Neighborhood disadvantage and lung cancer risk in a national cohort of never smoking Black women

Abstract

Background:

Compared to women of other races who have never smoked, Black women have a higher risk of lung cancer. Whether neighborhood disadvantage, which Black women experience at higher rates than other women, is linked to never-smoking lung cancer risk remains unclear. This study investigates the association of neighborhood disadvantage and lung cancer risk in Black never-smoking women.

Methods and materials:

This research utilized data from the Black Women’s Health Study, a prospective cohort of 59,000 Black women recruited from across the US in 1995 and followed by biennial questionnaires. Associations of lung cancer incidence with neighborhood-level factors (including two composite variables derived from Census Bureau data: neighborhood socioeconomic status and neighborhood concentrated disadvantage), secondhand smoke exposure, and PM_2.5 were estimated using Fine-Gray subdistribution hazard models.

Results:

Among 37,650 never-smokers, 77 were diagnosed with lung cancer during follow-up from 1995 to 2018. The adjusted subdistribution hazard ratio (sHR) of lung cancer incidence with ten unit increase in neighborhood concentrated disadvantage index was 1.30 (95 % CI: 1.04, 1.63, p = 0.023). Exposure to secondhand smoke at work was associated with increased risk (sHR = 1.93, 95 % CI: 1.21, 3.10, p = 0.006), but exposure to secondhand smoke at home and PM_2.5 was not.

Conclusion:

Worse neighborhood concentrated disadvantage was associated with increased lung cancer risk in Black women who never smoked. These findings suggest that non-tobacco-related factors in disadvantaged neighborhoods may be linked to lung cancer risk in Black women and that these factors must be understood and targeted to achieve health equity.

1. Introduction

Lung cancer accounts for 27 % of all cancer deaths worldwide and is the leading cause of cancer-related mortality in the United States (US), where nearly 150,000 lung cancer deaths occurred in 2019 [1]. Although most lung cancer cases are due to tobacco use, approximately 15–20 % occur in never smokers, who are defined as having smoked<100 cigarettes in their lifetime [2]. Unfortunately, the frequency of never smoker patients diagnosed with lung cancer has increased in the past decades. In the US, 17 % of the non-small cell lung cancer (NSCLC) cases diagnosed between 2011 and 2013 were in patients who never smoked, in contrast to 8.9 % from 1990 to 1995 [3].

Black women in the US who never smoked have a higher incidence of NSCLC compared with non-smokers of European and Asian descent [4,5]. Additionally, Black women who never smoked have higher mortality from NSCLC when compared to never smoking women of other races [6]. Despite these differences, there has been limited evaluation of the distinct etiologies that may drive non-smoking lung cancer risk in Black women. Specifically, neighborhood disadvantage is implicated in racial disparities across a variety of oncologic outcomes [7], but its impact on the incidence of lung cancer in Black women has not been well studied.

As a result of structural racism in the US, Black communities have disproportionately higher exposure to adverse neighborhood-level factors, many of which are linked to lung cancer risk (e.g. air pollution exposure, traffic proximity, low neighborhood socioeconomic status [SES]) [8,9]. Government-sanctioned racial residential segregation (redlining) and neighborhood disinvestment have created Black communities that were intentionally disadvantaged, possessing higher exposure to pollutants, crime, and deprivation while having lower access to healthy foods, green space and appropriate housing [10]. Studies have shown a link between these segregated neighborhoods and both lung cancer mortality and access to lung cancer surgical care in Black individuals [11,12], but there has been little investigation into the impact of neighborhood social conditions and lung cancer risk in Black individuals, especially among those who do not smoke.

Structural inequities have also contributed to decreased cumulative wealth and resources in the Black community. As such, Black families possess only 10 % of the wealth of non-Hispanic White (NHW) families and Black individuals live in poverty 3 times more frequently than NHW individuals [13,14]. Additionally, at the same levels of education and income, Black/African-American women are more likely than their NHW counterparts to live in neighborhoods of low SES [15,16]. Therefore, the fact that low SES for both individuals and neighborhoods has been positively associated with incidence of NSCLC [17,18], has significant importance in the Black community.

