Seroprevalence of Infection by Borrelia Species Responsible for Lyme Disease in the French Alps: Analysis of 27,360 Serology Tests,2015

Abstract

Objectives:

Lyme borreliosis incidence is increasing in several areas; moreover, it has recently gained the public's attention. Apart from erythema migrans, Lyme disease diagnosis relies (among others) on serology test; however, the prevalence of positive enzyme-linked immunosorbent assay (ELISA) and western blot (WB) assay has been poorly studied in the general population. We aimed to approach the seroprevalence of infection by Borrelia species responsible for Lyme disease in the French Isere department using city laboratories data.

Patients and Methods:

We retrieved all serological tests for Borrelia species responsible for Lyme disease performed in the two main networks of city laboratories between 2015 and 2020. All patients with both ELISA and WB IgG were considered seropositive.

Results:

We analyzed 27,360 tests (ELISA/ELISA+WB). Mean age was 50.9 ± 20.3 years (ranges: 0–101), with 57.1% females. Overall, 11.7% had IgG detected by ELISA, and 4.7% had IgG detected by both ELISA and WB assay. Seropositive status was more frequent in males (7.0% vs. 2.9%, p < 0.001). Seropositivity rate increased with age after a first peak in childhood; men aged 61–70 years had the highest seropositivity rate (10.3%). In addition, seropositivity rate was higher in persons from a rural area. In multivariate analysis, older age, male gender and living in a rural area were independently associated with seropositivity. Seropositivity rate was stable on the 2017–2020 period.

Conclusion:

The seroprevalence of infection by Borrelia species responsible for Lyme disease is high in Isere; this probably reduces the predictive positive value for Lyme disease of ELISA and WB IgG, suggesting that this serological test should not be performed for nonspecific symptoms.

Introduction

In the past two decades, increasing incidences of Lyme disease have been reported in Western Europe (Sykes and Makiello, 2017) (56/100,000) and in the United States (Nelson et al., 2015) (up to 123/100,000 in Connecticut). Most cases are erythema migrans, a relatively easy clinical diagnosis, particularly in subjects reporting a recent tick bite. However, Borrelia burgdorferi infection may also be associated with facial palsy, meningoradiculitis, arthritis, and myocarditis; in these cases, diagnosis can prove challenging, and relies (among others) upon a serology test.

In parallel, Lyme disease has recently gained much attention from the general public often through sensationalized media coverage (Gocko et al., 2021), and with allegations that chronic fatigue, chronic diffuse pains, chronic arthralgia, and other nonspecific chronic symptoms are due to Borrelia infection (Sigal and Hassett, 2002).

In this context marked by the higher incidence of Lyme disease, increased media coverage and the assumption that the disease is responsible for a wide range of symptoms, general practitioners may be tempted (and/or requested by patients) to prescribe a serological test for Borrelia infection to establish a diagnosis of Lyme disease in the case of nonspecific chronic symptoms. However, Borrelia serology is not a dynamic test, as it provides no information about the activity of the infection. High seroprevalence would thus lower the predictive positive value regarding the diagnosis of active Lyme disease.

We aimed to determine the seroprevalence of infection by Borrelia species responsible for Lyme disease in the Isere department of France using the results of serological tests performed in city laboratories between 2015 and 2020.

Methods

We performed a retrospective study of serological tests for Lyme disease (or, more precisely, for infections by Borrelia species responsible for Lyme disease) that were performed in private local laboratories in the Isere department (population of 1.26 million).

Several studies (Botman et al., 2018; Coumou et al., 2014; Fix et al., 1998; Gasmi et al., 2017; Hinckley et al., 2014; John and Taege, 2019; Ramsey et al., 2004; Vreugdenhil et al., 2020) have pointed to a high misuse of prescriptions for Borrelia serological testing. A high proportion of “tests for Lyme disease” are performed in patients whose symptoms are not necessary evocative of this diagnosis (e.g., chronic fatigue, chronic polyarthralgia without arthritis, feeling of chronic cognitive decline, chronic polyalgia). Therefore, the people undergoing a “Lyme disease serological test” may be considered to have similar characteristics to the general population. This led us to consider serological tests for Lyme disease, whatever their purpose, to conduct a seroprevalence study.

