This paper is in the following e-collection/theme issue:

Equity and Digital Divide (216) Demographics of Users, Social & Digital Divide (737) eHealth Literacy / Digital Literacy (389) Participatory Medicine & E-Patients (770)

Published on in Vol 5, No 3 (2019): Jul-Sep

Associations Between Immigration-Related User Factors and eHealth Activities for Self-Care: Case of First-Generation Immigrants From Pakistan in the Oslo Area, Norway

Associations Between Immigration-Related User Factors and eHealth Activities for Self-Care: Case of First-Generation Immigrants From Pakistan in the Oslo Area, Norway

Associations Between Immigration-Related User Factors and eHealth Activities for Self-Care: Case of First-Generation Immigrants From Pakistan in the Oslo Area, Norway

Authors of this article:

Naoe Tatara1 Author Orcid Image ;   Hugo Lewi Hammer1 Author Orcid Image ;   Jelena Mirkovic2 Author Orcid Image ;   Marte Karoline Råberg Kjøllesdal3, 4 Author Orcid Image ;   Hege Kristin Andreassen5, 6 Author Orcid Image
Article Authors Cited by (3) Tweetations (2) Metrics

    Original Paper

    1Department of Computer Science, Faculty of Technology, Art and Design, Oslo Metropolitan University, Oslo, Norway

    2Center for Shared Decision Making and Collaborative Care Research, Oslo University Hospital, Oslo, Norway

    3Department of Community Medicine and Global Health, Institute of Health and Society, Faculty of Medicine, University of Oslo, Oslo, Norway

    4Health Services Research, Norwegian Institute of Public Health, Oslo, Norway

    5Centre for Women's and Gender Research, UiT The Arctic University of Norway, Tromsø, Norway

    6Centre for Care Research, Norwegian University of Science and Technology, Gjøvik, Norway

    Corresponding Author:

    Naoe Tatara, PhD

    Department of Computer Science

    Faculty of Technology, Art and Design

    Oslo Metropolitan University

    Postboks 4

    St. Olavs plass

    Oslo, 0130

    Norway

    Phone: 47 67 23 86 79

    Fax:47 22 45 32 05

    Email: naoe.tatara@oslomet.no


    Background: Immigrant populations are often disproportionally affected by chronic diseases, such as type 2 diabetes mellitus (T2DM). Use of information and communication technology (ICT) is one promising approach for better self-care of T2DM to mitigate the social health inequalities, if designed for a wider population. However, knowledge is scarce about immigrant populations’ diverse electronic health (eHealth) activities for self-care, especially in European countries.

    Objective: With a target group of first-generation immigrants from Pakistan in the Oslo area, Norway, we aimed to understand their diverse eHealth activities for T2DM self-care in relation to immigration-related user factors specific to this target group: proficiency in relevant languages (Urdu, Norwegian, English), length of residence in Norway, and diagnosis of T2DM compared with general user factors (age, gender, education and digital skills, and self-rated health status).

    Methods: Data were from a survey among the target population (N=176) conducted in 2015-2016. Using logistic regression, we analyzed associations between user factors and experiences of each of the following eHealth activities for T2DM self-care in the last 12 months: first, information seeking by (1) search engines and (2) Web portals or email subscriptions; second, communication and consultation (1) by closed conversation with a few acquaintances using ICT and (2) on social network services; and third, active decision making by using apps for (1) tracking health information and (2) self-assessment of health status. Using Poisson regression, we also assessed the relationship between user factors and variety of eHealth activities experienced. The Bonferroni correction was used to address the multiple testing problem.

    Results: Regression analyses yielded the following significantly positive associations: between Urdu literacy and (1) information seeking by Web portals or email subscriptions (odds ratio [OR] 2.155, 95% CI 1.388-3.344), (2) communication and consultation on social network services (OR 5.697, 95% CI 2.487-13.053), and (3) variety (estimate=0.350, 95% CI 0.148-0.552); between length of residence in Norway and (1) communication and consultation by closed conversation with a few acquaintances using ICT (OR 1.728, 95% CI 1.193-2.503), (2) communication and consultation on social network services (OR 2.098, 95% CI 1.265-3.480), and (3) variety (estimate=0.270, 95% CI 0.117-0.424); between Norwegian language proficiency and active decision making by using apps for self-assessment of health status (OR 2.285, 95% CI 1.294-4.036); between education and digital skills and active decision making by using apps for tracking health information (OR 3.930, 95% CI 1.627-9.492); and between being a female and communication and consultation by closed conversation with a few acquaintances using ICT (OR 2.883, 95% CI 1.335-6.227).

    Conclusions: This study implies immigration-related factors may confound associations between general user factors and eHealth activities. Further studies are needed to explore the influence of immigration-related user factors for eHealth activities in other immigrant groups and countries.

    International Registered Report: RR2-DOI 10.2196/resprot.5468

    JMIR Public Health Surveill 2019;5(3):e11998

    doi:10.2196/11998

    Keywords

    immigrants; type 2 diabetes; self-care; information-seeking behavior; literacy; language 


    Chronic Diseases and eHealth Use for Self-Care Among Immigrant Populations

    Immigration surge made many European nations linguistically and culturally more diverse than ever, which poses a challenge to the health systems [1]. Immigrant populations are more likely to have challenges related to health care access due to language barriers and cultural differences [2-9]. Belonging to an ethnic minority as well as having low literacy in the official language of the host country are shown to be associated with poor health [10,11]. Furthermore, many immigrant groups have a lifelong exposure to low socioeconomic position [12], which leaves them at high risk of chronic diseases [13]. Different studies show that immigrant populations are often disproportionally affected by chronic diseases, such as type 2 diabetes mellitus (T2DM) [7,14-17]. The World Health Organization Regional Office for Europe recognized the importance to address these challenges, and advocated in its health policy framework “Health 2020” [18] that eHealth should be successfully implemented and utilized to reduce social health inequalities [19].

    However, eHealth content is often text-based and designed for the nation’s majority population, thus enhancing social health inequalities rather than reducing them [11,20-24]. Designing eHealth services also aiming to include minority user populations requires sensitivity to their language and cultural preferences [25]. For immigrant populations, recent studies show less extended use of eHealth for specific purposes within self-care than among host populations [26-31]. However, knowledge is limited and mainly from the United States regarding the association between eHealth use and user factors that are especially relevant to immigrants, such as language proficiency [28,30,32,33] or the length of residence in the host country [34]. Both measures are frequently examined in relation to immigrants’ health behavior, health status, and access to and utilization of health care services [4,5,35-42]. Given that eHealth use can be considered a part of health behavior and use of available health care resources, the association between eHealth use for self-care and host language proficiency versus proficiency in their primary language or length of residence would be worth investigating.

    In this paper, we discuss how this can be done in an explorative study identifying variations in the use of eHealth for self-care of T2DM among first-generation immigrants from Pakistan in the Oslo area, Norway. Self-care is defined by the World Health Organization as “keeping health, prevention of and dealing with illness” [43]. Therefore, we target not only those who are given diagnoses of T2DM but also those who are engaged with the prevention of T2DM.

