Protocol for the CHINT study: a cross-sectional observational study of risk factors for cancer and other non-communicable diseases in the Chinese community of Milan

Introduction

Cancer incidence rates vary markedly between and within countries, likely reflecting differences in exposure to risk factors and access to cancer prevention initiatives.1 Migration may also influence cancer incidence rates in the host country.2 Although most immigrants are economically below average for a host country, they do not always have poorer health. A 2016 study conducted in Europe found that all-cause and cancer-specific mortality varied in immigrants according to origin: those from east Asia had lower mortality and those from eastern Europe had higher mortality than the host population.3 A 2018 study found that immigrants from non-western countries had more favourable all-cancer morbidity and mortality than the populations of European host countries, but with considerable cancer site-specific risk variation.3 4 Thus, while the incidences of cancers related to infections experienced in early life, such as liver, cervical and stomach cancer were higher, incidences of colorectal, lung, breast, ovary, prostate, testicular and bladder cancer were lower in immigrants than the host population. The latter cancers are linked to a ‘western’ lifestyle, characterised by lack of physical activity, unhealthy diet, excessive alcohol consumption, smoking and obesity.4 A 2020 meta-analysis of prospective cohort studies estimated that the healthiest (least western) lifestyle was associated with 29% and 52% lower risks of cancer incidence and cancer mortality, respectively, compared with the least healthy lifestyle.5

The observation that immigrants to Italy generally exhibit healthier behaviours than the locals6 suggests that immigrants can contribute positively to overall population health. However, immigrants also tend to adopt elements of the lifestyle of the host country, so it is important to monitor old and new factors that can compromise health among immigrant populations.

The Chinese community of Italy, numbering 280 0007 is the largest in the European Union. Lombardy (Northern Italy) is the main destination for Chinese migrants to Italy. Milan and its surrounding area hosts over 41 000 Chinese persons, to constitute the largest Chinese community in Italy.7

While the effects of lifestyle behaviours on the health of people in China have been extensively investigated,8–10 limited information is available on Chinese immigrants living in Italy. An observational study on risk factors for cardiovascular diseases and type 2 diabetes on Chinese immigrants is ongoing in Tuscany (Central Italy) and preliminary results indicate that increasing time spent in Italy is associated with increased prevalence of hypertension and type 2 diabetes, but lower prevalence of hypercholesterolaemia.11

To our knowledge, no studies exploring cancer risk factors have been conducted on Chinese people living in Italy. We therefore designed the CHINT study to explore lifestyle-related risk factors for cancer and other NCDs in a sample of persons of Chinese origin residing in Milan or the surrounding area. We expect CHINT to reveal potentially unhealthy lifestyle behaviours among Chinese immigrants, which can be addressed in future prevention programmes.

Objectives and outcomes

The CHINT study aims to ascertain the prevalence, distribution and variation with time spent in Italy and other characteristics, of modifiable risk factors for cancer and other NCDs in the Chinese community of Milan and surrounding area. Therefore, our primary outcome is the presence of NCD risk factors: heavy drinking, smoking, unhealthy diet, high blood pressure, obesity and/or abdominal obesity. We have not specified a hierarchy for these risk factors and have therefore not defined secondary outcomes in this study.

Hypotheses

Lifestyle changes are effective in reducing the risk of chronic diseases.12 13 However, we lack information on the prevalence of increased-risk behaviours in immigrant populations. We also lack knowledge as to what strategies are effective in changing to healthier behaviours in such populations. We hypothesise that Chinese immigrants have been increasing their high-risk behaviours since their arrival in Italy, by adopting more convenient western-style diets and behaviours. Furthermore, because of stressful working lives and cultural/linguistic barriers, members of Chinese immigrant communities are likely to be excluded/self-excluded from disease prevention programmes. We further hypothesise that Chinese people living in and around Milan engage in potentially unhealthy lifestyle behaviours that can be addressed in future prevention programmes.