In the Isere department, 2 laboratory companies account for 75% of the analysis facilities in the area: Oriade (57 facilities) and Eurofins (11 facilities). We anonymously collected all the serological tests for Lyme disease in these two companies from 2015 to 2020 (Oriade) and from 2017 to 2020 (Eurofins).

The tests used were:

- For enzyme-linked immunosorbent assay (ELISA): LIAISON^® Borrelia IgG and IgM assays (Diasorin™) for both laboratories;

- For western blot (WB): LYMECHECK^® Optima (Biosynex™; Oriade Laboratory) and RecomLine Borrelia (Diasorin; Eurofins Laboratory).

In France, as in most countries, the serological test for Borrelia infection features an ELISA (IgG and IgM) for screening, completed by a WB assay (either IgG or IgM) in the case of ELISA positivity. In the analysis, we included all ELISA samples (IgG and IgM), and all WB assays (IgG and IgM) provided that the test was performed after a positive ELISA of the corresponding immunoglobulin class, for persons living in the Isere department (no age limit).

WB was not retrospectively reinterpreted: we collected the conclusions of the biologist in charge of the test (positive, doubtful, negative). To determine seroprevalence, we considered any person with ELISA positive for IgG and WB assay positive for IgG to be seropositive and all other persons to be seronegative (i.e., ELISA negative for IgG; ELISA positive for IgG but negative or doubtful WB assay for IgG). As IgM specificity is consistently less than IgG, we did not use the ELISA and WB assay results for IgM to determine seroprevalence.

The Isere department features 29 townships. We constituted 13 geographical zones: 5 mostly rural (220,000 inhabitants), 2 mostly urban (410,000 inhabitants), and 6 mixed (620,000 inhabitants).

A chi-squared test was used to assess the association between categorical variables, and a Spearman test to assess the association between continuous variables. A Mann-Whitney test was used to compare the distribution of continuous variables in different groups. For multivariate analysis, we used a binary logistic regression model; all variables with a p value <0.05 were added to the model (including age, which had a Gaussian distribution). Considering the number of participants, we did not perform any stepwise selection. R software was used.

The study was declared to the French National Commission for Freedom and Information Systems (CNIL) on July 7, 2021 (N° 2222963).

Results

We analyzed 27,360 serological tests. The mean age of the tested individuals was 50.9 ± 19.8 years; 57.1% were female (mean age 51.1 ± 19.3 years) and 42.9% were male (mean age 50.7 ± 20.5 years; p = 0.869 for the difference) (Table 1).

Table 1.

Number of Serological Tests Analyzed According to Age and Gender

Age class (years)	Total	Female	Male
≤10	810	383	467
11–20	1513	808	697
21–30	2210	1282	928
31–40	3493	2013	1479
41–50	4492	2703	1788
51–60	5186	3101	2083
61–70	4793	2628	2164
71–80	3376	1888	1488
81–90	1345	755	590
>90	80	49	31

Gender was not known in 33 cases and age in 2 cases.

ELISA and WB assay results

ELISA was positive for IgM in 9.5% of tests and for IgG in 11.7%. Among the tests with ELISA positive for IgM, WB assay for IgM was negative in 92.2% of cases, doubtful in 1.4%, and positive in 6.4% (in three cases [0.01%], IgG and IgM WB were performed although both ELISA were negative). Among the tests with ELISA positive for IgG, WB assay for IgG was not performed in 0.6%, negative in 57.7%, doubtful in 2.0%, and positive in 39.7%. Overall seropositivity rate (ELISA+WB IgG) was 4.7% (Table 2).

Table 2.

IgG Enzyme-Linked Immunosorbent Assay and Western Blot Results According to Gender

Serological result		Total		Female		Male
ELISA	WB	N	%	N	%	N	%
Negative	Not performed	24,151	88.3	14,131	90.5	9995	85.3
Positive	Not performed	20	0.1	11	0.1	9	0.1
	Negative	1852	6.7	980	6.3	865	7.4
	Doubtful	63	0.2	35	0.2	28	0.2
	Positive	1274	4.7	454	2.9	819	7.0

Gender was not known in 33 cases.