    Pakistani Immigrants in Norway

    At the beginning of 2018, first-generation immigrants comprised 14.1% of the whole population of Norway [44]. Pakistanis are one of the biggest immigrant groups from non-Western countries. The first Pakistani immigrants arrived in Norway in the 1970s. Steadily, new immigrants are coming from Pakistan for family reunification or establishment of a new family with Pakistani immigrants or their descendants in Norway [45]. The majority of this population concentrates in the Oslo area [46]. The completed level of education among the first-generation Pakistani immigrants in Norway varies widely [47].

    The primary language of this population is Urdu. The Urdu language is written in the Urdu alphabet, which is unlike the Roman alphabet [48]. To our best knowledge, there is no statistic regarding Urdu literacy level of first-generation immigrants from Pakistan in Norway. However, data from United Nations Educational, Scientific and Cultural Organization shows the literacy rate in Pakistan among the population aged 15 years and older was 57% in 2014, and there was a considerable gap between males (69%) and females (44%) [49]. Thus, a nonnegligible number of first-generation Pakistani immigrants who migrated to Norway after school age, especially women, may also have a limited literacy level in the Urdu alphabet despite speaking fluently. In addition, our target population often use Roman Urdu—the Roman alphabet used to express Urdu pronunciation [50]. Roman Urdu is especially useful when typing text by a QWERTY keyboard on an information and communication technology (ICT) device.

    Norwegian is the official and dominant language in Norway. According to a survey among first-generation immigrants with different countries of origin by Statistics Norway in 2016, 55% of the Pakistani participants self-reported high or very high Norwegian literacy level, whereas only 12% reported low or very low Norwegian literacy level [51]. However, the same survey shows that the immigrants who came at a later stage of life (after age of 17) are less likely to self-report a high or very high Norwegian literacy level (42%) than those who came to Norway at an earlier stage (age at immigration 0-6 years: 97%; 7-16 years: 89%) [51].

    Data are not found regarding proficiency in other languages among the population. However, given that English is the most used language for eHealth content globally, those with English proficiency may use eHealth for T2DM self-care to a larger extent than those without English proficiency.

    It has been demonstrated that the prevalence of diabetes is considerably higher among Norwegian-Pakistanis (women: 26.4%; men: 20.0%) than among their ethnic Norwegian counterparts (women: 2.7%; men: 6.4%) [14]. The reasons for the high diabetes risk are interrelated factors including both epigenetics and lifestyle; most of the diagnosis cases are T2DM [52]. Many experience significant changes in lifestyle when moving to Norway, including dietary changes [53]. Pakistani immigrants with diabetes have reported advice from Norwegian health care workers to be inadequate and difficult to comply with due to cultural differences [54].

    Objectives

    In this exploratory study, we examine how the immigration-related user factors specific to this target group are associated with their diverse eHealth use for T2DM self-care compared with general user factors that are typically used in similar studies, such as age, gender, education, digital skills, or general health status.

    Relevant immigration-related user factors to this target group include proficiency in the relevant languages (literacy in Urdu as their primary language; proficiency in Norwegian as the host language; having proficiency in other languages, such as English, because it is the most common language used for eHealth content globally), length of residence in Norway, and diagnosis of T2DM. In this paper, we refer to them as “target group-specific user factors.”


    Description of Survey

    This study is part of a larger survey carried out among first-generation Pakistani immigrants in the Oslo area (N=176) in 2015-2016 [50] to explore the use of eHealth in this population. The informants were recruited by purposive sampling with a multirecruitment strategy [55]. The inclusion criteria were (1) immigrated from Pakistan after the age of 18 years, (2) live in the Oslo area, (3) speak Urdu (the official language of Pakistan) as the primary language in their private life, (4) aged between 25 and 59 years, (5) have access to or interest in using ICT tools (PC, tablet, or smartphone) connected to the internet in daily life, and (6) motivated for and capable of activities for self-care of T2DM [50]. The first and third criteria reflect our intention to include those who are more likely than others to have language and cultural barriers to regular health services.

    Ethical approval was given to the project protocol by Norwegian Social Science Data Services in June 2015 (project number: 43549). The details of the survey have been described elsewhere [55].

    Variables

    Independent Variables

    The independent variables can be divided into two categories: target group-specific variables and general variables. The target group-specific variables included literacy in Urdu, language proficiency in Norwegian, proficiency in another language (for each language), length of residence in Norway, and diagnosis of T2DM. The general variables included gender, age, self-assessment of health status, and completed education (both in Pakistan and Norway) in the form of total number of years. We also included frequency of asking for help when using ICT as a measure of digital skills as a part of the general variables.

    We chose to use self-reported language proficiency by following the convention of similar studies of immigrants’ health [3,5,33,56-58]. Confidence in Urdu literacy and in the Norwegian language was assessed using a 5-point Likert scale. One of the survey’s inclusion criteria was speaking Urdu as the primary language in daily life [50]. However, as described in the Introduction, a nonnegligible proportion of participants may not have high-level literacy in Urdu despite speaking in Urdu daily. Also, Roman Urdu is often used for using ICT in general. Thus, for Urdu literacy, we were interested in reading and writing skills in both Urdu alphabet and Roman Urdu. For Norwegian language proficiency, we asked about confidence level in speaking as well as writing and reading. As there was no presumption about our target populations’ proficiency in other languages, the survey included a question asking informants to name any other languages that they had confidence in. Having proficiency in another specific language was based on each named language to this question (dichotomous variable).

    The informants specified the length of their residence in Norway to one of the categories of 0 to 1 year, 1 to 5 years, 5 to 10 years, and more than 10 years. Using ranges of years rather than exact numbers follows the convention of similar studies [3-5,37,40,56].

    Age was represented by a middle year of an age range because the survey used age range instead of actual age to avoid the risk that an informant could be identified by a combination of answers to survey questions.

    Self-assessment of health status was obtained by a multiple-choice question ranging from poor=0 to excellent=5. The frequency of asking for help when using ICT was given by a multiple-choice question with alternatives ranging from never=0 to always=4. This question captured a lack of digital skills. Further details are described elsewhere [55].

    Dependent Variables

    We considered the same dependent variables that were used in our previous study [55]; dichotomous answers to nine questions asking usage of different types of eHealth for T2DM self-care in the last 12 months, and the total number of positive answers. Among the nine questions, three questions received few positive answers. The three questions were asking about having experience of the following: seeking relevant information with T2DM self-care by searching for software programs on PCs or mobile apps that could be used as a look-up tool (8/176, 4.5%), communicating or consulting about T2DM on portals for peer-communication (9/176, 5.1%), and communicating or consulting about T2DM self-care by online consulting to experts in diabetes (1/176, 0.6%). With such few positive answers, reliable statistical analysis was not possible and we decided to do further analyses only for the rest of the dependent variables. First, for seeking T2DM-relevant information (1) by using search engines that require input of search terms or (2) on specific websites or by email subscriptions that can be navigated by only scrolling and clicking. Second, for communicating or consulting about T2DM self-care (1) by using ICT in general for closed conversation with a few specific acquaintances, such as voice, video, or text communication, or (2) via social networking service (SNS). Third, for active decision making on T2DM self-care by using Web or mobile apps (1) for tracking health information, such as diet, physical activity, weight, blood glucose level, and so on, or (2) to assess own health status with regard to T2DM [55].

