Differences in modifiable cancer risk behaviors by nativity (US-born v. Non-US-born) and length of time in the US

Previous studies have identified racial-ethnic disparities in modifiable risk factors for cancers. However, the impact of US nativity on these risks is understudied. Hence, we assessed the association between US nativity and length of time in the US on modifiable cancer risk factors. Utilizing the 2010 and 2015 National Health Interview Survey datasets, we analyzed 8,861 US-born and non-US-born adults. Key variables included age, sex, race-ethnicity, education, income, diet, body mass index, physical activity, alcohol consumption, and smoking. Statistical methods included descriptive statistics and regression. Most respondents were US-born (n = 7,370), followed by long-term (≥15 years, n = 928), and recent (<15 years, n = 563) immigrants. Moderate-to-vigorous physical activity was higher among US-born individuals (342.45 minutes/week), compared to recent (249.74 minutes/week) and long-term immigrants (255.19 minutes/week). Recent immigrants consumed more fruits (1.37 cups/day) and long-term immigrants more vegetables (1.78 cups/day) than US-born individuals. Multivariate analyses found recent immigrants had lower odds of consuming alcohol (AOR: 0.33, 95% CI: 0.21–0.50) and smoking (AOR: 0.30, 95% CI: 0.19–0.46), and higher odds of meeting fruit consumption guidelines (AOR: 2.80, 95% CI: 1.76–4.45) compared to US-born individuals. Long-term immigrants had lower odds of alcohol consumption (AOR: 0.56, 95% CI: 0.37–0.84) and smoking (AOR: 0.42, 95% CI: 0.30–0.59), and higher odds for meeting fruit (AOR: 1.87, 95% CI: 1.22–2.86) and fiber (AOR: 2.03, 95% CI: 1.02–4.05) consumption guidelines. Our findings illustrate the importance of considering the impact nativity and length of US residency has on health. Our findings underscore the need for culturally tailored public health strategies.


Introduction
Cancer disparities and inequities in the United States (US), such as differences in cancer incidence, mortality rates, and access to screening and treatment, have resulted in increased cancer

Data source and sample
Our study utilized merged 2010 and 2015 National Health Interview Survey (NHIS) datasets as these selected years included additional questions regarding cancer prevention and control from Cancer Control Supplement.The datasets were obtained via the IPUMS Health Surveys website, and dietary variables from National Cancer Institute (NCI) Epidemiology and Genomics Research Program website.These dietary variables were already computed using algorithms developed by their staff [20,21].These algorithms were specifically designed to convert screener responses from the NHIS dietary questionnaire into quantitative estimates of individual dietary intake.The method of estimation relied on predictive modeling, correlating NHIS screener responses with the detailed dietary data from NHANES.This correlation ensured a robust and validated approach to dietary intake assessment, leveraging NHANES as the reference standard for dietary information [20,21].The key dietary variables analyzed included fruits, vegetables, dairy (cup equivalents), added sugars (teaspoon equivalents), whole grains (ounce equivalents), and fiber (grams) as guided by the ACS Guideline for Diet and Physical Activity for Cancer Prevention [20,21].The NHIS is a data collection program conducted by the National Center for Health Statistics (NCHS), which is part of the US Centers for Disease Control and Prevention (CDC) [22].The NHIS collects data through continuous face-to-face household interviews and covers a wide range of health topics, with the survey content and structure periodically redesigned to meet evolving data needs [22].The Research Ethics Review Board of the National Center for Health Statistics approved these NHIS protocols [22].Our analyses were composed of adult respondents (aged 18 years and older) with a response for all variables included in our analyses of nativity and length of residence in the US (N = 8,861, unweighted).

