After removing duplicates, we identified a total of 11,370 records across the three databases, of which 78.6% were marked as not relevant after employing the ASReview tool. In the full-text screening phase, 442 papers were assessed for eligibility, and ultimately, 25 papers were included in the review (Fig. 1).

Table 1 and Table S3 present the general characteristics of the included studies. All studies were published from 2017, with the majority from 2020 (n = 16, 64.0%). The cross-sectional study design was the most adopted, investigating characteristics of exposures and outcomes of interest simultaneously (n = 21, 84.0%). Four studies (16.0%) employed a longitudinal design, investigating migrants’ characteristics during the transit phase and after their resettlement in destination countries [31,32,33,34]. In most cases (n = 19, 76.0%) researchers collected quantitative information. Only one study was qualitative (4.0%) [35], while mixed methods research was employed in the remaining five studies (20.0%).

The majority of the studies employed population-based samples, enrolling all individuals residing in camps and dedicated facilities during the study periods, while in other cases (n = 9, 36.0%) the investigators employed opportunistic samples. One study (4.0%) used secondary data, relying on information from medical records and health documentation [36]. The target population participated in the studies mainly while hosted in refugee camps/asylum centers (n = 15, 60.0%), or in facilities where migrants admitted for resettlement in Europe attended mandatory pre-departure educational activities (n = 4, 16.0%).

We observed that Higher Education Institutes (HEIs) were the most frequently involved organizations (n = 6, 24.0%) in the implementation of the studies. In other instances, HEIs were responsible for the studies with the support of IOM (n = 4, 16.0%). Médecins sans Frontierès (MSF) was the non-governmental organization (NGO) most frequently involved (present in four out of six studies carried out by NGOs). Public Health Institutes (such as the Swiss Tropical and Public Health Institute, the Finnish Institute for Health, the National Institute for Health, Migration and Poverty) were responsible for four (16.0%) of the included studies. In four studies (16.0%), multiple organisations were involved and collaborated in the study process.

In the included studies, data were collected over limited time periods, with mean durations of seven months. The data collection period ranged from a minimum of 15 days to a maximum of 26 months. In most instances, data collection was carried out by the study researchers with the assistance of cultural mediators (n = 7, 28.0%), healthcare workers with mediators (n = 6, 24.0%), or by researchers or healthcare workers alone (n = 10, 40.0%). The healthcare workers involved included physicians, nurses, psychologists, and others.

Table 1, Table S3 and Table S4 contain information regarding the populations’ characteristics. The sample sizes were relatively small, ranging from 30 to 2484, with a mean of 463 subjects. The distribution of males and females was homogeneous. Overall, 14 studies (56.0%) were conducted on young adults only, while 11 (44.0%) included also accompanied or unaccompanied minors. Among these, three studies (12.0%) were conducted on minors only, comprising a population aged between 6 and 17 years old [37,38,39]. Almost all studies defined their population as either refugees or asylum seekers. Although these terms are distinct, they were used interchangeably by study researchers to refer to study participants regardless of their legal status at the time of the research.

Figure 2 presents a comprehensive overview of the study locations and the main countries of origin of the included populations. Most of the studies involved populations originating from Syria (n = 11, 44.0%). Ten studies (40.0%) included populations from the African continent, especially Western Africa (Gambia, Senegal, Nigeria, Guinea, Mali). In other studies, the main country of origin was Afghanistan (n = 3, 12.0%), while one study included populations mainly originating from Middle Eastern or North African countries [40].

Study locations, main countries of origin and migratory routes towards Europe. The black arrow represents four longitudinal studies in which the initial phases were conducted in Lebanon, and follow-up was conducted in Norway

Regarding the location of the studies, the majority were conducted in Italy and Serbia (n = 7, 28.0% and n = 6, 24.0%, respectively), followed by Norway (n = 4, 16.0%). The remaining studies were conducted in Greece (n = 3, 12.0%), France (n = 2, 8.0%), Switzerland (n = 1, 4.0%), Bosnia-Herzegovina (n = 1, 4.0%) and Finland (n = 1, 4.0%). Not all studies provided information on the migratory routes used by migrants but based on countries/areas of origin of included populations and study locations we can infer two primary paths: the Central Mediterranean route, to reach Italy, France and Switzerland coming from Western, Northern and Eastern Africa, and the Balkan route, with migrants originating mainly from Syria, Afghanistan and Iraq. It is noteworthy that we did not identify any study on migrants travelling via the Western Mediterranean route.

