Health status of female Moldovan immigrants to Italy by health literacy level and age group

Background: Migration �ows from Eastern Europe to Italy have been large and continue to grow. The purpose of this study was to examine the health status of a population of Moldovan immigrant women, and their access to health care services in northern Italy, by age group and health literacy level. Methods: We administered an ad-hoc questionnaire to adult Moldovan women. A bivariate analysis was conducted to test the association between health literacy and age groups with other variables ( lifestyles, symptoms and diseases, access to health services). A stepwise logistic regression analysis was run to test the association between access to primary care and health literacy. Moreover, the study compare Moldovan women data with a sample of Italian women of the same age range living in North-Eastern region. Results: Our sample included 170 Moldovan women (aged 46.5 ± 12.3) in �ve occupational categories: home care workers (28.2%); cleaners (27.1%); health care workers (5.9%); other occupations (28.8%); and unemployed (10%). Active smokers were twice as prevalent among the women with a low health literacy. Health literacy level also determined access to primary healthcare services. For all age groups, the Moldovan sample reported a higher prevalence of allergies, lumbar disorders and depression than the Italian controls.


Background
International migration is one of the most predominant issues of our times.The migration of workers from Eastern to Southern Europe began in the late 1980s, after the dissolution of the former communist regimes, and increased enormously in subsequent decades (1,2).In Europe as a whole, many migrants from Eastern Europe have satis ed the increasing demand for a cheap labor force in low-skill and highstrain jobs.To give an example, only 6% of the domestic workers registered with the Italian National Social Security Institute in the early 1980s were of migrant origin, whilst the gure had reached 72% by 2006 (3,4).Moldovan migration, in particular, is highly feminized, and the ows from Ukraine and the Republic of Moldova largely consist of women who are independent migrants (5)(6)(7).
Worldwide, it has been reported that migrant workers in certain living and economic conditions may not seek health care -supposedly due to a limited knowledge of their rights or to cultural or language barriers (8).This seems to be particularly true in the case of work-related diseases or injuries (9), and even applies to migrant women with a good formal education (5,6).It has also been established internationally that, for people to use health care services appropriately, they need to be able to access and understand healthrelated information, as captured in the concept of "health literacy".This type of competence depends directly on a given individual's language, education and culture.It is often described as an individual's capacity both to process health-related information and actively choose a healthy behavior, and to actively and appropriately interact with health care services (10,11).
Moldova's worker migration ow is reportedly one of the highest in the world, with approximately 25% of the country's economically active population in 2010 returning from working abroad, still working abroad, or intending to work abroad, according to the "Moldovan Labor Force Survey" (12).Studies on the health status of these people and in particular of female migrants , and on their use of health care services in their destination countries, are lacking, however.Recognizing this target population's health problems and ascertaining their inclination to use public health services could be useful in order to establish an effective program of health promotion and disease prevention for this particular migrant group.
Hence this descriptive study to examine the health status of a sample of Moldovan immigrant women, and their recourse to health care services in Padova, a province in north-eastern Italy, by age group and health literacy level.

Context
The Italian National Health System (NHS) is a mainly public system nanced by general taxation.All residents registered with the country's NHS (be they Italians or regular immigrants) can access all healthcare services free of charge or by paying a small fee, and they are assigned a general practitioner (GP) of their choice.Italy is one of ve of the 27 EU Member States to freely provide immigrants with much the same range of services as Italian nationals (13).

Moldovan migration to Italy
The Republic of Moldova is not a member of the European Union, so its citizens cannot move freely within the EU.They need an appropriate document to cross its borders.Nonetheless, Moldovan migration to Italy has been quite considerable, and has increased in recent times.As at January 1st, 2019, there were 128,979 Moldovan citizens legally residing in Italy, and 66% of them (85,431) were women.In the province of Padua (in north-eastern Italy), their prevalence is akin to the national one, with 63% of women out of a total of 9,866 Moldovan immigrants.These Moldovans now account for 1% of the resident population (937,908) of the province of Padua (13).

Sample
Our study population consisted of adult Moldovan women who understood Italian, recruited from February to June 2019 by a convenience sampling, since it was impossible to carry out a random sampling on Moldovan women residing in Padua because of privacy law.Questionnaires were administered face to face in the Italian language at various times and at four different venues in Padua, following the time-space method for sampling hard-to-reach populations (14,15), though we did not randomize the selection of venues and times.The interviewer administered questionnaires at: the Moldovan Consulate in the mornings from Tuesday to Friday (opening days), holding interview sessions for the same number of times on each day of the week; at two local Moldovan Orthodox Churches on Sundays, holding interview sessions for the same number of times at each church; at the parking lot usually attended by Moldovans because it is from here that vehicles providing a transport service depart for Moldova on Saturdays; and at the park where Moldovan women tend to meet on Saturdays when the weather is ne.Verbal informed consent was obtained from all participants before their enrolment.
The trained researcher started interviewing 205 women, and the interview was completed with 173 of them.In this article, we analyze a sample of 170 Moldovan women living legally in Italy, having excluded 3 women without documents.At each interviewing session (37 in all), an average of four women refused to participate.Their refusal were mainly due to shortage of time -because they were going to an appointment with the consular o cer, for instance, or because they would be late for Mass.