Study variables
The main dependent variables were modifiable risk factors associated with cancer including physical activity, diet, alcohol consumption, smoking status, and body mass index (BMI).The main independent variable was US nativity.Following methods from prior studies using NHIS data, adults born in the US were defined as those originating from any of the 50 states, the District of Columbia, or a US territory and we categorized adults born outside the US and its territories as foreign-born [23][24][25][26][27]. Foreign-born individuals were further categorized as immigrants residing in the US for <15 years (recent immigrants), and �15 years (long-term immigrants), this endpoint is consistent with prior research using the NHIS dataset to measure nativity and length of time in the US as a proxy for time needed to see maximal effects of acculturation [26,27].In this study we used dietary, exercise, smoking status, alcohol consumption, BMI, and sociodemographic characteristics (age, sex, race-ethnicity, educational attainment, employment status, annual household income) variables.Details on operationalization of all the variables included in this study are depicted in Table 1.
The ACS Guideline for Diet and Physical Activity for Cancer Prevention is structured into four principal domains [11].The first domain is achieving and maintaining a healthy weight throughout life, which we operationalized in our study by defining adherence as a normal weight (measured as a BMI between 18.5 and 24.9).The second domain is physical activity, recommending adults engage in 150-300 minutes of moderate-intensity or 75-150 minutes of vigorous-intensity exercise weekly.Our study interpreted adherence to this domain as obtaining a minimum of 150 minutes of moderate-to-vigorous physical activity (MVPA) per week.The third domain pertains to dietary habits, promoting a healthy eating pattern across all ages.This includes a diet rich in nutrients that aid in maintaining a healthy body weight, with an emphasis on a variety of vegetables, fruits, and whole grains, and limiting or avoiding red and

RACE
Race-Ethnicity "What race(s) do you consider yourself to be?" and "Are you Hispanic or Latino?"RACENEW provides self-reported race information using the 1997 OMB standards for race and ethnicity classification.
1 = "Non-Hispanic White", 2 = "Non-Hispanic Black", 3 = "Hispanic", 4 = "Afro-Latin", 5 = "Non-Hispanic Asian" Derived from RACENEW and HISPYN SEX Sex "Are you male or female?"SEX indicates the person's sex as male or female, with options for "Unknownrefused" and "Unknown-don't know" starting in 2019.processed meats, sugar-sweetened beverages, and highly processed foods.In our analyses for dietary assessment, we analyzed the consumption patterns of various food groups, including fruits, vegetables, whole grains, fiber and added sugars, in the context of the guidelines for diet and physical activity for cancer prevention set forth by the ACS [11].In our analysis, adherence with this domain was assessed by daily consumption of fruits, vegetables, fiber, whole grains, and added sugars-measured by consumption of at least 2 cups of fruits, 2.5 cups of vegetables, 30 grams of fiber, 3 ounces of whole grains, and less than 12 tsp of added sugars per day.Lastly, the fourth domain advises avoiding or limiting alcohol consumption, categorized in our study as "ever" versus "never" drinkers.We also compared "ever" and "never" smokers to assess smoking status, considering its known association with cancer.