    Statistical Analyses

    Logistic regressions were used to assess the relationship between each type of eHealth use with the independent variables described previously. For total eHealth usage, which was a count variable, we used Poisson regression.

    To select the best regression models, all possible combinations of including and excluding independent variables were evaluated using the best-fit method in the statistical program R [59]. We used the Akaike information criterion (AIC) as the selection criterion between models [60]. The AIC awards models with a good fit to the data and at the same time punishes models that use many independent variables to avoid overfitting. Thus, the best model will have a trade-off between the fit to the data and the number of variables included in the model.

    Our previous study revealed that the scores for the questions “total years of education in Pakistan and Norway” and “frequency for asking for help when using ICT” were strongly correlated [55]. Therefore, we defined the variable “education and digital skills” as the mean score of these two questions. Before computing the mean, the following two steps were taken. First, to take into account that they were on different scales, the scores of the variables were standardized. Further, the scores of the frequency of asking for help when using ICT were multiplied by −1 because higher knowledge results in less frequency of asking. The internal consistency of the scores for these two questions, using the standardized Cronbach alpha [61], was .70.

    The questions about reading and writing in Roman Urdu and Urdu alphabet all reflect Urdu literacy. The internal consistency of the scores for these four questions, using the standardized Cronbach alpha, was .85. Therefore, we defined the “Urdu literacy” variable as the mean of these variables. Scores for the questions about reading, writing, and speaking the Norwegian language also showed a high internal consistency (Cronbach alpha=.92). Therefore, we also defined the “Norwegian language proficiency” variable as the mean of the three variables.

    To reduce the number of variables to estimate in the statistical model and take into account that time is a continuous variable, the categorical variables for the length of residence (0-1 year, 1-5 years, 5-10 years, and more than 10 years) were transformed to a continuous variable by setting each of the variables to 0.5, 3, 7.5, and 15 years, respectively. If one immigrates at the age of 18 years or older, it is reasonable to expect that the effect of the length of residence tails off with time (eg, one learns more about the Norwegian society the first year after arrival compared to after the tenth year). Therefore, in addition to using the length of residence directly in the regression models, we also evaluated two transformations: the square root and the logarithm of the length of residence. By comparing the resulting models based on the AIC, the logarithm transform resulted in the best models and was used in the analyses.

    A total of seven models were used to draw conclusions. To address the multiple testing problem, a Bonferroni-corrected significance value of 0.05/7=0.00714 was used in this paper.


    Characteristics of the Sample

    The distribution of the informants by the data concerning the general variables and the dependent variables has been described elsewhere [55] and is reproduced in Table 1.

    Table 1. Descriptive characteristics of the survey informants (N=176).a
    VariablesInformants, n (%)
    Demographic variables

    		Gender


    		Male42 (23.9)


    		Female134 (76.1)

    		Age group by birth year range


    		1981-199054 (30.7)


    		1971-198061 (34.7)


    		1956-197061 (34.7)

    		Total years of education from Pakistan and Norway


    		0 years14 (8.0)


    		5 years13 (7.4)


    		<10 years17 (9.7)


    		<12 years33 (18.8)


    		<14 years39 (22.2)


    		≥14 years55 (31.3)

    		Self-assessment of health status (score)


    		Excellent (5)11 (6.3)


    		Very good (4)27 (15.3)


    		Good (3)70 (39.8)


    		Fair (2)37 (21.0)


    		Going up and down (1)19 (10.8)


    		Poor (0)12 (6.8)

    		Frequency of asking for help when using ICTb


    		Always (4)18 (10.2)


    		Often (3)26 (14.8)


    		Sometimes (2)51 (29.0)


    		Seldom (1)12 (6.8)


    		Never (0)68 (38.6)
    Experience of eHealth use for T2DMcself-care in the last 12 months

    		For seeking relevant information


    		By using search engines that require input of search terms35 (19.9)


    		On specific Web sites or by email subscriptions that can be navigated by only scrolling and clicking63 (35.8)

    		For communication and consulting


    		By using ICT in general for closed conversation with a few specific acquaintances84 (47.7)


    		By social networking service58 (33.0)

    		For active decision making on self-care by using Web apps or apps for:


    		Keeping track of health information25 (14.2)


    		Self-assessment of health status38 (21.6)

    		Total number (variety) of eHealth types experienced


    		≥80 (0.0)


    		72 (1.1)


    		65 (2.8)


    		57 (4.0)


    		49 (5.1)


    		328 (15.9)


    		238 (21.6)


    		146 (26.1)


    		041 (23.3)

    aThis is a reproduction of Table 4 in our previous study [56]. There are modifications of labeling of the experience of eHealth use for T2DM self-care in the last 12 months as well as the omission of data not relevant to this paper.

    bICT: information and communication technology.

    cT2DM: type 2 diabetes mellitus.

    Table 2 shows the distribution of the informants by the survey data related to language, length of residence, and diagnosis of T2DM. A vast majority of the informants showed confidence in reading (162/176, 92.0% for a total of “strongly agree” and “agree”) and writing (152/176, 86.3%) in Urdu alphabet. The proportions of informants with confidence in reading (123/176, 69.9%) and writing (116/176, 65.9%) in Roman Urdu were smaller than for Urdu alphabet. Not a negligible number of informants expressed lack of confidence in reading (7/176, 3.9% for a total of “strongly disagree” and “disagree”) and writing (13/176, 7.4%) the Urdu alphabet. The proportions of the sample lacking confidence in reading (26/176, 14.8%) and writing (33/176, 18.7%) in Roman Urdu were even higher. The Urdu literacy score was mean 4.26 (SD 1.26).

    Regarding confidence level in the Norwegian language, the numbers concentrate around “neither” or “agree” in all three skills. The Norwegian language proficiency score was mean 3.35 (SD 1.08).

    In total, 91 informants (51.7%) named English as another language they had confidence in. Arabic and Punjabi were named by two informants each, which is too small of a number to be applied in the statistical analyses. Therefore, we interpreted the answer “English” as “having English proficiency” in the further statistical analysis.

    The majority of the sample (123/176, 69.9%) had lived in Norway for more than 10 years. The sample included only one informant (0.6%) who had lived in Norway for less than a year. A total of 27 informants (15.3%) answered that they had been diagnosed with T2DM.