ELISA, enzyme-linked immunosorbent assay; WB, western blot.

Factors associated with seropositive status

The seropositivity rate was different in the two laboratory companies: it was 3.5% for Eurofins and 4.8% for Oriade (p < 0.001). In addition, this rate was different along the years in the two companies (Supplementary Table S1).

Seroprevalence was higher in men (7.0% vs. 2.9%, p < 0.001) (Table 2); this difference was observed in all age classes (Fig. 1). After a peak in childhood, seroprevalence increased with age; it was 3.2% between 0 and 10 years and 3.0% between 11 and 20 years before decreasing to 1.6% between 21 and 30 years and then continuously increasing up to 7.5% for the 61–70 age class (Fig. 1). In subjects between the ages of 18 and 80 (N = 23,866), age and seropositivity rate were correlated (r = 0.854). The highest seroprevalence was observed in men aged between 61 and 70 years (10.3%, Fig. 1). Seroprevalence also significantly varied according to the geographical area, being higher in mostly rural areas (Table 3).

FIG. 1.

Seroprevalence for males and females in each age class. The size of the circle is proportional to the number of serology tests performed in the age class.

Table 3.

Serology Status According to Gender, Age, and Living Area

	Negative serology	Positive serology	p
Gender
Males	93.0%	7.0%	<0.0001
Females	97.1%	2.9%	<0.0001
Age
Median [IQR]	52 [37–66]	62 [48–71]	<0.0001
Mean (±SD)	50.5 (±19.8)	58.3 (±18.2)	<0.0001
Areas
Rural	92.9%	7.1%	<0.0001
Intermediate	95.5%	4.5%
Urban	96.2%	3.8%

Gender was not known in 33 cases and age in 2 cases. Positive serology: positive IgG ELISA and WB, negative serology: other cases.

IQR, interquartile range; SD, standard deviation.

In multivariate analysis, male gender (adjusted odds ratio, OR = 2.5 [95% confidence interval, CI 2.2–2.8], p < 0.001), age (adjusted OR = 1.022 [95% CI 1.019–1.025] per year, p < 0.001), living in a mostly rural area (adjusted OR = 1.7 [95% CI 1.5–1.9], p < 0.001), and the laboratory company Oriade (adjusted OR = 1.3 [95% CI 1.08–1.54], p < 0.001) were independently associated with seropositivity.

Seroprevalence did not increase from 2015 to 2020: it was 5.9%, 5.2%, 4.3%, 4.4%, 4.4%, and 4.4% for each year included in the analysis.

Discussion

Using serological tests performed in 75% of laboratories in the French Isere department, we observed a seroprevalence of 4.6% for infection by Borrelia species associated with Lyme disease. Previous studies evaluated this seroprevalence in the general population: 2.5% in Romania (Kalmar et al., 2021) (1,200 blood donors aged 18–65 years), 3.9% in Finland (van Beek et al., 2018) (2,000 persons aged 29 years and over), 4.2% in Scotland (Munro et al., 2015) (1,440 blood donors), 1–10% (according to the area) in Tyrol, Austria (Sonnleitner et al., 2015) (1,660 adult blood donors), 5% in Spain (Barreiro-Hurle et al., 2020) (316 blood donors), and 0.5% in California (Brummitt et al., 2020) (1,700 blood donors).

This heterogeneity probably reflects the distribution of vector ticks, although other factors may also play a role. As expected, this seroprevalence was higher in individuals with an occupational exposure to ticks: in forestry workers, it was 8.1% in South Korea (Acharya and Park, 2021), 21.6% in Belgium (De Keukeleire et al., 2017), 11.8% in Serbia (Jovanovic et al., 2015), 17.8% in Poland, and 15.2% in France (Zhioua et al., 1997).

