Study design

This part of the TBD STING-study was initiated in February of 2008 by advertisements on local television and in newspapers. Persons aged 18 years or older were recruited to the study if they had been bitten recently by a tick. The bitten person was asked to bring the tick(s) to one of 34 primary health care centres (PHCs) located in the regions of Southernmost, South Central, and Northern Sweden, and on the Åland Islands, Finland (Figure 1). At the PHC, the tick-bitten person signed a written consent to participate, donated the removed tick(s), provided a blood sample, completed a questionnaire (Additional file 1), and provided personal data (age and gender). The questionnaire contained questions about the date of detection of the attached tick(s) that led to the study inclusion, probable duration(s) of the tick attachment(s), the attachment location(s) of the tick(s) on the participant’s body, and the estimated number of tick bites the participant had contracted earlier that season.

At a follow-up visit to the PHC three months later, a second blood sample was collected and another questionnaire was completed (Additional file 2). This questionnaire inquired if the participant had had any more tick bites after the first registration. The questionnaire also contained inquiries about the participant’s general health condition during the study period, any symptoms suggestive of tick-borne disease(s) (LB, TBE etc.) and, whether or not, the participant had visited a health care provider due to such symptoms. If the participant had visit a health care provider his/hers medical records were scrutinized in order to determine if he/she was diagnosed with a tick-borne disease. Participants were also asked to collect and donate all ticks that they found attached to their bodies during the study period.

All samples were transported to Linköping University within three days, where the samples were frozen and stored at -70°C until examination.

Ethical considerations

Before a person was asked to participate in the study, staff of the PHC informed about the general outline and aims, that the participant was allowed to discontinue participation at any time; when so requested, all samples from that person would be discarded. Ethical permission for this study had been approved by the Regional Ethical Review Board, Linköping University (M132-06), and by the local Ethics Committee of the Åland Health Care, 2008-05-23.

Tick identification and tick feeding duration

Each tick was photographed dorsally and ventrally, using a USB-microscope (Dino-Lite Long AM4013TL, AnMo Electronics Corp., Taiwan) to determine species, life stage, and sex of adults based on [21–23]. To estimate the duration of blood feeding for adult female ticks and nymphs, the scutal index (the ratio of the length of the idiosoma to the width of the scutum) or the coxal index (the ratio of the distance between the basal coxae of the fourth pair of legs to the width of the scutum) were calculated as described by Gray and co-workers [16].

Statistical analyses

The potential relationships between duration of tick feeding and age classes of participants, between duration of tick feeding and attachment site, and between tick attachment site and gender of participants were evaluated with Chi-square test. This test was also used to compare the proportions of adult female ticks and nymphs that were removed before versus after 24 hours of blood feeding. Fisher’s exact test was used when the expected frequency was <5 in at least one of the cells of the contingency table. The Spearman rank correlation test was used to investigate if there was any significant association between the proportions of adult female ticks or nymphs attached > 24 hours and human age classes. Participants were categorized into one of following age classes (years): (19–29), (30–39), (40–49), (50–59), (60–69), (70–79), and (>80). Statistical analyses were performed and graphs were drawn using GraphPad Prism version 5.00 for Windows (GraphPad Software, San Diego, CA). P-values ≤ 0.05 were considered statistically significant.

Description of the tick-bitten participants and “their” removed ticks

The median age of the study population was 63 years (range 19–92). Participants aged 60–69 years was the largest age class (35%), while 19–29 year-old participants constituted the smallest group (2%). No significant differences were observed when the proportions of different tick stages were compared with age classes.

Seasonal tick infestation patterns in the studied regions

Of all participants, 39% were recruited from South Central Sweden (n = 688), 33% from the Åland Islands (n = 590), 27% from Southernmost Sweden (n = 477), and 1% from Northern Sweden (n = 15). Ninety-eight percent of the participants (2008, 824/836; 2009, 913/934) recorded the date when “their” ticks (n = 2056), found attached to the skin, were detected. In 2008, with no regard to the different stages of I. ricinus, tick infestations were recorded from mid-May to mid-October (Figure 2A). This was the case for all the studied regions except Northern Sweden where tick infestation on people in 2008 only occurred from early July to mid-August. Ticks collected from the other three regions exhibited a bimodal infestation pattern with a tick infestation depression between mid-July and early-August 2008. In 2009, tick bites were recorded from early April to early November (Figure 2B). This was the case for all the studied regions except Northern Sweden where tick infestation on people in 2009 only occurred from mid-June to early September. Ticks collected from Southernmost Sweden exhibited a bimodal infestation pattern with the first peak in mid-June and the second peak in mid-August. Ticks collected from South Central Sweden and from the Åland Islands exhibited a “dispersed” infestation pattern throughout the tick season of 2009.