    Association Between User Factors and eHealth Use

    Table 3 shows the Pearson correlation coefficients between the different variables. The highest correlation coefficient was .78 between Urdu literacy and education and digital skills (P<.001).

    Table 2. Descriptive characteristics of the survey informants regarding target group-specific user factors (N=176).
    User factorsInformants, n (%)
    Urdu literacy “I am very confident in”:

    		Reading Urdu alphabet


    		Strongly agree (5)147 (83.5)


    		Agree (4)15 (8.5)


    		Neither (3)7 (4.0)


    		Disagree (2)2 (1.1)


    		Strongly disagree (1)5 (2.8)

    		Writing Urdu alphabet


    		Strongly agree (5)137 (77.8)


    		Agree (4)15 (8.5)


    		Neither (3)11 (6.3)


    		Disagree (2)3 (1.7)


    		Strongly disagree (1)10 (5.7)

    		Reading Roman Urdu


    		Strongly agree (5)95 (54.0)


    		Agree (4)28 (15.9)


    		Neither (3)27 (15.3)


    		Disagree (2)6 (3.4)


    		Strongly disagree (1)20 (11.4)

    		Writing Roman Urdu


    		Strongly agree (5)94 (53.4)


    		Agree (4)22 (12.5)


    		Neither (3)27 (15.3)


    		Disagree (2)9 (5.1)


    		Strongly disagree (1)24 (13.6)
    Norwegian language proficiency “I am very confident in”:

    		Reading Norwegian


    		Strongly agree (5)26 (14.8)


    		Agree (4)71 (40.3)


    		Neither (3)46 (26.1)


    		Disagree (2)24 (13.6)


    		Strongly disagree (1)9 (5.1)

    		Writing Norwegian


    		Strongly agree (5)22 (12.5)


    		Agree (4)53 (30.1)


    		Neither (3)53 (30.1)


    		Disagree (2)33 (18.8)


    		Strongly disagree (1)15 (8.5)

    		Speaking Norwegian


    		Strongly agree (5)28 (15.9)


    		Agree (4)49 (27.8)


    		Neither (3)73 (41.5)


    		Disagree (2)18 (10.2)


    		Strongly disagree (1)8 (4.5)
    Any other language an informant is confident in

    		English91 (51.7)

    		Punjabi2 (1.1)

    		Arabic2 (1.1)
    Length of residence in Norway

    		0-1 year1 (0.6)

    		1-5 years20 (11.4)

    		5-10 years32 (18.2)

    		More than 10 years123 (69.9)
    Type 2 diabetes mellitus27 (15.3)
    Table 3. Correlation coefficients between the independent variables for the statistical analyses and P values.
    Variable123456789
    1. Being a female—a







    2. Age−.04







    3. Self-assessment of health status−.23b−.32c






    4. Norwegian language proficiency−.25b−.06.22b





    5. Type 2 diabetes−.02.32c−.33c−.16d




    6. English proficiency−.25b−.29c.22b.50c−.19d



    7. Education and digital skills−.31c−.37c.43c.54c−.26c.55c


    8. Urdu literacy−.22b−.34c.38c.51c−.28c.53c.78c

    9. Logarithm of years of residence in Norway.09.59c−.31c.16d.20b−.18d−.29c−.21b—

    aNot applicable.

    bP<.01.

    cP<.001.

    dP<.05.

    Table 4 shows the estimates of six multiple logistic regression models. The table shows results for the best regression model (based on best fit) for each dependent variable. To interpret the table, for example, “for seeking relevant information by using search engines that require the input of search terms” represents the dependent variable for the first regression analyses and the next three rows show the independent variables for this model. Note that estimates in the logistic regression analysis are the logarithm of the odds ratio, but we transformed them to odds ratios in Table 4 to make it easier to interpret. Table 5 shows the results of the Poisson regression analysis. Again, the best regression model is shown.

    In the models in which language-relevant variables remained, Urdu literacy and Norwegian language proficiency were positively related to use of different types of eHealth for T2DM self-care. Urdu literacy was positively associated with the following: seeking relevant information to T2DM self-care on specific websites or by email subscriptions, communication and consulting about T2DM self-care by SNS, and with the total number of eHealth types experienced. Norwegian language proficiency was positively related to use of Web or mobile apps for self-assessment of health. Having English proficiency was neither positively nor negatively related to any dependent variable.

    Regarding years of residence in Norway, the logarithm of this value was positively associated with communication and consulting about T2DM self-care by using ICT for a closed conversation with a few specific acquaintances, and by SNS, as well as with the total number of eHealth types experienced. Having been diagnosed with T2DM was not related to any dependent variable.

    The composite variable education and digital skills appeared positively associated with the use of Web or mobile apps for keeping track of health information. Being a female was positively associated with using ICT for closed communication and consultation about T2DM self-care with a few specific acquaintances. Age had no significant association with any dependent variable. Self-assessment of health status did not remain in any model.

    Table 4. Logistic regression analysis of the association between eHealth use and user factors.
    Dependent and independent variablesOdds ratio (95% CI)P valuea
    For seeking relevant information


    		By using search engines that require theinputof search terms



    		Intercept0.022 (0.002-0.209).001


    		T2DMb3.576 (1.301-9.832).01


    		Urdu literacy1.649 (1.026-2.651).04

    		On specific Web sites or by email subscriptions that can be navigated by only scrolling and clicking


    		Intercept0.019 (0.002-0.144)<.001


    		Urdu literacy2.155 (1.388-3.344).001


    		Logarithm of years of residence in Norway1.371 (0.966-1.947).08
    For communication and consulting


    		By using information and communication technology in general for a closed conversation with a few specific acquaintances


    		Intercept0.396 (0.199-0.788).008


    		Being a female2.883 (1.335-6.227).007


    		Logarithm of years of residence in Norway1.728 (1.193-2.503).004

    		By social networking service



    		Intercept0.002 (0-0.207).008


    		Age0.951 (0.903-1.001).06


    		Having English proficiency0.379 (0.166-0.863).02


    		Urdu literacy5.697 (2.487-13.053)<.001


    		Logarithm of years of residence in Norway2.098 (1.265-3.480).004
    For active decision making on T2DM self-care by usingWebor mobile apps for


    		Keeping track of health information



    		Intercept4.253 (0.256-70.507).31


    		Age0.909 (0.844-0.979).01


    		Education and digital skills3.930 (1.627-9.492).002


    		Logarithm of years of residence in Norway1.753 (0.983-3.127).06

    		Self-assessment of health



    		Intercept0.263 (0.023-2.951).28


    		Age0.935 (0.887-0.985).01


    		Norwegian language proficiency2.285 (1.294-4.036).004


    		Having English proficiency0.418 (0.154-1.139).09


    		Education and digital skills2.414 (1.104-5.28).03

    aA Bonferroni-corrected significance value of 0.00714 was used to interpret the P values.

    bT2DM: type 2 diabetes mellitus.