The migration journey experience was divided into two categories: journey characteristics (n = 24 studies, 96.0%) and specific risk factors experienced during travel (n = 21 studies, 84.0%) (Table S3). A total of 19 studies (76.0%) reported information on the time since arrival in the country where the study was conducted, expressed in median/mean number of days/months/years. In the included studies, migrants had been in those countries for a few days to several months, and in some cases, for more than a year. Two studies (8.0%) stated that migrants were newly arrived without providing further details [40, 41]. A total of five studies (20.0%) provided detailed information on the reasons behind migration, which was predominantly driven by factors such as war and conflict, the pursuit of international protection, insecurity, political instability and persecution (Table S3). The duration of travel was reported in 19 studies (76.0%), either as the median time of travel duration (and I-III quartile) or using specific categories. In the included studies, the journeys lasted from one to 60 months (see Additional file 2). Only eight studies (32.0%) specified the journey type, indicating whether migration occurred by sea, air, or land. Additionally, data on the presence of a transit country, the number of transit countries, the median duration of stay in transit countries, and the transit countries crossed were obtained from 11 out of 25 studies (44.0%) (Table S3).

Specific risk factors pertaining to the transit phase included, among others, violent events and trauma experiences, potentially traumatic events, and difficulties encountered during the journey. These risk factors were presented in the included studies either by considering the total number of events experienced by the subjects or by providing detailed information regarding the type of events (torture, threats, sexual or physical violence, robbery, incarceration, forced labor and others). While journey characteristics were consistently assessed through specifically developed questions, validated questionnaires were utilized for evaluating the risk factors experienced during travel. These instruments were either specifically designed for journey experiences or adapted from more generic questionnaires on trauma [42] (Table S3).

Most of the included studies focused on mental health (n = 15, 60.0%) as the main health outcome of interest, investigating the presence of anxiety and depression disorders, post-traumatic stress disorder (PTSD) and other psychological symptoms. Studies on physical health (n = 4, 16.0%) examined the presence of non-communicable diseases (NCDs) and infectious diseases. One study was exclusively focused on perceived health, including well-being and perceived quality of life in participants (n = 1, 4.0%) [34]. Five studies (20.0%) addressed multiple health outcomes (Fig. 3a). The following heatmap (Fig. 3a) describes all potential combinations of exposures and outcomes investigated in the included papers. Exposures were categorised into specific risk factors experienced during the journey and journey characteristics. Health outcomes were classified into mental health, including anxiety, depression, other psychological symptoms and PTSD symptoms, non-communicable diseases, infectious diseases, well-being and quality of life.

Heatmap for combinations of investigated exposures and outcomes (a) and associations between investigated exposures and outcomes (b)

For each cell, we reported the total number of studies providing information for that combination of exposure and outcome. The colour intensity in each cell reflects the number of studies, indicating that the majority of the available literature focused on mental health problems in combination with the total number of traumatic events experienced, the duration of the journey, and the time since arrival in the countries where the studies were conducted. As previously stated, only 14 out of 25 studies (56.0%) formally assessed the associations between exposures and health outcomes [31,32,33,34, 42,43,44,45,46,47,48,49,50,51]. The associations evaluated in these studies are presented in the heatmap (Fig. 3b). Most of these studies focused on the relationship between mental health problems and the total number of traumatic events, showing that increased exposure to various types of traumatic experiences during the journey was associated with higher risks of experiencing anxiety, depression and PTSD [31, 42, 44, 51]. Furthermore, a higher risk of mental health problems was positively linked to the length of the journey and the time since arrival [45, 46, 49, 50]. In addition, a few studies also examined the positive association between exposure to sexual and physical violence, the selection of specific journey routes, and the occurrence of infectious diseases [43, 47].

With regard to the characteristics evaluated in the risk of bias assessment (Table S5), it was observed that nearly all included studies adequately defined their inclusion criteria and described subjects and study settings. Furthermore, all studies employed valid methodologies for measuring exposure and outcomes, utilising validated questionnaires or surveys specifically developed for the study’s purposes. However, it is noteworthy that while all studies identified possible confounding factors, some of them (n = 4, 16.0%) did not employ specific strategies to address these confounders in the statistical analysis. This may be attributed to the limited sample sizes available in the included studies.