Statistical analyses
All analyses conducted accounted for the NHIS complex survey design by applying appropriate sampling weights provided by NHIS.All statistical tests were two-tailed, with an alpha <0.05 considered statistically significant, and SAS version 9.4 (SAS Institute, Inc., Cary, NC, US) was used for statistical analysis.Each model was fitted using stepwise selection and Variance Inflation Factor (VIF) was assessed testing multicollinearity of variables.The results from the stepwise selection analyses indicated diverse variables that were retained for each model.Therefore, we employed a systematic approach in selecting variables uniformly across all adjusted multivariable models.Key variables such as region of birth, length of time in the US, annual income, race-ethnicity, gender, educational attainment, age, and health insurance coverage were used in all models to maintain consistency.This choice was guided by the study aims, theoretical considerations relevant to public health research, and methods employed by previous studies on nativity and length of time in the US.The robustness and appropriateness of this strategy were affirmed by the significant results (p-values <0.05) from F-tests and Wald Chi-Square tests across all models for all retained variables, along with numerous significant associations observed in each model [23,[25][26][27][28].The absence of multicollinearity was confirmed by the VIF values (all <2.5), further validating reliability of the models.There were no missing variables for all analyses conducted in this study.We used survey year variables from NHIS datasets in our multivariable analyses for both survey cycles.This adjustment helped account for varying impacts of survey years on dependent variables between 2010 and 2015, reducing confounding effects caused by temporal changes in our outcome variables.Descriptive statistics and chi-squared tests were employed to describe the sociodemographic characteristics and modifiable risk factors of the sample, stratified by nativity and length of residence.Additionally, sample means with confidence limits for continuous variables (age, daily dietary intake, BMI, and weekly minutes of moderate, vigorous, and moderate-to-vigorous physical activity) were computed and were compared between groups.
Univariate and multivariable binary logistic regression models were used to examine the impact of nativity with and without controlling for sociodemographic characteristics on modifiable risk factors based on the ACS cancer prevention guidelines [11].The dependent variables for each individual model were BMI, physical activity, smoking status, alcohol consumption, and dietary consumption patterns (fruit, vegetable, fiber, whole grains, and added sugar consumption).Independent variables included nativity and length of residence in the US, race-ethnicity (Black Hispanic, White Hispanic, non-Hispanic Black, non-Hispanic Asian, and non-Hispanic White), age groups (18-34, 35-54, and 55+ years), educational attainment (less than high school education, high school education, some college, and college graduate), sex (male and female), health insurance status (yes or no), and annual family earning brackets (<$25,00, $25,000 to $55,000, 55,000 to $74,999, and $75,000+).
When examining the region of birth by immigrant status and duration in the United States, the data revealed distinct patterns among US-born individuals, recent immigrants (those with less than 15 years in the U.S.), and long-term immigrants (those with 15 or more years in the U.S.).Unsurprisingly, all US-born respondents (100%) originated from the United States.Among immigrants, those from Mexico, Central America, and the Caribbean Islands constituted the majority, with 57.0% in the recent immigrant category and 61.6% in the long-term immigrant category.This was followed by immigrants from Asia, who accounted for 21.4% of recent immigrants and 18.0% of long-term immigrants, reflecting a significant presence in the U.S. immigrant population.Notably, South America, Europe, Russia (and former USSR areas), Africa, and the Middle East also contributed to the U.S. immigrant landscape, albeit in smaller proportions.For instance, South American-born individuals represented 6.6% of In terms of gender, recent immigrants had a higher proportion of males (69.6%) compared to US-born (58.3%) and long-term immigrants (67.0%).Females were more prevalent among US-born (41.8%) and long-term immigrants (33.1%) than recent immigrants (30.4%).