    Table 5. Poisson regression analysis of the association between the variety of experienced eHealth activities and user factors.
    Total number (variety) of eHealth types experiencedEstimate (95% CI)P valuea
    Intercept−0.158 (−1.321, 1.005).79
    Age−0.019 (−0.035, −0.003).02
    T2DMb0.310 (−0.012, 0.633).06
    Having English proficiency−0.209 (−0.462, 0.044).11
    Education and digital skills0.181 (−0.037, 0.399).10
    Urdu literacy0.350 (0.148, 0.552).001
    Logarithm of years of residence in Norway0.270 (0.117, 0.424).001

    aA Bonferroni-corrected significance value of 0.00714 was used to interpret the P values.

    bT2DM: type 2 diabetes mellitus.


    Principal Findings and Implications for Future Studies

    This study explored diverse eHealth use for T2DM self-care by first-generation immigrants from Pakistan in the Oslo area in relation to two types of user factors: (1) target group-specific user factors including proficiency in relevant languages, length of residence in the host country, and a diagnosis of T2DM, and (2) general user factors including education level, age, gender, self-assessment of health status, and digital skills. The results of multiple regression analyses showed that in all the final models at least one of the target group-specific variables remained. In addition, in most models target group-specific variables were most strongly associated with eHealth use. Therefore, in our survey sample, the inclusion of target group-specific user factors yielded better-fitting models predicting use of eHealth for different purposes compared with only including general user factors.

    Informants with higher Urdu literacy were more likely than those with lower Urdu literacy to seek relevant information on Web portals or obtain such information by email subscriptions, to use SNS for communication and consulting about T2DM self-care, and to use a wider variety of eHealth services. For these types of eHealth use, Norwegian language proficiency did not remain in the final models. Use of mother tongue as the first choice for seeking information of interest is probably a natural behavior, as long as users expect or know that they can reach the information they need. SNS is often used for keeping online connection and communication with people one already knows. Therefore, it is also reasonable to speculate that informants communicate by SNS with their Pakistani family, friends, or alike who share an interest in T2DM self-care.

    Norwegian language proficiency was positively associated with only use of Web or mobile apps for self-assessment of health as a part of T2DM self-care. It is reasonable to speculate that informants who are fluent in Norwegian have a better chance to know and use such services provided in the Norwegian language and that Norwegian eHealth services are of value for this subgroup of immigrants. The result leaves a question about why Urdu language literacy was not associated with the use of these eHealth services. Implications here could be either that such services are unavailable in Urdu language or that the informants were unaware of them despite their availability. Thus, it will be worth further examining if the informants have ever tried to search the internet for or been aware of such services offered in the Urdu language. Given the results that Norwegian language proficiency and Urdu literacy were associated with different types of eHealth activities, it would also be worth further investigating the difference in information or advice they receive from the different channels as well as the difference in user experiences. Another interesting question to pursue in future studies would be if the use of this type of eHealth service provided in the Norwegian language can positively affect the overall social integration in the Norwegian society for minority populations and vice versa, as well as health behavior.

    Approximately half of the sample answered they are also confident in English. Nevertheless, English proficiency was not associated with any type of eHealth use for T2DM self-care. It is difficult to speculate the reason for irrelevance between having English proficiency and eHealth use for T2DM self-care among the sample. However, the implication here is that the target population may not use eHealth content in English for T2DM self-care as much as those in the Norwegian or Urdu language. As described in the Introduction, there are only a limited number of studies available that have investigated the relevance between immigrant populations’ eHealth use and language proficiency or similar [28,30,32,33]. All these studies are from the United States, where English is the primary language. Because of this, as well as differences in study design, our results cannot be simply compared with these studies. However, most of the studied cases found having English proficiency or similar user factors to be associated with eHealth use or confidence regarding eHealth use [28,30,32,33]. The results call for the need for more knowledge by studying other immigrant groups in European countries where English is not the primary language of either the immigrant groups or the host country.

    In its logarithm form, years of residence in Norway was positively associated with closed communication with a few acquaintances and use of SNS, as well as the total number of eHealth types experienced in the last 12 months. An interpretation here could be that the longer immigrants live in Norway, the more people they connect with who also are at high risk of, or having, T2DM. Further, they may be more exposed to eHealth possibilities than those with shorter length of residence in Norway, although such effect may not be linear given the relatively short history of eHealth and needs to be seen by controlling the age of users. Other studies about immigrants’ use of eHealth in the United States showed that the effect of length of residence on eHealth use is not monotonically increasing; the group in the middle segment showed the highest eHealth engagement compared with those who lived in the host country the shortest or the longest [30,34]. These studies in the United States cannot be simply compared with our study because they include immigrants regardless of age at immigration, whereas all the informants in this study immigrated to Norway after the age of 18. Also, the types of eHealth activity they explore are not exactly the same as this study. Nevertheless, given the natural relevance between age and years of residence, future studies with larger sample sizes should further explore the relation between these factors by, for example, segmenting the sample and comparing the subgroups with different lengths of residence.

    Despite remaining in the final model of using search engines to seek relevant information on T2DM self-care and the variety of eHealth activities experienced, having a diagnosis with T2DM was not significantly associated with any type of eHealth activity. One of the inclusion criteria of the survey was “being motivated for and capable of performing activities for self-care of T2DM”; thus, having a diagnosis of T2DM in itself may not be relevant enough to differentiate users and nonusers of eHealth for self-care of T2DM among the survey sample. Self-assessment of health did not remain in any model. Being in good condition could be a consequence of—and being in bad condition could be the motivation for—using eHealth for self-care. Thus, self-assessment of health at the time of participation in a survey may not be a good indicator of eHealth use for T2DM self-care.

    Implications for Development and Dissemination of Future eHealth Services to Reduce Social Health Inequalities

    The results present valuable insight about relevant user factors of diverse eHealth activities for T2DM self-care among one of the little-studied, vulnerable user groups. The yielded knowledge can be used to guide further research on designing, developing, and disseminating eHealth tools for this vulnerable population to achieve the goal to reduce social health inequalities.

    Despite an inclusion criterion of being “motivated for and capable of activities for self-care of T2DM,” not all types of eHealth activities were common enough among the sample. The findings of this study indicate that there is a certain divide among the target population regarding eHealth activities. First, those with high Norwegian language proficiency, and those with high education and digital skills, are more likely to use apps for self-assessment of health and to keep track of health information by using apps, respectively. Second, those with high Urdu literacy were more likely to go online for relevant information seeking and communication via SNS in connection with T2DM self-care than those with low Urdu literacy. In addition, nearly half of the sample had used ICT for closed conversation with specific acquaintances for consulting about T2DM, which is strongly associated with the longer period of residence in Norway. The implication here could be the importance of disseminating reliable health information or self-care apps through their social network. Relevant studies on eHealth use by immigrant populations also support use of the target populations networks [34] as well as stakeholders to promote available eHealth services [30,62].