We observed a much higher seroprevalence in males. This has been found in most previous studies conducted in Romania (Kalmar et al., 2021), Finland (van Beek et al., 2018), Austria (Sonnleitner et al., 2015), and Spain (Barreiro-Hurle et al., 2020) but not in Scotland (Munro et al., 2015). This predominance in males may be related to their higher exposure to tick areas for work- or leisure-related activities. This is also supported by an American study (Marx et al., 2021) showing that in patients admitted to the emergency room for tick bites, males outnumbered females. Even though such stereotypes (i.e., males exposed to more opportunities for tick bites) may be less true in recent years, we still observed this gender-based difference in the younger age classes.

This is also consistent with an American study (Schwartz et al., 2017) and a Canadian study (Kulkarni et al., 2019), which showed that the diagnosis of Lyme disease is more frequent in males, and another American study (Kugeler et al., 2022), which observed a greater increase in incidence in males, although this gender-based difference was not observed in other countries (Beltrame et al., 2021; Enkelmann et al., 2018). By contrast, a Slovenian study observed that the incidence of erythema migrans was higher in females (Strle et al., 2013). A Swedish study (Bennet et al., 2007) showed that females older than 40 years of age were more frequently bitten by ticks than males of the same age with a similar exposure; to our knowledge, no other work has investigated whether ticks are more likely to bite females.

Interestingly, the seroprevalence for another tick-borne pathogen (among other transmission routes), Coxiella burnetti, was also higher in males in a study from the Netherlands (van Roeden et al., 2018).

We observed an increase in seroprevalence with age, which was also reported in some European studies (Barreiro-Hurle et al., 2020; Kalmar et al., 2021; van Beek et al., 2018) but not from Scotland (Munro et al., 2015). This pattern was observed with most infections that are not typically acquired in childhood, as observed with hepatitis E in France, for example (Mansuy et al., 2011). Even if older patients may have recently been bitten by ticks, antibodies may persist many years after an infection, as observed 8–15 years after treated Lyme borreliosis (Peltomaa et al., 2003). We observed a higher seroprevalence in children compared with adults aged 20–30 years, which is supported by studies on the epidemiology of the disease showing a similar peak during childhood in the United States (Kugeler et al., 2022), Germany (Enkelmann et al., 2018), and Canada (Kulkarni et al., 2019).

The main implication of our study relates to the positive predictive value of serology for the diagnosis of Lyme disease. Indeed, given the high seroprevalence in the study area, particularly in older males, “Lyme disease serological test” should not be performed when the clinical picture is not suggestive of Lyme disease, as the frequently positive results might falsely lead to the conclusion that Borrelia genus is responsible for such nonspecific symptoms (Forestier et al., 2018). It is notable that most of serological tests studied here were performed in women; meanwhile, more men were positive. This finding suggests that physicians misuse this test more often with females. Moreover, as a longitudinal survey concluded that only 5% of positive serological tests are associated with Lyme disease (Fahrer et al., 1998), this particularly low predictive positive value should be kept in mind when prescribing this test.

One strength of our study is the large number of serological tests analyzed, making it the largest sample to date. Another strength is that the participating laboratory networks covered 75% of the area (Isere department). These two strengths contribute to providing a representative view of seroprevalence. The main weakness of our study is our use of serological tests performed to establish or rule out a diagnosis of Lyme disease, which may introduce biases (and would theoretically overestimate the seroprevalence). However, as stated, it is likely that this test is frequently misused (being performed in clinical situations not consistent with Lyme disease), meaning that the sampled individuals can be considered a representative sample of the general population (despite the sex ratio).

To assess disease seroprevalence in a territory, the most direct way is to perform blood sampling of the population or a fraction of a population. However, as this is both a complex and ethically questionable method, other sources should be considered. Blood donation may provide a high number of samples, although the representativeness of the sampled population may be disputable. Retrieving the serological results from private local laboratories provides a large number of samples from a broader age range (compared with blood donation) and with existing analyses. For example, the use of this method in Epstein-Barr virus infection (Fourcade et al., 2017) demonstrated a change in seroprevalence over the years.

Conclusion

Seroprevalence in the Isere department in France is high, particularly in males, and increases with age. This suggests that the serology test for Borrelia species responsible for Lyme disease should not be performed in the case of a nonspecific clinical picture.

Data Availability

The datasets generated analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

Footnotes

Author Disclosure Statement

No conflict of interest to declare.

