in the food processing industry themselves. In addition, our study is relatively small and underpowered to identify small differences in risk.

It is worth noting that our diagnostic testing did not distinguish between SFGR species. Although SFGR is typically associated with Rickettsia rickettsii, the causative agent of Rocky Mountain Spotted Fever, several other species have been found to cause clinical symptoms. In particular, Rickettsia ambylommatis is one of the most common species identified among lone star ticks in North Carolina, and some have theorized that it contributes to the high RMSF diagnosis rates in this region [25,26]. In addition, there is known cross-reactivity between Ehrlichia species (E. chaffeensis, E. ewinigii, E. muris).

Our study has several limitations. First, it was primarily designed to examine occupational risk factors for transmission of SARS-CoV-2 early in the COVID-19 pandemic, which may introduce bias. We did not collect data about many potential risk factors previously associated with tickborne illnesses such as pet ownership, self-reported tick exposures, or non-occupational outdoor activities [27–29]. As our study relies on self-report, we are likely missing data on medical comorbidities of our participants. Additionally, we cannot be certain that seropositive participants acquired either infection in North Carolina, although the vast majority of our participants had a fixed residence. Our study was relatively small and may be underpowered to detect small differences in risk between groups or result in overfitting of multivariate models. As our study focused on Latino communities in food processing industries in North Carolina, the external validity of our results to other populations may be limited. Finally, we do not have a comparison group or a robust existing body of literature to estimate tickborne disease seroprevalence in the general population in nearby geographic areas.

Overall, our study found high seroprevalence of SFGR and Ehrlichia antibodies among Latino food processing workers in North Carolina, raising concern for exposure to tickborne infections among food processing workers, particularly farmers, and the broader North Carolina population. Our study is one of the first to investigate seroprevalence of tickborne infections among Latino Americans, a population that remains vulnerable and often overlooked despite its growing size. We also identified a high Ehrlichia seroprevalence among young children, which could be a sentinel of a potentially increasing public health concern in North Carolina that warrants more surveillance. Given the nationwide rise of tickborne illness, larger epidemiological and entomological surveillance studies are needed to better quantify geographic trends in tickborne illness and identify occupational and socioeconomic risk factors.

Supporting information