Methods and analysis

Study design and participants

The CHINT study is a cross-sectional observational study aiming to recruit a sample of about 600 Chinese immigrants of age ≥18 years who have lived in Italy at least since 2019, from among the 41 000 Chinese persons living in and around Milan, Lombardy, Italy.7 Those recruited will include first-generation (born in China), second-generation (born in Italy, with at least one China-born parent) and later generation immigrants. A logic model that includes inputs, outputs and short-term, medium-term and long-term outcomes of CHINT is shown in figure 1.

Figure 1
Figure 1

Logic model of the CHINT study.

Power calculations indicated that sample size of 600 persons provides sufficient power (0.80, alpha level 0.01) to measure with high precision the risk factors being assessed in the study. Examples are: for daily alcohol consumption, 6 mL/day is the detectable difference between two equal-size groups with SD 21 mL/day; for waist circumference 3 cm is the detectable difference between two equal-size groups with SD 10 cm; for daily intake of fruit and vegetable 50 g/day is the detectable difference between two equal-size groups with SD 175 g/day; and for systolic blood pressure 6 mm Hg is the detectable difference between two equal-size groups with SD 21 mm Hg.

Patient and public involvement

Previous studies on immigrant populations12 13 indicate that involvement of community leaders and stakeholders improves compliance, and can reduce refusal and dropout rates. This is particularly true for Chinese communities which are largely self-contained in part because of cultural barriers that include non-familiarity with the western medical concept of prevention.14

It is for this reason we set up the CAB, consisting of stakeholders and including in particular members of Chinese and Chinese-Italian associations. Study researchers and the CAB met several times during the planning stage of the study to better understand the healthy behaviours and health literacy challenges facing the community. When developing the survey tools, the CAB’s contributions were fundamental in crafting the wording of questions so as not to offend sensitivities and in insisting that anthropometric measurements be taken without undressing. The CAB advised against taking blood samples since most Chinese are averse to giving a blood sample.

CAB members and Chinese community members were actively involved in recruitment, and members of the Chinese community (both study staff and volunteers) supported researchers during physical examinations and the completion of the questionnaires.

Meetings with the CAB to discuss the project progress, and initiatives involving peers to convince Chinese community members to participate, will continue until the recruitment ends. Study results will be disseminated at meetings to which the CAB and the entire Chinese population residing in Lombardy will be invited.

Recruitment and first contact

It is not possible to recruit a random sample from the Chinese community since there is no population list that the study can access. To recruit a sample that is as representative of the entire community as possible, we are adopting the following expedients.

Recruitment from members of the 18 Chinese associations based in Milan/Lombardy

Each association is asked to disseminate information about the CHINT study to its members and forward the names of those interested in participating to the study call centre. The call centre is run by an Italian-Chinese association: it contacts persons (via telephone or social media) expressing in interest in participating, and arranges an appointment at the recruitment centre for these persons, and also with persons who contact it directly. The 18 Chinese associations have a combined membership of about 4000; we expect to recruit 300–400 people from among them.

Employees

Manual workers and those with low-skill jobs are under-represented in the Chinese associations. To address this, the directors/owners of Chinese businesses/firms in Milan and hinterland are contacted by the CAB, informed about the study, and asked to disseminate information on the study to all their employees, with instructions on what to do if an employee is interested in participating. To further facilitate study recruitment, Chinese businessmen/women are asked to allow study researchers to visit their businesses to encourage recruitment and to be able to recruit during working hours.

Volunteers

Any eligible persons who present without appointment at a recruitment centre are welcome to participate in CHINT.

The elderly

All recruited persons are informed that their parents would be welcome to join the study and are encouraged to bring them along to the recruitment office for enrolment.

Procedures at recruitment

When potential recruits present at the recruitment centre, COVID-19 safety procedures (eg, green pass, body temperature measurement, hand sanitising, Filtering Face Piece 2) are applied in accordance with evolving Italian legislation. Potential recruits are then conducted to a waiting room to read the informed consent. Researchers and Chinese mother-tongue assistants are always on hand to answer queries and resolve difficulties. Persons who sign the informed consent form are considered to have entered the study.

Minimal personal details and contact information (mobile or email) of the participant are then entered into an electronic offline password-protected recruitment database. A number (personal information number (PIN)) is generated that uniquely identifies the participant and also entered on the database. A card with the newly generated PIN is given to the participant who is accompanied to have his/her physical measurements taken, and complete the lifestyle questionnaire.