    On the other hand, translation into English seems not to be equally relevant, as having English proficiency was found to have no association with eHealth use for this population despite that more than half the sample showed confidence in English. Hence, to reduce social inequalities through eHealth services, these need to be offered both in Norwegian and minority languages that are especially relevant to high-risk groups of social health inequalities. The importance of provisioning easy-to-read, culturally appropriate health information in their own language is also outlined in relevant studies [28,30,32]. The quality and amount of information in eHealth content provided in the nondominant language should not be deteriorated because such a problem is reported in relevant studies [63,64]. Immigrants are heterogeneous in terms of linguistic and cultural background. Provision of eHealth services in many minority languages that are adapted to the cultural background would be highly demanding and costly. Machine translation seems not to have reached a level of sufficient accuracy, but it could be useful when used in combination with human translation [65,66]. Moreover, further development of machine translation technology has great potential to enable provision of eHealth services in minority languages at less cost and at a more trustable level than today.

    Limitations

    Limitations regarding the survey sample and the survey methods we reported in our previous study [55] apply to this study as well. In addition, we should note that this study may suffer from nonprobability sampling, albeit the multirecruitment method we took was more appropriate for recruiting the target population than ordinary probability sampling [67].

    For this particular study, we should note the following. Data about English proficiency was obtained in an indirect manner by asking about any language that the informants were confident in. This method may have caused underestimation of the number of informants with high confidence in English proficiency to some extent but should not have caused overestimation. Given the results of the regression analyses, even if the number of informants with high English proficiency increases to some degree, the results would not change drastically.

    Our study relates to one specific population in one specific context. We invite future research to explore if these findings could be replicated in other contexts and populations and explore potential generalizability of the results.

    Conclusion

    The contribution of this study is advancing knowledge on user factors associated with diverse eHealth activities among one of the immigrant populations in European countries, which is a vulnerable and little-known user group. Particularly, this study showed the importance of examining the target group-specific user factors for their associations with experience of diverse eHealth activities when the target group is a vulnerable immigrant population. This study also implied that to facilitate and enhance eHealth use by immigrant populations, eHealth services should be provided in minority languages rather than using English as a common language for foreigners. Future studies are needed for a further understanding of other target user groups vulnerable to poor health or low socioeconomic position due to the same or similar reasons as our sample.

    Acknowledgments

    The Department of Computer Science, the Faculty of Technology, Design and Art, Oslo Metropolitan University, funded this research in December 2014. The authors are grateful to Monica Morris and Anica Munir for serving as research assistants.

    Authors' Contributions

    NT and HLH conceived this study. NT drafted the study design and the first version of the manuscript. HLH designed and conducted the statistical analyses and wrote the draft of the relevant part. The other three authors contributed to further development of the study design and to finalizing the manuscript by giving comments to all versions of the manuscript draft. All authors read and approved the final manuscript.

    Conflicts of Interest

    None declared.