The link between elevated NSCLC risk in low SES/disadvantaged communities is assumed to be secondary to the fact that low SES individuals have higher smoking rates [8]. There has been little evaluation of whether the association between low SES/disadvantaged communities and increased NSCLC risk persists in never smokers. This finding would suggest that non-tobacco components within disadvantaged neighborhoods may drive this link. Some studies have evaluated the impact of neighborhood deprivation and lung cancer risk in cohorts that are predominantly smokers [8,19-21]. Singh et al found that individuals in more deprived areas or those groups with lower education and income had higher lung cancer incidence rates than their more affluent counterparts [19]. Hystad et al found that residents living in low SES neighborhoods had increased risk of lung cancer and that smoking was the primary mediator of this link [8]. Sanderson et al found that current and short-term former smokers in deprived areas had higher risk of lung cancer whereas never smokers in deprived areas did not have increased risk of lung cancer [20].

The findings from the Sanderson study, which is the only study to date to examine the role of neighborhood deprivation in lung cancer among non-smokers, suggest that neighborhood SES has no significant impact on the incidence of lung cancer in never smokers because smoking is the primary mediator of the link between low SES and lung cancer risk. However, the study was limited in its ability to focus on Black women never smokers because of their relatively small numbers. Because of the remaining gap in our understanding of whether neighborhood SES impacts the incidence of NSCLC in Black women who have never smoked, we investigated this association in the prospective data from the Black Women’s Health Study (BWHS). A deeper understanding of this relationship could improve our ability to accurately identify communities at elevated risk for lung cancer.

2. Materials & methods

2.1. Study population

We included all participants who enrolled in the Black Women’s Health Study (BWHS) and reported never having smoked on their baseline survey. Briefly, the BWHS enrolled 59,000 Black women in 1995 from across the US, recruiting from subscribers to Essence magazine and members of the Black Nurses’ Association. Enrolled women ranged in age from 20 to 70 years. Baseline questionnaires asked participants about demographic characteristics, medical history, and lifestyle factors, including a detailed smoking history. Participants completed biennial follow-up surveys to provide updates on major health conditions, lifestyle factors, and other exposure, including updates of smoking status. Lung cancer status was determined by self-report through the follow-up surveys and by linkage with state cancer registries and the National Death Index Plus. Confirmation of the diagnosis and information on tumor characteristics were obtained by review of medical records and/or cancer registry data. In this analysis, participants who indicated never having smoked at baseline were included from the initial enrollment period through follow up to the timepoint when their non-smoking status changed to smoking or up to December 31, 2018. BWHS activities, including a waiver of documentation of informed consent, were approved by the Boston University Institutional Review Board (IRB) and analyses were approved by the City of Hope IRB. This study followed Strengthening and Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.

2.2. Individual-level covariates

At baseline in 1995, participants reported their age, smoking status, marital status, body mass index (BMI), health status, and levels of education (high school degree or less, some college, college degree and above). Smoking status was updated at subsequent biennial surveys. In 1997, participants began reporting exposure to secondhand smoke at work and at home (being in a room with a smoker for at least one hour/day for ≥ 12 consecutive months), household income attainment and whether they had health insurance (yes, no).

2.3. Neighborhood-level covariates

Exposure to ambient air pollution and neighborhood-level social or economic factors was assigned based on the census tract in which the participant lived at baseline. Briefly, fine particulate matter<2.5 μm in diameter (PM_2.5) concentrations was estimated at all participant home addresses during participation from 1995 to 2018 using nationwide monitoring data from the U. S. Environmental Protection Agency and a two-stage modeling approach using both Bayesian Maximum Entropy and Land Use Regression, with a final model R² = 0.79 [22]. Monthly PM_2.5 estimates were generated at the baseline home address of each woman and averaged to estimate the mean exposure during the participation period.