Physical measurements

Operative training

The operatives who take the physical measurements (eg, waist and hip circumference) of study participants were trained according to international guidelines.14 In summer 2021, trainee operatives performed repeated measurements on 60 Italian volunteers. Training continued until interoperative and intraoperative precision, as technical error of measurement, was less than 8% for all measurements.15

In a pilot study (November–December 2021), operatives performed the physical measurements on a group of 30 Chinese participants (early recruits to the study). After assessing the comments of the participants and CAB members, the measurement procedure was further refined. For example, participants are now requested to empty their bladder prior to the measurement session, and measurements done with the participant in light clothing.

Weight and height

Weight and height are measured in light clothes, without shoes. Weight is measured twice using an electronic scale (KERN MGD100K-1S05, Germany). The average is taken and rounded to the nearest 100 g. The scale is calibrated every 3 months during data collection. Height is measured with stadiometer (seca Gmbh Co. KG, Germany), with any head gear removed, feet together, heels against the backboard, knees straight and the participant looking straight ahead (Frankfort position).16 The measurement is rounded to the nearest 0.1 cm and recorded.

Waist and hip circumference

Waist and hip circumference are measured with a tape measure with results rounded to the nearest 0.1 cm. Waist circumference is measured at the midpoint between the bottom of the last rib and the top of the hip.16 Hip circumference is measured at the maximum circumference above the buttocks. Removing a garment is not necessary: the thickness of clothes is measured with a calliper and this measurement is used to correct circumferences measured over clothes.

Blood pressure and heart rate

Blood pressure and heart rate are measured using a validated automatic device (Omron HBP-1320)17 according to standard procedures: blood pressure is measured in an upper arm after 5 min of sitting. A second measurement is taken after an additional 2 min of sitting, and is usually the measurement recorded.18

Muscle strength

Muscle strength is assessed as handgrip strength during maximum voluntary contraction, using a hydraulic dynamometer, following standard procedures.19 Participants perform six contractions (three per hand) each separated by 30 s of rest. Mean measures are recorded for each hand.

The study questionnaires

The CHINT study employs two questionnaires: a lifestyle/personal history/medical history questionnaire, and a food frequency questionnaire (FFQ) designed for Chinese people living in Italy. The questionnaires are available in Chinese, Italian and English. Participants usually complete the lifestyle questionnaire online, using a tablet, during the recruitment session. The FFQ is completed later. Error checking algorithms ensure that only valid data are entered. For recruits with low literacy in Chinese or Italian, a mother tongue assistant is available to help participants complete the questionnaires. Printed or PDF versions of the questionnaires are available.

The lifestyle questionnaire

The core questions were taken from pre-existing validated questionnaires20 21 and also the WHO STEPS (STEPwise approach to chronic disease risk factor surveillance questionnaire).16 Behaviours relating to the preceding year are usually enquired about, but for some behaviours (eg, tobacco use) lifetime history is asked about. From May to October 2021, the lifestyle questionnaire was tested on a convenience sample of Italians to assess understanding and compilation times. To minimise the data input burden to participants, it was decided that the questionnaire should take less than 30 min to complete; to achieve this predefined categorical variables were used whenever possible and questions found to cause difficulties were simplified.

From October 2021 to February 2022 the questionnaire was tested on a convenience sample of 50 Chinese volunteers and study assistants. In cooperation with the CAB, the questions and replies were assessed for acceptability and clarity. Because several Chinese volunteers had difficulties with written Chinese, they preferred to complete the Italian version. As a result, the Italian version was simplified and retested on another group. The final version of the questionnaire has 13 sections: 01 general questions; 02 physical activity; 03/04 reproductive history (men and women); 05 tobacco use; 06 second-hand smoking; 07 personal medical history and attitudes to health; 08 alcohol consumption; 09 fruit and vegetable consumption; 10 salt consumption; 11 sleep; 12 cancer screening; and 13 willingness to participate in future studies.