    1. Marschang S. Health Inequalities and eHealth: Report of the eHealth Stakeholder Group. 2014 Feb 21.   URL: http://ec.europa.eu/newsroom/dae/document.cfm?doc_id=5170 [accessed 2018-01-04] [WebCite Cache]
    2. Morey OT. Digital disparities. J Consum Health Internet 2007 Dec;11(4):23-41. [CrossRef]
    3. Kim G, Aguado LCX, Chiriboga DA, Jang Y, Parmelee P, Allen RS. Limited English proficiency as a barrier to mental health service use: a study of Latino and Asian immigrants with psychiatric disorders. J Psychiatr Res 2011 Jan;45(1):104-110. [CrossRef] [Medline]
    4. Chang E, Chan KS, Han H. Factors associated with having a usual source of care in an ethnically diverse sample of Asian American adults. Med Care 2014 Sep;52(9):833-841. [CrossRef] [Medline]
    5. Chang E, Chan KS, Han H. Effect of acculturation on variations in having a usual source of care among Asian Americans and non-Hispanic whites in California. Am J Public Health 2015 Feb;105(2):398-407. [CrossRef] [Medline]
    6. Isaacs T, Hunt D, Ward D, Rooshenas L, Edwards L. The Inclusion of ethnic minority patients and the role of language in telehealth trials for type 2 diabetes: a systematic review. J Med Internet Res 2016 Sep 26;18(9):e256 [FREE Full text] [CrossRef] [Medline]
    7. Testa R, Bonfigli AR, Genovese S, Ceriello A. Focus on migrants with type 2 diabetes mellitus in European Countries. Intern Emerg Med 2016 Apr;11(3):319-326. [CrossRef] [Medline]
    8. King-Shier K, Lau A, Fung S, LeBlanc P, Johal S. Ethnocultural influences in how people prefer to obtain and receive health information. J Clin Nurs 2018 Apr;27(7-8):e1519-e1528. [CrossRef] [Medline]
    9. Alzubaidi H, Mc Namara K, Browning C. Time to question diabetes self-management support for Arabic-speaking migrants: exploring a new model of care. Diabet Med 2017 Dec;34(3):348-355. [CrossRef] [Medline]
    10. Heubusch K. Access + understanding. The role of health literacy in patient-centric health IT. J AHIMA 2010 May;81(5):32-34; quiz 35. [Medline]
    11. Sarkar U, Karter AJ, Liu JY, Adler NE, Nguyen R, Lopez A, et al. The literacy divide: health literacy and the use of an internet-based patient portal in an integrated health system-results from the diabetes study of northern California (DISTANCE). J Health Commun 2010;15 Suppl 2:183-196 [FREE Full text] [CrossRef] [Medline]
    12. Nazroo JY, Smith GD. The contribution of socioeconomic position to health differentials between ethnic groups: evidence from the United States and Britain. In: McBeth H, Shetty P, editors. Health and Ethnicity. London: CRC Press; Nov 23, 2000:41-58.
    13. Sommer I, Griebler U, Mahlknecht P, Thaler K, Bouskill K, Gartlehner G, et al. Socioeconomic inequalities in non-communicable diseases and their risk factors: an overview of systematic reviews. BMC Public Health 2015 Sep 18;15:914 [FREE Full text] [CrossRef] [Medline]
    14. Jenum AK, Diep LM, Holmboe-Ottesen G, Holme IM, Kumar BN, Birkeland KI. Diabetes susceptibility in ethnic minority groups from Turkey, Vietnam, Sri Lanka and Pakistan compared with Norwegians - the association with adiposity is strongest for ethnic minority women. BMC Public Health 2012;12:150 [FREE Full text] [CrossRef] [Medline]
    15. Tran AT, Straand J, Diep LM, Meyer HE, Birkeland KI, Jenum AK. Cardiovascular disease by diabetes status in five ethnic minority groups compared to ethnic Norwegians. BMC Public Health 2011 Jul 13;11:554 [FREE Full text] [CrossRef] [Medline]
    16. Meeks KA, Freitas-Da-Silva D, Adeyemo A, Beune EJ, Modesti PA, Stronks K, et al. Disparities in type 2 diabetes prevalence among ethnic minority groups resident in Europe: a systematic review and meta-analysis. Intern Emerg Med 2016 Apr;11(3):327-340. [CrossRef] [Medline]
    17. Fernando E, Razak F, Lear SA, Anand SS. Cardiovascular disease in South Asian migrants. Can J Cardiol 2015 Sep;31(9):1139-1150. [CrossRef] [Medline]
    18. World Health Organization Regional Office for Europe. About Health 2020   URL: http:/​/www.​euro.who.int/​en/​health-topics/​health-policy/​health-2020-the-european-policy-for-health-and-well-being/​about-health-2020 [accessed 2016-02-14] [WebCite Cache]
    19. World Health Organization Regional Office for Europe. 2017 Jun. eHealth-where are we now?   URL: http://www.euro.who.int/en/health-topics/Health-systems/e-health/ehealth-where-are-we-now [accessed 2018-01-05] [WebCite Cache]
    20. Bacigalupe G, Askari SF. E-Health innovations, collaboration, and healthcare disparities: developing criteria for culturally competent evaluation. Fam Syst Health 2013 Sep;31(3):248-263. [CrossRef] [Medline]
    21. Kim H, Xie B. Health literacy in the eHealth era: a systematic review of the literature. Patient Educ Couns 2017 Jun;100(6):1073-1082. [CrossRef] [Medline]
    22. Sheon AR, Bolen SD, Callahan B, Shick S, Perzynski AT. Addressing disparities in diabetes management through novel approaches to encourage technology adoption and use. JMIR Diabetes 2017 Jul 13;2(2):e16 [FREE Full text] [CrossRef] [Medline]
    23. Lee C, Ramírez AS, Lewis N, Gray SW, Hornik RC. Looking beyond the Internet: examining socioeconomic inequalities in cancer information seeking among cancer patients. Health Commun 2012;27(8):806-817 [FREE Full text] [CrossRef] [Medline]
    24. Norman CD, Skinner HA. eHealth literacy: essential skills for consumer health in a networked world. J Med Internet Res 2006 Jun;8(2):e9 [FREE Full text] [CrossRef] [Medline]
    25. Latulippe K, Hamel C, Giroux D. Social health inequalities and eHealth: a literature review with qualitative synthesis of theoretical and empirical studies. J Med Internet Res 2017 Apr 27;19(4):e136 [FREE Full text] [CrossRef] [Medline]
    26. Mesch G, Mano R, Tsamir J. Minority status and health information search: a test of the social diversification hypothesis. Soc Sci Med 2012 Sep;75(5):854-858. [CrossRef] [Medline]
    27. Zeissig SR, Singer S, Koch L, Zeeb H, Merbach M, Bertram H, et al. Utilisation of psychosocial and informational services in immigrant and non-immigrant German cancer survivors. Psychooncology 2015 Aug;24(8):919-925. [CrossRef] [Medline]
    28. Gonzalez M, Sanders-Jackson A, Emory J. Online health information-seeking behavior and confidence in filling out online forms among Latinos: a cross-sectional analysis of the California Health Interview Survey, 2011-2012. J Med Internet Res 2016 Jul 04;18(7):e184 [FREE Full text] [CrossRef] [Medline]
    29. Nölke L, Mensing M, Krämer A, Hornberg C. Sociodemographic and health-(care-)related characteristics of online health information seekers: a cross-sectional German study. BMC Public Health 2015 Jan 29;15:31 [FREE Full text] [CrossRef] [Medline]
    30. Nguyen A, Mosadeghi S, Almario CV. Persistent digital divide in access to and use of the Internet as a resource for health information: results from a California population-based study. Int J Med Inform 2017 Jul;103:49-54. [CrossRef] [Medline]
    31. Massey PM, Langellier BA, Sentell T, Manganello J. Nativity and language preference as drivers of health information seeking: examining differences and trends from a U.S. population-based survey. Ethn Health 2017 Dec;22(6):596-609. [CrossRef] [Medline]
    32. Bender MS, Choi J, Arai S, Paul SM, Gonzalez P, Fukuoka Y. Digital technology ownership, usage, and factors predicting downloading health apps among caucasian, filipino, korean, and latino americans: the digital link to health survey. JMIR Mhealth Uhealth 2014;2(4):e43 [FREE Full text] [CrossRef] [Medline]
    33. De Jesus M, Xiao C. Predicting internet use as a source of health information: a “language divide” among the Hispanic population in the United States. Policy Internet 2012 Aug 10;4(2):1-11. [CrossRef]
    34. Zhao X, Yang B, Wong C. Analyzing trend for U.S. immigrants' e-Health engagement from 2008 to 2013. Health Commun 2018 May 16:1-11. [CrossRef] [Medline]