Seasonal infestation of the different stages of I. ricinus

In 2008, with no regard to geographical origin of the ticks, adult ticks and nymphs were detected from mid-May to mid-October (Figure 3A). Larvae were detected from early June to mid-September. The nymphal infestation pattern on humans was bimodal with an infestation depression between mid-July and early August. Adult and larval infestation patterns were unimodal with peaks in mid-August and early August, respectively. In 2009, adult ticks and nymphs were detected from early April and mid-April, respectively, to mid-September and early November, respectively (Figure 3B). Larvae were detected from mid-May to mid-September. The nymphal infestation peaked in mid-June and early August, while peaks of adult and larval infestations took place in mid-July.

Tick feeding sites on the human body

Of the 1770 participants, 93% reported the attachment site for a total of 1881 ticks. Among these 1881 I. ricinus ticks, a significantly greater proportion of adult female ticks was attached to the skin of the head/neck area (P < 0.001), on the skin of the torso/dorsum area (P < 0.001), and in the groin/genital area (P < 0.01) compared to the proportions of nymphs attached to the corresponding locations (Figure 4). In contrast, greater proportions of nymphs were found on the arms (P < 0.001), and legs (P < 0.001), compared to the proportions of adult female ticks on these extremities. Due to small number of adult male ticks and larvae, they were excluded from the statistical analyses.

For both men and women, the legs were the major location of tick attachment (51% of 597 ticks on men, and 49% of 1051 ticks on women; Figure 5). This was followed by the torso/dorsum (20% and 24%, respectively) and arms (19% and 17%, respectively). A significantly (P = 0.001) greater proportion of men (9%) than women (5%) recorded ticks attached to the groin/genital area. In contrast, a significantly (P < 0.001) greater proportion of women (5%) than men (1%) recorded ticks attached to the head/neck area. No other significant differences regarding tick attachment sites on men compared to those on women were detected.

No significant differences were found when tick attachment sites were analysed in relation to age classes of the human hosts.

Duration of tick attachment

Based on the scutal and coxal indices of the ticks [16], we were able to estimate the duration of attachment (= feeding time) for 1710 ticks (410 adult female ticks and 1300 nymphs).

Among nymphs (n = 1300), 37% had been attached< 24 hours and 63% > 24 hours (Table 2). The head/neck area appeared to have the greatest proportion of nymphs (71%) attached > 24 hours, while the arms seemed to have the lowest proportion of nymphs (54%) attached > 24 hours. However, when nymphal data on attachment time was statistically analysed in relation to site of attachment, no significant differences were revealed. When attachment time, i.e. nymphs removed < 24 hours versus nymphs removed > 24 hours was analysed in relation to participant’s gender, women removed a significantly greater proportion of nymphs (40%) within 24 hours compared to that of men (32%, P < 0.001) (Table 3).

Calculated time of tick feeding in relation to the participants’ self-estimated time of tick attachment

Calculated times in hours (h) of tick feeding, based on the scutal and coxal indices (intervals 0–24 h, 24–48 h, 48–72 h, and 72–96 h), together with the participants’ self-estimated times of tick attachment were available for 748 ticks. For the calculated interval 0–24 h of tick feeding, participants’ self-estimated the time of tick attachment to 0–91 h (median 15 h, interquartile range [IQR] 7–23 h); for the calculated interval 24–48 h the self-estimated times ranged from 0 to 92 h (median 20 h, IQR 9–33 h); for the interval 48–72 h to 0–91 h (median 20 h, IQR 10–40 h); and for the calculated interval 72–96 h the self-estimated times of tick attachment ranged from 0 to 98 h (median 29 h, IQR 19–50 h).

Seasonal distribution of infesting ticks

The general seasonal infestation pattern of nymphs during both years was bimodal, with peaks in June-July and August. A similar bimodal seasonal activity pattern of host-seeking nymphs was recorded during most study years in two different field studies in south-central Sweden [7, 8], and in one year a unimodal activity pattern was recorded [8]. The depression in nymphal infestation on humans, observed in the present study may, therefore, be a consequence of the depression in nymphal host-seeking activity. The general seasonal infestation pattern of adult ticks during both years was unimodal rather than bimodal. The reason for this is unknown but the adult ticks, compared to nymphs, may have a greater resistance to relative humidity and as a result, they may not exhibit a depression in host-seeking activity pattern during the hottest and driest part of the summer. The numbers of larvae in this study were too low to enable us to draw any conclusions about their seasonal infestation pattern.