Education levels varied.Recent immigrants had the highest percentage of individuals with less than high school education (27.9%), followed by long-term immigrants (28.1%) and USborn individuals (6.6%).College graduates were evenly distributed across all groups, with USborn individuals (31.0%) slightly leading.
Annual income differences were present, with a higher proportion of recent immigrants (55.9%) earning less than $25,000 compared to US-born individuals (40.1%).In contrast, a larger percentage of US-born individuals (13.5%) earned over $75,000 compared to recent (10.8%) and long-term immigrants (12.7%).
Health insurance coverage was significantly lower among recent immigrants (55.9% without coverage) compared to US-born (84.7% with coverage) and long-term immigrants (74.4% with coverage).Physical activity levels showed that a lower percentage of recent immigrants (42.7%) met the criteria compared to US-born individuals (55.0%) and long-term immigrants (46.5%).
US-born individuals had the highest percentage of ever-smokers (40.1%), while recent immigrants had the lowest (22.1%).
Dietary factors such as fruit, vegetable, whole grain, added sugars, and fiber intake showed minimal differences across groups, except for fruit consumption where recent (86.3%) and long-term immigrants (84.4%) met the criteria more than US-born individuals (77.0%).
Mean age, daily dietary intake, weekly minutes of physical activity, and body mass index (BMI) by nativity and length residence in the US, NHIS 2010 and 2015 (N = 8,861, unweighted).When assessing sample means from Table 3, US-born individuals had a mean age of 37.24 years (95% CI: 36.78-37.69),which was higher than the immigrants who had been in the US for less than 15 years, with a mean age of 33.17 years (95% CI: 32.18-34.16),but lower than those who had been there for �15 years, with a mean age of 42.38 years (95% CI: 41. 41-43.35).
In terms of daily dietary intake, US-born individuals consumed fruits at a mean of 1.11 cups/day (95% CI: 1.09-1.13)and vegetables at 1.63 cups/day (95% CI: 1.61-1.65).Immigrants who had been in the US for less than 15 years consumed more fruits, at 1.37 cups/day (95% CI: 1.27-1.47),and vegetables, at 1.79 cups/day (95% CI: 1.70-1.88).Immigrants residing in the US for �15 years consumed fruits at 1.27 cups/day (95% CI: 1.20-1.33)and vegetables at 1.78 cups/day (95% CI: 1.72-1.84).For fiber intake, US-born participants consumed an average of Univariate and binary multivariable logistic regression analyses.The results of the binary multivariate analyses showed differences based on immigrant status by ACS adherence variables (Table 4).
In terms of smoking status, both recent and long-term immigrants had lower odds compared to US-born individuals in both univariate and multivariate models, with recent immigrants having an OR of 0.43 (95% CI: 0.33-0.55)and an AOR of 0.30 (95% CI: 0.19-0.46),and long-term immigrants having an OR of 0.65 (95% CI: 0.54-0.80)and an AOR of 0.42 (95% CI: 0.30-0.59).
For meeting fruit consumption guidelines, both recent and long-term immigrants had higher odds in both models, with recent immigrants showing a significant increase in the multivariate model (AOR: 2.80, 95% CI: 1.76-4.45).
Vegetable consumption guidelines adherence was not statistically significant for recent immigrants, but long-term immigrants had higher odds in the univariate model (OR: 2.84, 95% CI: 1.78-4.51).The findings were not statistically significant in the multivariate model for either group.
Both recent and long-term immigrants had higher odds of adherence to fiber consumption guidelines in the univariate model, but only long-term immigrants had significant findings in the multivariate model (AOR: 2.03, 95% CI: 1.02-4.05).No significant differences were observed for meeting added sugar and whole grain consumption guidelines in the multivariate models for both recent and long-term immigrants.However, long-term immigrants had increased odds of meeting whole grain guidelines in the univariate model (OR: 1.08, 95% CI: 0.70-1.67).