Two neighborhood-level social factor scores were previously derived using factor analysis after linking BWHS participant addresses at time of enrollment to year 2000 US Census data at the block group level [23-25]. The neighborhood SES score is based on median household income; median housing value; percentage of households receiving interest, dividend, or net rental income; percentage of adults 25 years or older that have completed college; percentage of employed persons aged 16 years or older that are in occupations classified as managerial, executive, or professional specialty; and percentage of families with children that are not headed by a single female. Factor analyses were conducted and the factor loadings were used to weigh the variables and sum them for an overall neighborhood SES score, with higher scores signifying higher neighborhood SES [23].

The neighborhood concentrated disadvantage score [24,25] was based on factor analysis of six census variables at the block group level, including percentage of individuals below the poverty line; percentage of individuals on public assistance; percentage of female-headed households; percentage unemployed; percentage of individuals below age 18; and percentage of Black residents. Factor scoring coefficients from each factor analysis were used to weigh the six variables. The variables were summed to create a neighborhood concentrated disadvantage score. Higher neighborhood concentrated disadvantage score indicated greater disadvantage while higher neighborhood SES score indicated higher SES [23-25].

2.4. Statistical analysis

Using Fine–Gray models to account for death and initiation of smoking as competing events, we assessed the association of patient characteristics, environmental risk factors, and neighborhood-level social factors with the incidence of lung cancer. Participants accrued person-time beginning at the time of exposure assessment in 1995. Although income was collected in 2003 and secondhand smoking data in 1997, they were assumed to also apply in 1995. The follow-up ended at lung cancer diagnosis, death, initiation of smoking, or December 31, 2018, whichever came first. Thus, participants’ follow-up was censored at the time they noted initiation of smoking. We accounted for clustering of participants within census tracts by using a robust sandwich variance estimator of the covariance. Verified self-reported lung cancer was the event of interest. We tested for proportional hazards by creating interactions of the time-dependent predictors (such as age) and survival time in Fine-Gray regression models, which were not significant.

Individual-level variables considered included age and BMI, which were treated as continuous variables, secondhand smoke exposure at home and at work (yes, no), health insurance status (yes, no), individual educational attainment (high school degree or less, some college, college degree and above), and individual income attainment (<$25,000, $25,000 - $50,000, $50,000 - $100,000, above $100,000). Neighborhood-level variables considered were annual PM_2.5 exposure (treated as continuous), and the two neighborhood-derived composite variables - neighborhood concentrated disadvantage (continuous and in quintiles) and neighborhood SES (continuous and in quintiles). In multivariable analysis, we included covariables known a priori to be associated with lung cancer risk (age, BMI, PM_2.5, individual educational attainment, individual income attainment, insurance status, and secondhand smoke exposure at home) regardless of their p-values, to assess the association between secondhand smoke at work, neighborhood concentrated disadvantage, and neighborhood SES, treated as continuous and categorical variables. Interactions between neighborhood-level variables and secondhand smoke exposure were also tested. Missing data were multiply imputed and the results were combined using the R (version 4.0.3) package MICE [26]. Subdistribution hazard ratios (sHRs) and 95 % confidence intervals (CIs) were calculated. All tests were 2-sided, and statistical significance was defined as p-value < 0.05. Statistical analyses were performed using SAS 9.4 (SAS Institute, Cary, NC).

3. Results

Of 59,000 BWHS participants, 37,650 (64 %) reported never having smoked at baseline and were included in the analysis. The mean age at enrollment of the never-smoking study cohort was 37 years. Of those, 77 (0.2 %) were diagnosed with lung cancer at a mean age of 59.41 ± 13.2 (range, 30–88). The median follow-up time for the study cohort was 22.6 years (range, 0.46 – 23.88). The demographic distributions of the study cohort are shown in Table 1. The mean BMI was 27.6 ± 6.7 (range 14–83). The percentage of the cohort exposed to secondhand smoke at home and at work was 68 % and 39 %, respectively. The percentage of individuals with a high school degree or less was 16 %, and 43 % reported an individual income under $50,000. The mean PM_2.5 was 16.0 ug/m³ ± 3.8 and the average neighborhood concentrated disadvantage and neighborhood SES scores were −0.55 ± 8.9 (Range: −19.76, 79.61) and 0.42 ± 9.4 (Range: −30.36, 45.72), respectively.

Table 1.