Questionnaire
The tool used for this study was an ad hoc questionnaire administered to our sample population.It consisted of 55 multiple-choice or open-answer questions administered by a trained interviewer, covering the following domains: socio-demographic factors; lifestyles; perceived health symptoms; self-reported diagnoses of certain speci c diseases; and recourse to public health services.The questionnaires were anonymized for the purposes of our analysis.
The variables measured for each domain were as follows: 1. Socio-demographic factors: Age, expressed in years, and grouped as: "20-34", "35-44", "45-54", "55-64" and ">65" years; Schooling, expressed in years of school attended; Years living in Italy; Employment, classi ed by risks of exposure for the purposes of this study in 5 groups of occupations as follows: home care-workers (live-in and live-out paid caregivers); cleaners (live-in and live-out domestic workers or industrial cleaners); health care workers (nurses and social care operators); or other occupations (all forms of employment other than those mentioned above); and unemployed; Health literacy, measured with the "Single Item Literacy Screener" (SILS), a single-item questionnaire used to identify any impairment in adults' understanding of health material: "How often do you need to have someone help you when you read instructions, pamphlets or other written material from your doctor or pharmacist?" with 5 possible answers (1.Never; 2. Rarely; 3. Sometimes; 4. Often, 5. Always).Scores higher than 2 indicate some di culty in understanding the meaning of printed health-related material (16).Respondents' scores were used to divide our sample into groups with a high health literacy (SILS questionnaire scores of 1 or 2) or low health literacy (for scores of 3, 4 or 5).A previous study judged that the Italian version of the SILS -as an indicator of limited reading and understanding ability regarding health information -is a better tool for measuring HL than more complex functional HL measurement instruments (17).

Lifestyle factors:
Smoking habit, regardless of the number of cigarettes smoked a day.Ex-smokers were classi ed as non-smokers for the purposes of this study; Drinking habit, expressed as the mean daily alcohol intake measured in units of alcohol (UA), where 1 UA = 12 g ethanol, i.e. approximately 250 ml beer, 75 ml wine or 25 ml spirits; Sport or physical exercise in free time in the previous 12 months: rated as "Never", "Less than once a week", or "At least once a week"; Weight (kg) and height (cm), from which we calculated the respondent's Body Mass Index (BMI, kg/m 2 ), then divided our sample into BMI categories: "Underweight" (BMI less than 18.5); "Normal weight" (BMI between 18.5 and 24.9); "Overweight" (BMI between 25 and 29.9); and "Obese" (BMI over 30).

Items concerning self-reported diagnoses were drawn from the Italian version of the "European
Health Interview Survey" (EHIS, Eurostat) adopted by the Italian National Institute of Statistics (ISTAT) (18,19).The assessment concerned whether any of the following diseases had been diagnosed by the respondent's doctor: asthma, allergies, bronchitis, myocardial infarction, coronary diseases, hypertension, diabetes, lumbar and cervical disorders, arthritis/arthrosis, depression and anxiety.
4. Items regarding perceived symptoms were extracted from the "Health and Work Performance Questionnaire" (HWPQ) (17).Speci c symptoms were chosen because they are often associated with chronic stress (20) which is reportedly a risk to migrant care workers' mental health (21).The questions concerned whether respondents had suffered any of the following symptoms in the previous 6 months: headache, trouble sleeping, fatigue, lack of appetite, di culty concentrating, gastro-intestinal disturbances, dizziness, shortness of breath, and di culty relaxing.Respondents answered on a 6-point scale ranging from "Never" (0) to "Daily", as reported elsewhere (6) (22).Then they were pooled into three groups ("Never", "Sometimes" and "Daily") for our analysis.
5. To assess aspects of health care, we rst inquired whether respondents were covered by free medical insurance and social security bene ts, and whether they had their own, trusted doctor in Italy or in Moldova.To examine their access to health care services, our questionnaire included other elements drawn from the EHIS (23.The questions concerned whether respondents had: "seen a GP in the previous four weeks"; "been examined by a specialist in the previous four weeks"; "gone to an Emergency Department in the previous 12 months"; and/or "been hospitalized in the previous 12 months".Women in the appropriate age groups were also asked if they had undergone mammography and PAP smear/HPV testing at least once in their life (Italian NHS prevention programs recommend mammography screening for women aged 50 to 74, and HPV screening [HPV test or PAP smear] for women aged 25-64).
Regarding the completeness of the data, there were no missing values for the occupational category variable, but for the items regarding the health literacy measure and age group there were 3 (1.8%) and 6 (3.5%) missing values, respectively.We did not apply any imputation techniques, given the low proportions of cases with missing data.