Main findings
We found nativity and length of residency in the US impacted health-related behaviors associated with cancer risks including physical activity, smoking status, alcohol consumption, and  dietary patterns.Prior studies have established that modifiable risk factors including physical activity, diet, alcohol consumption, smoking status, and BMI significantly influence cancer risk [11,[29][30][31][32].Our study extends this understanding by showing that these risk factors vary among immigrants based on nativity and length of stay in the US, highlighting the need for culturally tailored cancer prevention strategies.We observed that longer US residency often leads to the adoption of local health behaviors, a phenomenon known as the "acculturation", which modifies immigrants' cancer risk profiles [33].This may be attributed to the stark income disparities observed among recent immigrants compared to US-born individuals, as highlighted in the presented findings, underscore the critical economic challenges foreignborn US residents face.
A surprising finding in our study the difference in health insurance coverage among nativity groups, which contrasts with past research noting a coverage gap between recent immigrants and US-born individuals.Contrary to expectations, recent immigrants (<15 years in the US) have higher health insurance coverage compared to US-born individuals (15.3%) and long-term immigrants (�15 years in the US) (25.6%) [34].This discrepancy may be due to our sample of recent immigrants benefiting from specific insurance programs or employer-based coverage, as well as their proactive efforts to secure health insurance as part of their integration process.Conversely, the high percentage of uninsured US-born individuals (84.7%) indicates significant coverage challenges that could exacerbate health disparities and negatively impact cancer prevention and outcomes [34].
A key finding is the distinct behavioral patterns among immigrants, particularly those who have resided in the US for shorter durations.These patterns include variations in dietary habits, physical activity levels, and substance use (smoking status and alcohol consumption), which collectively contribute to varying cancer risks compared to US-born individuals.We observed that recent immigrants tended to have healthier dietary patterns and lower substance use in our models, as found in previous research, suggesting an overall protective effect against cancer risk [35].However, this effect appears to diminish with longer residence in the US, indicating a potential impact of acculturation on health behaviors.This is consistent with findings by Gubernskaya et al, who observed an increased risk of obesity and deteriorating health among Hispanic immigrants in the US for longer periods of time [36].
The "Hispanic/Latino Paradox," originally formulated in 1986, describes the lower all-cause mortality rates experienced by Hispanics/Latinos, despite their higher poverty rates, lower educational attainment, and lower health insurance coverage rates compared to non-Hispanic Whites [37][38][39][40][41].This relative advantage has been attributed to a combination of several factors and cultural norms, such as "familismo," which is a protective mechanism from unfavorable health outcomes among Hispanics compared to non-Hispanic Whites [37,42].However, this paradox may not uniformly apply across all Hispanic/Latino subgroups.The variance among subpopulations, such as differences in health outcomes between foreign-born and US-born Hispanics, as well as among Afro-Latinos and Afro-Caribbeans, is often overlooked.These subgroups exhibit distinct socio-economic, cultural, and health profiles, potentially leading to different health outcomes and access to healthcare services [43][44][45][46][47].This was evidenced by findings from Pinheiro et al., 2011 who explored cancer survival patterns related to this paradox and found no evidence of "Hispanic advantage" found in cancer survival [42].Notably, their findings suggested that by grouping Hispanics as a homogenous group, rather than by subgroup, existing survival disparities are missed [42].Similarly, studies focusing on other racial-ethnic groups, such as Asians and Blacks, have also revealed significant intragroup heterogeneity, further emphasizing the need for disaggregated data in health disparities research [43,48,49].This heterogeneity among racial-ethnic subpopulations has been studied among other immigrant racial-ethnic groups including Asians and Blacks and has been coined the "healthy immigrant paradox" in past studies [50][51][52][53][54]. Therefore, exploring inter-and intra-group differences (i.e.Black Hispanic, White Hispanic, foreign-born, US-born) across different racial-ethnic subpopulations is a crucial endeavor to truly understand the modifiable risk factors as well as behavioral characteristics to mitigate health disparities in the US [53].