In Southernmost Sweden, South Central Sweden, and on the Åland Islands the tick infestation on participants began one month earlier and ended one month later in 2009 compared to 2008 (from early April to early November and from mid-May to mid-October, respectively). According to the Swedish Meteorological and Hydrological Institute (SMHI), in Southernmost Sweden and in South Central Sweden the mean temperature during April 2009 was higher than the mean temperature during April 2008 [24, 25]. In addition, the mean temperature for November in these regions was higher in 2009 than in 2008 [26, 27]. This may, at least partly, explain why the participants from Southernmost Sweden and South Central Sweden contracted tick bites during an extended time period in 2009 compared to 2008. For both study years in Northern Sweden, the tick infestation on participants lasted for a shorter time period (from early-June to mid-August in 2008 and from mid-June to early September in 2009). The shorter tick infestation period in Northern Sweden may reflect the generally lower abundance of ticks [2], which is a function of a relatively low tick density and relatively low diel and seasonal tick activities, which reflect the generally lower environmental temperature and the shorter growing season in northern Sweden compared to the southern regions. However, to elucidate how temperatures and other climate and weather parameters may influence the seasonal tick infestation pattern in a certain region, frequent sampling and analysis of ticks that have infested humans over multiple seasons are needed. Moreover, the seasonal tick infestation patterns found in this study are influenced not only by the tick’s particular seasonal activity pattern, which may differ among different regions, but also by the varying activities of people, e.g. when people tend to visit tick-infested areas for berry- or mushroom picking or other purposes.

Ixodes ricinus stages and “predilection sites” on humans

In the present study, only few larval ticks were removed. However, this tick stage is considered to be a much less important vector of B. burgdorferi s.l. and TBEV infections to humans; the unfed larva is almost never infected with LB-causing bacteria [28], nor with TBEV [29]. The majority of the I. ricinus ticks removed were nymphs. This stage is considered to be the most important stage in the transmission of borreliae and TBEV to humans. This is because in nature nymphs are much more numerous than adult ticks, and because nymphs, compared to adult female ticks, are more easily overlooked due to their smaller size and less conspicuous colouration. Even if fewer adult ticks were removed from the participants, adult ticks, compared to immature ticks, are, in general, more often infected with Borrelia bacteria [30]. Men removed a greater proportion of adult female ticks compared to women. In contrast, women removed a greater proportion of larval ticks compared to men. However, this does not necessarily imply that a particular tick stage of I. ricinus has a preference for a certain gender of Homo sapiens. Rather, it may reflect morphological, behavioural and physiological differences between men and women. Such differences, e.g. the usually more hairy skin of men may result in differences between men and women in their capacity to rapidly detect a tick on the skin. Morphological and other differences between men and women and between children and older persons may result in apparently tick-stage-specific “predilection sites” on the human body which may depend on how easy it is for a certain tick stage to find a suitable attachment site. Such “preferred” feeding sites on the host’s body are, most likely, much dependent also on the host’s grooming behaviour.

The majority of the ticks had attached to the legs of the participants. This site of infestation corresponds to the anatomical location where the classical sign of erythema migrans (EM) the hallmark rash of early LB, usually appears. Bennet and co-workers recorded that the most common anatomical localisations of EM among LB patients (n = 118) were the legs (63.6%) followed by torso/dorsum (24.6%), arms (10.2%) and genitalia (1.7%) [31]. These proportions correspond arbitrarily to the anatomical distribution of tick bites recorded by the participants in the present study.