Data for comparison
The results of the "European Health Interview Survey 2015" (EHIS) (18) questionnaire administered to a sample of Italian women (n=1827) living in the country's north-eastern region, and aged between 20 and 74, were compared with the results of the present study on Moldavan immigrants.

Statistical analysis
A descriptive analysis was conducted.We calculated the means and standard deviations for quantitative variables, and the relative and absolute frequencies for categorical variables.Although a convenience time-space method for sampling hard-to-reach populations was applied, we used statistical inference, assuming that this sampling method ensures that the sample is to some degree representative of the entire population.In particular, a chi-squared test was used to test the difference in the distribution of a categorical variable, but Fisher's exact test was applied when the expected frequencies were <5.Differences in quantitative measures by group were tested with ANOVA.A backward stepwise logistic regression model with an exit probability of 0.10 was used to test whether access to primary care (the dependent variable was the item "Had you seen a GP in the previous four weeks?Yes/No") was associated with health literacy, adjusting for sociodemographic variables, health behavior variables and previously-diagnosed diseases.The "R: A language and environment for statistical computing" (R Foundation for Statistical Computing, Vienna, Austria) was used for the analysis (24).

Results
Table 1 shows the sample's characteristics.They ranged in age from 21 to 69 years (mean 46.5; SD 12.3).They had attended school for a mean 12.7 years (SD 3.7).The sample's distribution in the different job categories revealed that 28.2% were home care workers, 27.1% were cleaners, 5.9% were health care workers, 28.8% had other occupations, and 10% were unemployed.Table 2 shows the bivariate analysis conducted on the variables considered in the questionnaire by age group.There was a high prevalence of allergies and lumbar disorders at all ages, and a high prevalence of depression clustered in the intermediate age groups.Older age was associated with a higher BMI, and with an increase in the reported prevalence of some diseases, such as hypertension, arthritis/arthrosis, cervical disorders, and diabetes.Table 3 shows the distribution of respondents' lifestyles and usage of health care services by their health literacy level.The prevalence of active smokers in the low health literacy group was more than twice as high as in the high health literacy group.There was also a signi cant association between high and low levels of health literacy and respondents' occupations (p 0.016).In particular, nearly half of respondents with low levels of health literacy were employed as cleaners.There was also evidence of the group with a high health literacy making more visits to GPs (see Fig 1).The stepwise logistic regression showed that the odd of access to primary care (in the previous 4 weeks) was increased with higher BMI (OR 1.1, 95%C.I. 1.0-1.2) and lumbar disease (OR 2.3 95%C.I. 1.1-4.7),instead was reduced, approaching statistical signi cance, in case of low levels of health literacy (OR 0.4 95%C.I. 0.2-1.07)(data not shown).Table 4 shows the gures for the prevalence of various diseases in the sample of 20-to 74-year-old Italian women living in the north-east of the country.When compared with the self-reported prevalence of diseases in the sample of Moldavan women (Table 2), there emerged a higher prevalence of all the diseases considered in the Moldovan women.This was true for all age groups, except in the case of diabetes, for which the Moldovan women reported a higher prevalence only among the 45-54, 55-64 and over 65-year-olds.