The lifestyle questionnaire is designed to be completed by the participant online, using a tablet, during the recruitment session. If a participant is unable to use a computer or has other difficulties, the questions can be read out by an assistant who fills in the questionnaire on the participant’s behalf. For participants who prefer later compilation, the questionnaire is sent to the participant via an SMS link. In this case too, an assistant can be contacted who can help complete the questionnaire via telephone in the event of difficulties. The mode in which the lifestyle questionnaire was completed (with/without help, at the recruitment centre/later, online/on paper) is recorded.

The food frequency questionnaire

The FFQ for Chinese people living in Italy is still being developed. It will enquire about food consumed over the year up to compilation. It is derived from the T-FFQ—an instrument validated specifically on the population of Taizhou city, Jiangsu province, China.22 Most of the Chinese who emigrated to Milan come from Wenzhou city (Zhejiang province). Weinzhou city is geographically close to Taizhou but eating habits differ. We originally planned a Taizhou-Wenzhou adaptation of the T-FFQ followed by a Wenzhou-Milan adaptation. However, because of COVID-19, the Taizhou-Wenzhou adaptation proved impossible, so we are adapting the T-FFQ directly to the Milan-Chinese diet. This ongoing project comprises the following phases:

  1. An open-ended face-to-face 24-hour diet recall (24hdr) interview with a sample (n=50) of Chinese living in the Milan area, conducted by a trained mother-tongue assistant. Foods (not their quantities) recalled in the 24hdr and not included in the T-FFQ are added to the T-FFQ food list.

  2. Food items from the validated EPIC-Italy FFQ23 that are not present the T-FFQ food list are added to it.

  3. A prototype FFQ is developed based on the updated T-FFQ food list. Original images from T-FFQ are used. New photographs of food items/dishes are added as required.

  4. The new prototype FFQ is validated on Chinese residents of Milan of varying age and socioeconomic status: first the 24hdr is administered telephonically to each resident by the trained mother-tongue assistant. Two-to-four weeks later, the new prototype FFQ is administered to the same residents, again telephonically by the trained mother-tongue assistant. Only frequencies of consumption (not quantities) are assessed and compared between the two instruments. The repeatability of the prototype FFQ is tested by administering it again 1–3 months after first administration and comparing the results of the two tests.

  5. We expect to be able to use the new FFQ, validated as above, by March 2024. It will be administered by telephone interview to CHINT recruits who have previously received the food list with photographs of dishes.

Strategies to improve the representativeness of study population

To monitor the representativeness of the study population, as it is being recruited, we periodically compare the recruited sample distribution by sex, age and occupation type with distributions from official data on Chinese immigrants in Italy.7 Comparison at the beginning of 2023 showed that the CHINT sample contained 59% women (against 51% in the entire community), that 7% were under age 30 years (against 21%), that 88% of the recruited men were employed (against 79%), and that 75% of the recruited women were employed (against 60%). Furthermore blue collar/manual workers were under-represented in our sample (22% of employed persons against 33% in the community).

Based on these findings and after consultation with the CAB, we are making efforts to recruit more blue collar males, more males<30 years, and also more retired and unemployed people. To do this, we are approaching Chinese associations and Chinese businesses and asking them to send out invitations (to join the study) to individual members and employees specifically belonging to these categories. These initiatives will continue into 2024.

Maximising recruitment, compliance and retention

A campaign to inform about the study began in September 2021 and continued through public meetings, the dissemination of leaflets and brochures, and the showing of short films on local TV channels and Chinese outlets. The roles of the CAB and Chinese business leaders in facilitating reaching out to potential recruits and facilitating recruitment have been described above. To further encourage a wide range of people to join the following measures have been implemented.

Easing logistic hurdles

Recruitment offices are placed close to where Chinese immigrants are to be found: the Chinatown neighbourhood, and Milan suburbs where Chinese blue collar workers live, work and congregate. Recruitment offices are open on days and at times compatible with Chinese working hours and Chinese holidays (as communicated by the CAB). The Chinatown recruitment office is open at convenient times so that people can drop-in without appointment for information and also to be recruited.

If people are pressed for time at the recruitment meeting, after the measurement session (15–20 min) the lifestyle questionnaire can be completed elsewhere later on.

For people who are computer illiterate or have difficulty in reading, questionnaire replies can be dictated to a Chinese mother tongue assistant, either face-to-face or telephonically.