The most commonly infested anatomical location, i.e. legs, is approximately within the same height above the ground where nymphs and adults of I. ricinus quest in the vegetation [32]. This suggests that most ticks, searching for an optimal attachment site on a recently encountered human host, will walk only a short vertical distance before they will attach and start to feed. However, the site of tick attachment was related to the stage of I. ricinus. Greater proportions of nymphs, compared to adult female ticks, were removed from the extremities of the participants, i.e. from legs and arms, compared to other parts of the body. In contrast, greater proportions of adult female ticks were removed from the skin of the torso/dorsum area, head/neck area, and groin/genital area compared to other parts of the body. Falco and co-workers recorded a similar behaviour, i.e. an apparent “preference” for certain body parts on human hosts by nymphs and adult females of I. scapularis[10]. Stage-related differences may, at least partly, be related to the level at which a particular tick stage quests in the vegetation. Adults of I. ricinus usually quest at a higher level above ground, compared to nymphs [32]. Thus, the first contact by adult ticks, compared to nymphs, on human hosts should take place further up the human body, closer to the torso/dorsum area and head/neck area. Tick-stage related “preferences” for site of attachment have been observed on other vertebrate hosts. On the white-tailed deer, Odocoileus virginianus, the adult I. scapularis feed mainly on the anterior dorsal body regions: 87% of adult ticks attached to the ears, head, neck and brisket [33]. These feeding sites by adults of I. scapularis correspond arbitrarily to the feeding sites selected by adult females of I. ricinus on humans (i.e. torso/dorsum and head/neck areas), found in the present study. On horses, attachment by adult female I. scapularis was largely restricted to the under-body areas, which was considered to reflect avoidance of direct sunlight by the ticks [33]. On the European roe deer, larvae, nymphs and adult females of I. ricinus show high degrees of intrastadial spatial aggregation [34]: larvae aggregate mainly to the forelegs and to the head of roe deer, nymphs aggregate mainly to the head, and adult females aggregate mainly to the neck of roe deer. Stage-specific degrees of tolerance of desiccation may be one among factors, which explain how stage-specific “preferences” for attachment sites have evolved. However, the host’s grooming behaviour and capacity to remove ectoparasites from particular parts of the host’s body should have a great effect on the evolution of feeding sites “preferred” by the ectoparasites.

The site of tick attachment was not influenced by the age of the bitten person. However, women, compared to men, removed a greater proportion of ticks from the head and neck area. Berglund and co-workers [9] found that LB patients bitten on the head or neck more often presented neurologic manifestations compared to LB patients bitten on other parts of the body. This suggests that the site of tick attachment on the skin of the human host may be of particular clinical significance. We also found that men, compared to women, removed a greater proportion of ticks from the groin/genital area. Similar results were recorded in the study of Berglund et al. [9]. Consequently, this body region seems to be a “preferred site” for blood-seeking ticks, presumably since here they should be relatively well protected from sunlight, desiccation and host grooming activity.

Duration of tick attachment

Among adult female ticks and nymphs, 63% were removed later than 24 hours of attachment. When the calculated duration of tick-feeding (based on scutal and coxal indices) were compared to the participants’ self-estimated durations of tick attachment, we found that the tick-bitten persons usually underestimated the duration of tick attachment. A person who removes an attached tick from the skin later than 24 hours of tick attachment is more likely to develop localised and systemic symptoms of tick-borne diseases compared to if the tick is removed earlier [15]. Therefore, it is important to find and remove any tick from the skin as early as possible. For instance, even if the virions of TBEV can be transmitted within 1 hour after tick attachment [12], it is advisable to remove any tick as quickly as possible since the amount of virus particles in the tick salivary gland seems to increase with duration of tick feeding [35]. It is reasonable to assume that the more time a TBEV-infected tick feeds, the higher will be the virus dose transmitted to the host.

The location of the attachment site seemed to influence how soon a tick was detected and, therefore, the duration of tick attachment. Ticks attached to the groin/genital area or to the head/neck area were apparently more difficult to detect than ticks attached to other sites. We did not find any significant differences between adult female ticks and nymphs regarding the time from attachment until the tick was detected and removed. We expected that the larger, adult female ticks would be detected sooner than the smaller and more inconspicuous nymphs. This was reported in a study on I. ricinus feeding on humans in Switzerland [36]. The discrepancy between that study and ours, regarding the duration of tick attachment until the nymphs or the adult female ticks were removed, could be due to differences in the participants’ awareness as well as people’s knowledge about tick-infested habitats.

In the present study, the majority of ticks were removed 24–48 hours after the beginning of attachment. Only 5% of the adult female ticks and nymphs were removed after ≥ 72 hours. In transmission experiments using rodents, a high level of Borrelia transmission is reached after ≥ 72 hours of tick attachment [13, 14, 37]. Several studies have shown that few people (6.2-9.0%) become infected with Borrelia when they are bitten by a Borrelia-infected tick [18, 38, 39]. One explanation for the low risk of contracting a Borrelia infection could be that few ticks (only 5%, in our study) are still attached to the skin after 72 hours.

Older people are likely to have poorer vision and impaired physical sensitivity compared to younger people. This may explain why we found that older participants, compared to younger participants, detected the attached nymphs after a longer attachment time. Similar observations on I. scapularis nymphs removed from humans were reported by Falco and co-workers [10]. Moreover, we found that men, compared to women, usually detected any attached tick after a longer duration of tick attachment; this suggests a higher risk of Borrelia-transmission to men if they are bitten by Borrelia-infected ticks.