Demographic characteristics of the never-smoked study cohort.

	N = 37,650
Age at enrollment, Mean ± SD (Range)	36.8 ± 10.5 (20, 70)
Body mass index, Mean ± SD (Range)	27.6 ± 6.7 (14, 83)
Secondhand smoking, exposed at home
Yes, no. (%)	25,626 (68)
No, no. (%)	12,024 (32)
Secondhand smoking, exposed at work
Yes, no. (%)	14,566 (39)
No, no. (%)	23,084 (61)
Health insurance status
Yes, no. (%)	34,926 (93)
No, no. (%)	2,724 (7)
Individual Education
High school degree or less, no. (%)	5,999 (16)
Some college no. (%)	12,983 (34)
College or Graduate degree, no. (%)	18,668 (50)
Individual Income, no. (%)
<$25,000	4,528 (12)
$25,000 - <$50,000	11,691 (31)
$50,000 - <$100,000	14,835 (39)
$100,000 +	6,596 (18)
PM2.5, Mean (SD)	16.0 (3.8)
Neighborhood Concentrated Disadvantage Index, Mean ± SD (Range)	−0.55 ± 8.9 (−19.76, 79.61)
Neighborhood SES Index, Mean ± SD (Range)	0.42 ± 9.4 (−30.36, 45.72)

SD: standard deviation; PM_2.5: particulate matter < 2.5 μm in diameter; SES: socioeconomic status.

The results of age-adjusted univariable and multivariable analyses examining the association between lung cancer incidence and patient-, environmental-, and neighborhood-level risk factors are shown in Tables 2 and 3, respectively. In the age-adjusted univariable analysis (Table 2), the only variables considered statistically significant at p < 0.05 were neighborhood concentrated deprivation score (sHR = 1.30, 95 % CI: 1.03, 1.63) and secondhand smoke exposure at work (sHR = 1.93, 95 % CI: 1.21, 3.08).

Table 2.

Age-adjusted univariate analysis of individual and neighborhood-level factors and lung cancer incidence among 37,650 never-smokers in the Black Women’s Health Study.

Factor	NSCLC cases (N)	Age-adjusted sHR (95 % CI)
Age at enrollment	77	NA
Body mass index	77	0.99 (0.96, 1.03)
PM2.5, ug/m3 (continuous)	77	1.01 (0.95, 1.08)
Educational attainment
High school degree or less	20	1.05 (0.61, 1.81)
Some college	33	1.02 (0.61, 1.69)
College degree and above	24	Reference
p-value (test of homogeneity)		0.98
Annual family income ($)
<25,000	14	0.49 (0.21, 1.23)
25,000–50,000	25	0.58 (0.29, 1.54)
50,000–100,000	24	0.64 (0.33, 1.25)
>100,0000	14	Reference
p-value (test of homogeneity)		0.36
Health insurance status
No	4	1.40 (0.52, 3.75)
Yes	73	Reference
Secondhand smoke at home
No	17	Reference
Yes	60	1.29 (0.74, 2.26)
Secondhand smoke at work
No	26	Reference
Yes	51	1.93 (1.21, 3.08)*
Neighborhood Concentrated Disadvantage Index per 10 unit increase	77	1.30 (1.03, 1.63)*
Neighborhood SES Index per 10 unit increase	77	0.82 (0.63, 1.07)

^*p-value < 0.05.

CI: confidence interval; NA: not applicable; NSCLC: non-small cell lung cancer; PM_2.5: particulate matter <2.5 μm in diameter sHR: subdisribution hazard ratio. Test of homogeneity compares the equality of estimates across the categories within a variable.

Table 3.