Incentives to individuals

A personalised feedback report is provided to recruits (by SMS link) after completion of at least 80% of the core assessment (physical measurements and lifestyle questionnaire). The following personal data are contained in the report: alcohol consumption, tobacco consumption, fruit and vegetable consumption, anthropometry (extent of overweight and abdominal obesity), blood pressure and muscle strength. These personal characteristics are compared with Italian and Chinese recommendations for healthy levels of these characteristics. The reports also include links to sites that provide more information and suggest how to improve personal risk profile. The sites are Italian (Italian Ministry of Health, CREA), Chinese (Chinese Health Mystery and Chinese CDC) and international (WHO, WCRF, FAO).

Incentives to the community

Criticisms raised by the Chinese associations revolved around the perception that CHINT was not immediately useful to participants or the community: ‘You are measuring us but if we have a health problem you do not help us solve it’ or ‘You talk about disease prevention, but do not explain how to implement disease prevention practices’. The concern was also voiced that access to treatments was difficult because of linguistic and cultural barriers. To go some way to responding to these concerns, two additional projects were set up to proceed in parallel with the main CHINT study: Prevention Afternoons and Help Desk for Chinese Cancer Patients.

Prevention Afternoons are meetings open to the public, but aimed at Chinese, in which INT epidemiologists and oncologists provide information on the prevention of NCDs: the focus is on reducing NCD risk factors, including high blood pressure, which was found to be particularly common in the Chinese population that participated in the CHINT pilot study. The presentations are bilingual (Italian/Chinese) and organised in collaboration with the CAB. Chinese mother tongue assistants are on hand to clarify difficult points and help participants to ask questions.

The Help Desk for Chinese Cancer Patients is a project funded by the INT Scientific Director’s office in collaboration with the CAB. The Help Desk will be located at the INT, and will be designed and publicised to attract to Chinese patients and their families who attend the INT (major cancer hospital in Milan). A trained Chinese mother tongue assistant will welcome patients, give advice about cancer treatment and intercede on their behalf with INT doctors.

Data security

The data management and data security procedures for CHINT were approved by the INT Data Protection Office. Signed informed consent forms and questionnaire responses on paper are stored in a locked filing cabinet in PI’s office.

The PIN assigned to each participant at recruitment is the only personal identifier in the database. The database is stored on a password-protected server. Database integrity is ensured by daily backups. Keys to allow the association of survey data with recruit names and contact details are stored in a separate offline password-protected database as prescribed by Italian data protection legislation (GDPR 2016/679 and legislative decree No. 196/2003).

Quality assurance and analysis of data

The data entry platform is designed such that inadmissible values and inconsistencies cannot be entered. For example, a history of smoking (years being a smoker) must be congruent with participant age and the declared dates (years) of starting and stopping smoking. If an inconsistency occurs during compilation, input is blocked and a native speaker mediator is alerted in order to clarify things with the participant, and resolve the inconsistency.

At the end of the compilation, the programme points out any questions that have not been answered and alerts a mediator to help with completion. These features massively reduce input errors and the need for data checking.

So far missing values are due to loss of connection during data entry, or computer/tablet malfunction, some system/equipment failure, or unsatisfactory technicalities while recording the values. As missing values are likely to be missing at random, people with missing values are not excluded from the analyses of non-missing variables and no imputing procedures will be applied. Only subjects who have missing information regarding the study inclusion criteria (age, sex, arrival in Italy) will be excluded from the analyses.

Descriptive statistics will be used to summarise participant characteristics and exposure variables (frequencies of consumption of foods and beverages, tobacco and alcohol; physical activity, and physical measurements). The exposure variables will be analysed as continuous variables and collapsed into dichotomous groups (healthy vs unhealthy) where appropriate. Multivariable linear regression modelling will be used to identify characteristics (eg, age, education, occupation, time in Italy) associated with exposures. Multivariate logistic regression models will be used to analyse characteristics associated with a healthy versus an unhealthy condition. Variables such as socioeconomic status, confidence in the Italian national health service, confidence in western medicine, migration history, and so on, will be analysed to identify predictors of willingness to enter future intervention studies. The analyses will be performed using Stata V.16.0 (Stata Corp); p<0.05 will be considered significant.