    35. Himmelgreen DA, Pérez-Escamilla R, Martinez D, Bretnall A, Eells B, Peng Y, et al. The longer you stay, the bigger you get: length of time and language use in the US are associated with obesity in Puerto Rican women. Am J Phys Anthropol 2004 Sep;125(1):90-96. [CrossRef] [Medline]
    36. Lebrun LA. Effects of length of stay and language proficiency on health care experiences among immigrants in Canada and the United States. Soc Sci Med 2012 Apr;74(7):1062-1072. [CrossRef] [Medline]
    37. Sanchez-Vaznaugh EV, Kawachi I, Subramanian SV, Sánchez BN, Acevedo-Garcia D. Differential effect of birthplace and length of residence on body mass index (BMI) by education, gender and race/ethnicity. Soc Sci Med 2008 Oct;67(8):1300-1310. [CrossRef] [Medline]
    38. Zhang W, Ta VM. Social connections, immigration-related factors, and self-rated physical and mental health among Asian Americans. Soc Sci Med 2009 Jun;68(12):2104-2112. [CrossRef] [Medline]
    39. Oza-Frank R, Cunningham SA. The weight of US residence among immigrants: a systematic review. Obes Rev 2010 Apr;11(4):271-280. [CrossRef] [Medline]
    40. Oza-Frank R, Stephenson R, Narayan KM. Diabetes prevalence by length of residence among US immigrants. J Immigr Minor Health 2011 Feb;13(1):1-8. [CrossRef] [Medline]
    41. Jonnalagadda SS, Diwan S. Health behaviors, chronic disease prevalence and self-rated health of older Asian Indian immigrants in the U.S. J Immigr Health 2005 Apr;7(2):75-83. [CrossRef] [Medline]
    42. Commodore-Mensah Y, Ukonu N, Obisesan O, Aboagye JK, Agyemang C, Reilly CM, et al. Length of residence in the United States is associated with a higher prevalence of cardiometabolic risk factors in immigrants: a contemporary analysis of the National Health Interview Survey. J Am Heart Assoc 2016 Nov 04;5(11):1 [FREE Full text] [CrossRef] [Medline]
    43. The Role of the Pharmacist in Self-Care and Self-Medication. 1998.   URL: http://apps.who.int/medicinedocs/pdf/whozip32e/whozip32e.pdf [accessed 2015-12-09] [WebCite Cache]
    44. Dzamarija MT, Steinkellner A. Statistics Norway. 2018 Mar 05. 14 per cent of population are immigrants   URL: https://www.ssb.no/en/befolkning/artikler-og-publikasjoner/14-per-cent-of-population-are-immigrants [accessed 2018-03-13] [WebCite Cache]
    45. Sandnes T. Statistics Norway. 2016 Mar 08. More family reunifications than new marriages   URL: https:/​/www.​ssb.no/​en/​befolkning/​artikler-og-publikasjoner/​more-family-reunifications-than-new-marriages [accessed 2018-03-15] [WebCite Cache]
    46. Pettersen SV. Statistics Norway. 2009. Innvandrere i norske kommuner: Demografi, levekår og deltakelse i arbeidsstyrken   URL: https://www.ssb.no/a/publikasjoner/pdf/rapp_200936/rapp_200936.pdf [accessed 2015-12-18] [WebCite Cache]
    47. Statistics Norway. 2015. Level of education for immigrants 16 years and older   URL: http:/​/www.​ssb.no/​270248/​level-of-education-for-immigrants-16-years-and-older.​country-background.​numbers [accessed 2016-09-16] [WebCite Cache]
    48. Encyclopaedia Britannica. 2009 May 28. Urdu language   URL: https://www.britannica.com/topic/Urdu-language [accessed 2018-03-15] [WebCite Cache]
    49. UNESCO Institute of Statistics. Pakistan   URL: http://uis.unesco.org/en/country/PK [accessed 2018-03-22] [WebCite Cache]
    50. Tatara N, Kjøllesdal MK, Mirkovic J, Andreassen HK. eHealth use among first-generation immigrants from Pakistan in the Oslo Area, Norway, with focus on diabetes: survey protocol. JMIR Res Protoc 2016 Apr 25;5(2):e79 [FREE Full text] [CrossRef] [Medline]
    51. Vrålstad S, Wiggen KS. Statistics Norway. 2017. Levekår blant innvandrere i Norge 2016   URL: https:/​/www.​ssb.no/​sosiale-forhold-og-kriminalitet/​artikler-og-publikasjoner/​_attachment/​309211?_ts=160ea9e4890 [accessed 2018-05-31] [WebCite Cache]
    52. Bhopal RS. A four-stage model explaining the higher risk of Type 2 diabetes mellitus in South Asians compared with European populations. Diabet Med 2013 Jan;30(1):35-42. [CrossRef] [Medline]
    53. Mellin-Olsen T, Wandel M. Changes in food habits among Pakistani immigrant women in Oslo, Norway. Ethn Health 2005 Nov;10(4):311-339. [CrossRef] [Medline]
    54. Fagerli RA, Lien ME, Wandel M. Experience of dietary advice among Pakistani-born persons with type 2 diabetes in Oslo. Appetite 2005 Dec;45(3):295-304. [CrossRef] [Medline]
    55. Tatara N, Hammer HL, Andreassen HK, Mirkovic J, Kjøllesdal MKR. The association between commonly investigated user factors and various types of eHealth use for self-care of type 2 diabetes: case of first-generation immigrants from Pakistan in the Oslo Area, Norway. JMIR Public Health Surveill 2017 Oct 05;3(4):e68 [FREE Full text] [CrossRef] [Medline]
    56. Abe-Kim J, Takeuchi DT, Hong S, Zane N, Sue S, Spencer MS, et al. Use of mental health-related services among immigrant and US-born Asian Americans: results from the National Latino and Asian American Study. Am J Public Health 2007 Jan;97(1):91-98. [CrossRef] [Medline]
    57. Mui AC, Kang S, Kang D, Domanski MD. English language proficiency and health-related quality of life among Chinese and Korean immigrant elders. Health Soc Work 2007 May;32(2):119-127. [Medline]
    58. Pottie K, Ng E, Spitzer D, Mohammed A, Glazier R. Language proficiency, gender and self-reported health: an analysis of the first two waves of the longitudinal survey of immigrants to Canada. Can J Public Health 2008;99(6):505-510. [Medline]
    59. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing   URL: https://www.r-project.org/ [accessed 2005-12-09] [WebCite Cache]
    60. Akaike H. A new look at the statistical model identification. IEEE Trans Automat Contr 1974 Dec;19(6):716-723. [CrossRef]
    61. Cronbach LJ. Coefficient alpha and the internal structure of tests. Psychometrika 1951 Sep;16(3):297-334. [CrossRef]
    62. Morahan-Martin JM. How internet users find, evaluate, and use online health information: a cross-cultural review. Cyberpsychol Behav 2004 Oct;7(5):497-510. [Medline]
    63. Li J, Lowres N, Jin K, Zhang L, Neubeck L, Gallagher R. Quality and cultural sensitivity of linguistically appropriate cardiovascular disease information for Chinese immigrants: a review of online resources from heart foundations. J Cardiovasc Nurs 2018;33(3):269-280. [CrossRef] [Medline]
    64. Bastos A, Paiva D, Azevedo A. Quality of health information on acute myocardial infarction and stroke in the world wide web. Acta Med Port 2014;27(2):223-231 [FREE Full text] [Medline]
    65. Dew KN, Turner AM, Choi YK, Bosold A, Kirchhoff K. Development of machine translation technology for assisting health communication: a systematic review. J Biomed Inform 2018 Sep;85:56-67. [CrossRef] [Medline]
    66. Chen X, Acosta S, Barry AE. Machine or human? Evaluating the quality of a language translation mobile app for diabetes education material. JMIR Diabetes 2017 Jun 29;2(1):e13 [FREE Full text] [CrossRef] [Medline]
    67. Hussain-Gambles M, Leese B, Atkin K, Brown J, Mason S, Tovey P. Involving South Asian patients in clinical trials. Health Technol Assess 2004 Oct;8(42):iii, 1-iii109 [FREE Full text] [Medline]

    ‎
    AIC: Akaike information criterion
    ICT: information and communication technology
    OR: odds ratio
    SNS: social network service
    T2DM: type 2 diabetes mellitus

    Edited by G Eysenbach; submitted 21.08.18; peer-reviewed by M Gonzalez, C Calvo Lobo, D Arigo, S Balsari, M Ghajarzadeh; comments to author 20.03.19; revised version received 11.07.19; accepted 28.07.19; published 16.08.19

    Copyright

    ©Naoe Tatara, Hugo Lewi Hammer, Jelena Mirkovic, Marte Karoline Råberg Kjøllesdal, Hege Kristin Andreassen. Originally published in JMIR Public Health and Surveillance (http://publichealth.jmir.org), 16.08.2019.

    This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Public Health and Surveillance, is properly cited. The complete bibliographic information, a link to the original publication on http://publichealth.jmir.org, as well as this copyright and license information must be included.