Funding Information

No funding was allocated to this study.

Supplementary Material

Supplementary Table S1

References

Acharya

, Park

. Seroepidemiologic survey of Lyme disease among forestry workers in national park offices in South Korea. Int J Environ Res Public Health, 2021; 18:2933.

Barreiro-Hurle

, Melon-Garcia

, Seco-Bernal

, et al. Seroprevalence of Lyme disease in southwest Asturias. Enferm Infecc Microbiol Clin (Engl Ed), 2020; 38:155–158.

Beltrame

, Rodari

, Mauroner

, et al. Emergence of Lyme borreliosis in the province of Verona, Northern Italy: Five-years of sentinel surveillance. Ticks Tick Borne Dis, 2021; 12:101628.

Bennet

, Stjernberg

, Berglund

. Effect of gender on clinical and epidemiologic features of Lyme borreliosis. Vector Borne Zoonotic Dis, 2007; 7:34–41.

Botman

, Ang

, Joosten

JHK

, et al. Diagnostic behaviour of general practitioners when suspecting Lyme disease: A database study from 2010–2015. BMC Fam Pract, 2018; 19:43.

Brummitt

, Kjemtrup

, Harvey

, et al. Borrelia burgdorferi and Borrelia miyamotoi seroprevalence in California blood donors. PLoS One, 2020; 15:e0243950.

Coumou

, Hovius

, Van Dam

. Borrelia burgdorferi sensu lato serology in the Netherlands: Guidelines versus daily practice. Eur J Clin Microbiol Infect Dis, 2014; 33:1803–1808.

De Keukeleire

, Vanwambeke

, Cochez

, et al. Seroprevalence of Borrelia burgdorferi, Anaplasma phagocytophilum, and Francisella tularensis infections in Belgium: Results of three population-based samples. Vector Borne Zoonotic Dis, 2017; 17:108–115.

Enkelmann

, Bohmer

, Fingerle

, et al. Incidence of notified Lyme borreliosis in Germany, 2013–2017. Sci Rep, 2018; 8:14976.

10.

Fahrer

, Sauvain

, Zhioua

, et al. Longterm survey (7 years) in a population at risk for Lyme borreliosis: what happens to the seropositive individuals?. Eur J Epidemiol, 1998; 14:117–123.

11.

Fix

, Strickland

, Grant

. Tick bites and Lyme disease in an endemic setting: problematic use of serologic testing and prophylactic antibiotic therapy. JAMA, 1998; 279:206–210.

12.

Forestier

, Gonnet

, Revil-Signorat

, et al. [Pathway to diagnosis and real-life experience of patients believing they are affected by “chronic Lyme disease”]. Rev Med Interne, 2018; 39:912–917.

13.

Fourcade

, Germi

, Guerber

, et al. Evolution of EBV seroprevalence and primary infection age in a French hospital and a city laboratory network, 2000–2016. PLoS One, 2017; 12:e0175574.

14.

Gasmi

, Ogden

, Leighton

, et al. Practices of Lyme disease diagnosis and treatment by general practitioners in Quebec, 2008–2015. BMC Fam Pract, 2017; 18:65.

15.

Gocko

, Tattevin

, Lemogne

. Genesis and dissemination of a controversial disease: Chronic Lyme. Infect Dis Now, 2021; 51:86–89.

16.

Hinckley

, Connally

, Meek

, et al. Lyme disease testing by large commercial laboratories in the United States. Clin Infect Dis, 2014; 59:676–681.

17.

John

, Taege

. Appropriate laboratory testing in Lyme disease. Cleve Clin J Med, 2019; 86:751–759.

18.

Jovanovic

, Atanasievska

, Protic-Djokic

, et al. Seroprevalence of Borrelia burgdorferi in occupationally exposed persons in the Belgrade area, Serbia. Braz J Microbiol, 2015; 46:807–814.

19.

Kalmar

, Briciu

, Coroian

, et al. Seroprevalence of antibodies against Borrelia burgdorferi sensu lato in healthy blood donors in Romania: An update. Parasit Vectors, 2021; 14:596.

20.