Similarities and/or differences with findings from other studies
Our results corroborate previous findings that foreign nativity influences cancer risk in the US, as demonstrated by McRoyet al. (2018) who found that immigrants had better health outcomes compared to their US-born counterpoint notwithstanding access to fewer resources, thus confirming the healthy immigrant paradox [50].Interestingly, our study adds to the existing literature by emphasizing the role of duration of US residence in shaping cancer risk profiles.This finding aligns with Gallo et al. (2009), who noted that the health advantages associated with being non-US-born tend to wane with increased time spent in the US [55].Further, our findings align with Miller et al. (2018), who found significant variations in cancer incidence based on nativity and length of residence, highlighting the importance of assessing the heterogeneity of the US Hispanic population beyond ethnicity [56].Our study thus bridges a critical gap in understanding how acculturation affects health outcomes, particularly concerning cancer risk.

Clinical and public health implications from our findings
Our findings affirm the importance of considering nativity status and length of US residence in cancer risk assessment and prevention strategies.Healthcare providers should be aware of these variations when screening for cancer risk factors [57].Public health initiatives should aim to develop targeted health promotion and cancer screening programs that are culturally sensitive and accessible to diverse populations, including recent immigrants.

Strengths and limitations from our findings
Although our study provides a nuanced understanding of the interplay between nativity, duration of residence in the US, and modifiable risk factors, it has some limitations.Firstly, the reliance on self-reported cross-sectional data may introduce biases and limits our ability to establish causal relationships, and the tracking of behavioral changes over time.Further, we may not have captured all factors influencing cancer risk, such as undocumented lifestyle modifications post-immigration, cultural practices that affect health behaviors, mental health aspects which influence of stress-related factors on cancer risks among immigrant communities, and H. pylori infection (which was been shown to have significant association with foreign nativity and the incidence of gastric cancer) [57].Moreover, we were not able to distinguish between different categories of non-US-born participants, such as temporary visa holders, permanent residents, and disaggregate the precise country of origin as countries were grouped by region.This lack of detail prevents a nuanced analysis of how varying immigration statuses might influence stress levels, resource access, and consequently, cancer risk behaviors.Additionally, the use of self-reported race is limited by its subjective nature, as individuals may interpret their racial identity differently depending on their country of origin and cultural norms [58].
Despite these limitations, our study is comprehensive due to its analysis of a large, diverse population that is largely understudied.The NHIS data set provided robust data on modifiable behaviors implicated in cancer risk and detailed information regarding nativity.

What makes our study novel
The impact of nativity and acculturation on diet, exercise, obesity, smoking, and alcohol usage in not well understood.Our analysis adds valuable contributions to the existing literature related to cancer risks and migrant US populations by presenting evidence of the changes in health behaviors among immigrants over time, emphasizing the role of acculturation in the adoption of behaviors associated with increased cancer risk particularly less fruit consumption and substance use.

Importance and relevance of the study
This knowledge is crucial for health policymakers and practitioners who work with immigrant communities and minorities.Foreign born populations make up a significant proportion of minority subpopulations living in the US [13,59,60].Therefore, there is a need for culturally tailored health promotion strategies that can address the unique challenges faced by immigrants in maintaining healthy behaviors throughout their years residing in the US such as health insurance coverage and access to health care services.

Future directions
Future research should explore the impact these modifiable risk factors may have on cancer mortality.Additionally, longitudinal studies should be employed to track the evolution of health behaviors among immigrants and explore interventions that effectively support healthy lifestyle choices, and work to help immigrant maintain current healthy behaviors as they acculturate.Moreover, investigating the underlying factors driving these behavioral changes, such as socioeconomic status and cultural influences, would further enhance our understanding of health risks among immigrant populations.Finally, culturally targeted considerations for lifestyle interventions where dietary choices and exercise modalities which resonate with immigrant populations and their dynamic cultures must be created to aid in effective lifestyle modification interventions for these immigrant populations.

Table 2 . Distribution of sociodemographic and cancer risk factors by nativity in the US (NHIS 2010 and 2015, N = 8,861, unweighted).
recent immigrants and 4.8% of long-term immigrants.European-born immigrants were more prevalent among long-term residents (8.4%) compared to recent arrivals (3.9%), suggesting earlier waves of migration from Europe.Similarly, immigrants from Russia and former USSR areas, though small in percentage (0.6% recent, 1.3% long-term), added to the diversity of the immigrant population.African and Middle Eastern immigrants, while constituting a minor portion (African: 8.6% recent, 4.1% long-term; Middle Eastern: 1.9% for both recent and longterm), further illustrated the wide-ranging origins of the U.S. immigrant populace.

Table 4 . Univariate (OR, 95% CI) and Multivariate logistic regression models (AOR, 95% CI) for the association between US nativity/length of residence in US and non-adherence to American Cancer Society (ACS) diet and physical activity recommendations among respondents (N = 8,861, unweighted) from 2010 and 2015 National Health Interview Surveys (NHIS).
** US-Born served as the referent for all models.** Adjusted models were adjusted for nativity, region of birth, length of residence in the US among immigrants, race-ethnicity, age, education level, health coverage, and income level.This adjustment is applied to all variables except the specific variable under examination in each instance of interpretation.** Bolded values are significantly different from US-Born respondents (referent group)