Current status and study timeline

Recruitment to CHINT started in March 2022. We expect to complete baseline recruitment by September 2024 and to have sent out FFQs for remote compilation by August 2024. The study Gantt chart is shown in figure 2.

Figure 2
Figure 2

Gantt chart.

Ethics and dissemination

The CHINT study is being conducted in compliance with the latest revisions of the declaration of Helsinki24 and the declaration of Oviedo.25 The study protocol was approved by the INT ethical committee (INT Prot.147/21). Study documentation is first produced in Italian, translated into English and Chinese (simplified) and then back-translated in Italian. Initially, project findings will be disseminated to the scientific community by publications in scientific journals and presentations at scientific conferences. The findings will also by communicated to lay people, and particularly the Chinese immigrant community, through freely accessible newsletters, brochures, social media, seminars, articles and public meetings.

Discussion

The ongoing CHINT study is the first investigation exploring cancer risk factors conducted on Chinese people living in Italy. The prevalences of obesity, smoking, heavy drinking, physical inactivity, unhealthy diet, high blood pressure, obesity and/or abdominal obesity, and adherence to cancer screening programme are being investigated together with associated personal and socioeconomic characteristics.

The main strength of the study is that it has enlisted the active cooperation of stakeholders within the Chinese community: The CAB is involved not only in ensuring that study instruments are acceptable to participants, but also in reaching out to specific sectors of the population (eg, persons of lower socioeconomic status) who are rarely recruited to epidemiological studies.

The composition of the sample is being monitored during recruitment to verify the representativeness of the study population by comparison with official data on Chinese immigrants in Italy.7 Subsequently participants from less represented subgroups are actively sought to render the sample as representative as possible. This strategy is expected to overcome, to some extent, the main study limitation: that a random sample of the Chinese community is not being recruited. This limitation implies that it will not be possible to exclude bias resulting from a ‘volunteer effect’.

In designing the study, we adhered to the CAB’s advice to use the least invasive assessment methods possible. We therefore decided not to collect the biological samples that would provide additional information on risk factors in the target population. However, we are confident that a study that ascertains lifestyle and performs physical measurements will enable us to reliably identify modifiable risk factors for NCDs in the population and hence indicate what priorities need to be addressed in subsequent intervention studies to reduce NCD risk factors. Such intervention studies will be facilitated by the trust that has been created within the Chinese community as a result of this present study.

The purpose of this survey is to provide a fairly unbiased glimpse of an unknown reality: for the first time to identify modifiable risk factors for NCDs in a distrustful population with literacy problems and presumed health problems. Study results will be disseminated not only to the scientific community but also at meetings to which the CAB and the entire Chinese population residing in Lombardy will be invited. We are confident that a study that ascertains lifestyle and performs physical measurements will reliably identify modifiable risk factors for NCDs and hence indicate what needs to be done in subsequent intervention studies to reduce NCD risk in this population.

Acknowledgments

We sincerely thank the participants of the CHINT study. We also thank the members of the Union of Chinese associations of Milan who made it possible for us to enter the heart of the Chinese community and render the study possible. We are indebted to Dr. Maria Rosa Azzolina and Dr. Lidia Casti of “Istituto Italo Cinese Aps”, to Ms. Suping Huang of the “Associazione delle donne cinesi a Milano” and also to Dr. Cai Maosheng from the “Association of Alumni of the Zhejjang University in Italy” who contributed importantly to the design, method development and the conduct of the study. We are also indebted to Zhao Jianfeng, Federica Lamberti, Hua Jiao, Zhang Haibo, Du Xingling, Ye Lingling, Liu Xin, Jin Nan, Eros Micheli, and Jiang Zhewen for their cultural and linguistic mediation during recruitment. We also thank Dr. Giuseppe Calabrò, medical director of the “Istituto Scientifico Euromedica” and his staff for their continuous support in the recruitment and involvement of Chinese patients in the study. Finally, we thank Don Ward for help with the English of the published paper.

This post was originally published on https://bmjopen.bmj.com