Discussion
As expected, this study showed a higher prevalence of several diseases, such as hypertension and diabetes in older age groups.More interestingly, almost all the illnesses considered showed a higher overall prevalence among the Moldovan immigrant women than among the Italian controls.The former made more use of health care services than the latter too.An association also emerged between health literacy level and both lifestyle and recourse to health care services.
Our sample of Moldovan immigrant women included a sizable proportion who were overweight (30.5%), or obese (17.7%).These gures are higher than the 22.1% for overweight and 10.1% for obesity among Italian women, but only half the percentages for Moldovan women in their home country, where 60.1% are overweight and 31% are obese (25).Another nding concerns the clustering of higher BMIs in the older age groups, while the younger immigrant women had lower BMIs.This may be a sign of an adaptive effect, with the immigrants' lifestyles, such as their dietary habits, approaching those of their adopted country.Another possible interpretation of this phenomenon, however, is that women lose weight after migrating because of the hard living and working conditions they nd in their adopted country.
The answers to our questionnaire indicate that less than one in ve Moldavan immigrant women are smokers -a proportion almost in line with the Italian reference gure (21.3%) -and the prevalence of smokers was similar in all age groups.The Moldovan women's reported alcohol consumption was moderate-to-high (nearly 1.5 UA a day), and one in three of our respondents exceeded the recommended limit for women (1 UA a day).In its "Global Alcohol Report" for the Republic of Moldova, the WHO indicated that alcohol consumption by women over 15 years old averaged 2 units/day (26).Our results suggest that our Moldovan immigrant women drink less than their counterparts at home, and slightly more than Italian women in the same age range, whose average alcohol consumption is 1 UA a day (27).
Concerning physical exercise, half of our sample reported engaging in some form of physical exercise in their free time, but one in three said they never did so.This level of sedentariness is higher than reported by Moldovans in their own country, which is 24.5% for adults generally.
As for the overall health status of our sample, there was a noticeably higher prevalence of several diseases compared with the Italian reference values.We identi ed a more than twofold self-reported prevalence of hypertension, arthritis/arthrosis, cervical disorders, diabetes, and allergies, and a threefold prevalence of lumbar disorders, depression and anxiety.If we look at the reported prevalence of allergies (36% for the Moldovan group versus 14.8% for the Italian controls), this may re ect the numerous studies in the literature indicating that migration to a highly-industrialized country favors the development of respiratory allergies in immigrants (28,29).As regards lumbar and cervical disorders, back pain has been found directly related to mental health disorders and stress in fact stress could contribute to the onset or the persistence of chronic pain (30).Another plausible explanation for these conditions is work-related, given the large proportion of our respondents who were home care workers and cleaners (jobs that involve the manual lifting of sometimes heavy loads).Analyzing our women by age group, the ratio for the prevalence of Moldovan and Italian women with lumbar disorders declines linearly from 6.3 for the younger women to 1.7 for the older age groups.The same trend could be seen for hypertension, for which the ratio went from 10 for the younger women to 1.3 for the older age groups.
Depression was reported by more than 10% of our Moldovan sample, with a slightly higher prevalence in the intermediate age groups.This is three times higher than the prevalence of 4.3% reported by a sample of 1827 Italian women living in the north-east of the country (19).This difference is more evident among the younger age groups, the prevalence ratio being 7.2 in the youngest age group and dropping gradually to 1.2 for the older women.When we investigated the issue of anxiety, the prevalence of this condition was a remarkable ve times higher in the Moldovan women aged 45-54 and 55-64 than in their Italian counterparts, as opposed to a twofold prevalence in the other age groups.Analyzing symptoms usually associated with anxiety, depression and burnout (21) we found quite a high overall prevalence of daily headache, trouble sleeping, and extreme fatigue, possibly as a direct consequence of underlying stress.
These symptoms were distributed throughout our Moldovan sample, with no signi cant differences between the various age groups considered.This could be also explained by high prevalence of chronic pain as described above, in fact chronic pain could be emotionally stressful (30).Chronic pain in fact is known to change the levels of stress hormones and these can affect your mood, thinking and behavior.
Moreover, chronic pain can affect ability to function at home or work making also di cult to participate in social activities and hobbies, which could lead to decreased self-esteem.In addition, chronic pain could provide sleep disturbances, fatigue, trouble concentrating, decreased appetite.These negative changes can dampen overall mood; and this can result in depression and anxiety.In addition, vulnerability to stress has already been described in migratory groups, especially for Eastern European citizens migrating westwards (31).The stress of migration per se can lead to depression and anxiety (32) or somatization (33) which are frequently underestimated.Such conditions of malaise can also be carried to the migrants' home countries when they return.In fact, increasing attention is being paid to what has been called the "Italy syndrome", which is a sort of psycho-social distress suffered by Eastern European migrant women (34).The scienti c literature on this phenomenon is quite limited, while many newspaper investigations discuss it.Cozzi (35) traces the genealogy of the term showing that it was invented by two Ukrainian psychiatrists, Andriy Kiselyov and Anatoliy Faifrych, who identi ed a speci c medical case affecting women returning from Italy: bad mood, sadness, weight loss, loss of appetite, insomnia, tiredness, loss of motherliness, and split identity.
When questioned about their recourse to health care services, our sample of Moldovan women of all ages reported a large number of visits to GPs and specialists.This can be interpreted as a sign of their integration, and proof of the good functioning of the Italian NHS.The proportions of women reportedly seeing a GP or a specialist in the previous month, or being hospitalized in the previous year were 47.6%, 37.2% and 12.2%, respectively.These gures are much higher than those of our Italian controls, which were 35.6%, 22.2% and 6.7%, respectively.
The proportion of Moldovan immigrant women of suitable screening age who reporting having undergone HPV testing or a PAP smear at least once in their life was much the same as for their Italian

Figures
Figures

Table 2 .
Results of bivariate analysis: distribution of different questionnaire variables by age group

Table 3 .
Results of bivariate analysis: distribution of different questionnaire variables by level of health literacy

Table 4
Results of bivariate analysis: 95% confidence intervals (CI) for different "European Health Interview Survey 2015" questionnaire variables in north-eastern Italian women aged 20-74, by age group