Kugeler

, Mead

, Schwartz

, et al. Changing trends in age and sex distributions of Lyme disease—United States, 1992–2016. Public Health Rep, 2022; 137:655–659.

21.

Kulkarni

, Narula

, Slatculescu

, et al. Lyme disease emergence after invasion of the blacklegged tick, Ixodes scapularis, Ontario, Canada, 2010–2016. Emerg Infect Dis, 2019; 25:328–332.

22.

Mansuy

, Bendall

, Legrand-Abravanel

, et al. Hepatitis E virus antibodies in blood donors, France. Emerg Infect Dis, 2011; 17:2309–2312.

23.

Marx

, Spillane

, Beck

, et al. Emergency department visits for tick bites—United States, January 2017-December 2019. MMWR Morb Mortal Wkly Rep, 2021; 70:612–616.

24.

Munro

, Mavin

, Duffy

, et al. Seroprevalence of lyme borreliosis in Scottish blood donors. Transfus Med, 2015; 25:284–286.

25.

Nelson

, Saha

, Kugeler

, et al. Incidence of clinician-diagnosed Lyme disease, United States, 2005–2010. Emerg Infect Dis, 2015; 21:1625–1631.

26.

Peltomaa

, Mchugh

, Steere

. Persistence of the antibody response to the VlsE sixth invariant region (IR6) peptide of Borrelia burgdorferi after successful antibiotic treatment of Lyme disease. J Infect Dis, 2003; 187:1178–1186.

27.

Ramsey

, Belongia

, Chyou

, et al. Appropriateness of Lyme disease serologic testing. Ann Fam Med, 2004; 2:341–344.

28.

Schwartz

, Hinckley

, Mead

, et al. Surveillance for Lyme disease—United States, 2008–2015. MMWR Surveill Summ, 2017; 66:1–12.

29.

Sigal

, Hassett

. Contributions of societal and geographical environments to “chronic Lyme disease”: the psychopathogenesis and aporology of a new “medically unexplained symptoms” syndrome. Environ Health Perspect, 2002; 110(Suppl 4):607–611.

30.

Sonnleitner

, Margos

, Wex

, et al. Human seroprevalence against Borrelia burgdorferi sensu lato in two comparable regions of the eastern Alps is not correlated to vector infection rates. Ticks Tick Borne Dis, 2015; 6:221–227.

31.

Strle

, Wormser

, Mead

, et al. Gender disparity between cutaneous and non-cutaneous manifestations of Lyme borreliosis. PLoS One, 2013; 8:e64110.

32.

Sykes

, Makiello

. An estimate of Lyme borreliosis incidence in Western Europe. J Public Health (Oxf), 2017; 39:74–81.

33.

van Beek

, Sajanti

, Helve

, et al. Population-based Borrelia burgdorferi sensu lato seroprevalence and associated risk factors in Finland. Ticks Tick Borne Dis, 2018; 9:275–280.

34.

van Roeden

, Holsboer

, Oosterheert

, et al. Seroprevalence of Coxiella burnetii antibodies and chronic Q fever among post-mortal and living donors of tissues and cells from 2010 to 2015 in the Netherlands. Euro Surveill, 2018; 23:17-00384.

35.

Vreugdenhil

, Leeflang

, Hovius

, et al. Serological testing for Lyme Borreliosis in general practice: A qualitative study among Dutch general practitioners. Eur J Gen Pract, 2020; 26:51–57.

36.

Zhioua

, Rodhain

, Binet

, et al. Prevalence of antibodies to Borrelia burgdorferi in forestry workers of Ile de France, France. Eur J Epidemiol, 1997; 13:959–962.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.01 MB

0.00 MB

Seroprevalence of Infection by Borrelia Species Responsible for Lyme Disease in the French Alps: Analysis of 27,360 Serology Tests,2015–2020

Abstract

Objectives:

Patients and Methods:

Results:

Conclusion:

Introduction

Methods

Results

ELISA and WB assay results

Factors associated with seropositive status

Discussion

Conclusion

Data Availability

Footnotes

Author Disclosure Statement

Funding Information

Supplementary Material

References

Supplementary Material