Multivariable models of neighborhood-level factors associated with lung cancer incidence among never-smokers in the Black Women’s Health Study. Each neighborhood-level factor was adjusted for a priori risk factors⁺ of lung cancer and second-hand smoke at work.

	sHR, adjusted for variables in the multivariable models+ (95 % CI)
Neighborhood Concentrated Disadvantage Index per 10 unit increase #	1.30 (1.04, 1.63)*
Neighborhood SES Index per 10 unit increase#	0.82 (0.63, 1.06)
Effects of neighborhood-level risk factors were adjusted for the following variables in the multivariable model (regardless of the p value)
Age at enrollment	1.06 (1.04, 1.08)***
Body mass index, per unit increase	1.01 (0.97, 1.05)
Health insurance status
No	1.18 (0.41, 3.36)
Yes	1.0
Educational attainment
High school degree or less	1.27 (0.71, 2.28)
Some college	1.06 (0.65, 1.73)
College or Graduate degree	1.0
Annual family income ($)
<25,000	0.74 (0.31, 1.80)
25,000 - <50,000	0.68 (0.33, 1.37)
50,000 - <100,000	0.67 (0.34, 1.34)
>100,000	1.0
Second hand smoke at home
No	1.0
Yes	1.12 (0.63, 1.96)
Second hand smoke at work
No	1.0
Yes	1.96 (1.23, 3.13)**
PM2.5, per unit increase	1.01 (0.95, 1.08)

⁺A priori risk factors included in the multivariable model regardless of statistical significance were age at enrollment, body mass index, health insurance status, education, income, secondhand smoke at home, and PM_2.5.

^*p-value < 0.05.

^**p-value < 0.01.

^***p-value < 0.0001.

CI: confidence interval; PM_2.5: particulate matter < 2.5 μm in diameter; sHR: subdistribution hazard ratio.

In multivariable analysis, in which the effects of a priori known risk factors of lung cancer (age, BMI, PM_2.5, individual educational attainment, individual income attainment, insurance status, and secondhand smoke exposure at home), were adjusted for regardless of their statistical significance, secondhand smoke exposure at work was significant (sHR = 1.96, 95 % CI: 1.23, 3.13). The interaction effects of secondhand smoke at home and at work were not significant (p = 0.41). We then examined the association between lung cancer incidence and the two neighborhood-level variables adjusted for the a priori variables as well as secondhand smoke exposure at work, which was significantly linked to lung cancer incidence on multivariate analysis. We found that lung cancer incidence was not significantly associated with neighborhood SES (sHR = 0.82 per 10 unit increase, 95 % CI: 0.63, 1.06). However, the neighborhood concentrated disadvantage score was significantly associated with increased lung cancer incidence (sHR = 1.30 per 10 unit increase in the deprivation score, 95 % CI: 1.04, 1.63; p = 0.023) (Table 3). To better understand the relationship with lung cancer risk, we categorized the neighborhood concentrated disadvantage variable into quintiles. The sHR for the highest (most deprived) quintile compared to the lowest (least deprived) quintile was elevated, but not statistically significant (sHR = 1.51, 95 % CI: 0.80, 2.88). We did not detect a significant interaction between the neighborhood disadvantage score and secondhand smoke exposure at work.

4. Conclusion

We found that never smoking Black women were more likely to develop lung cancer if they lived in neighborhoods of concentrated disadvantage. This result remained consistent after adjustment for other neighborhood and individual-level characteristics, including secondhand smoke exposure. Our results suggest that concentrated neighborhood disadvantage, independent of the contribution of ambient air pollution exposure and secondhand smoke, may play an important role in the development of lung cancer among never smoking Black women.

This finding is important because it counters the notion that high smoking rates are the sole link between disadvantaged neighborhoods and lung cancer incidence and supports the idea that other non-tobacco-related factors within disadvantaged neighborhoods likely drive lung cancer risk in never smoking Black women. There are several adverse factors within disadvantaged neighborhoods that may contribute to elevated cancer risk. Disadvantaged neighborhoods lack access to supermarkets [27], job opportunities [28], well-funded schools [29], appropriate housing [30], and safe spaces for recreation [31]. Residents of these neighborhoods are also exposed to more toxic waste, pollution, and physiologic stress secondary to neighborhood crime, limited perceived safety, low social cohesion, and insufficient resources to promote health [32]. Furthermore, socioeconomic deprivation is associated with elevated lifetime cancer risk by air toxics [33], and a national US study identified significant positive associations between socioeconomic deprivation and respiratory and cancer hazardous air pollutants [34]. These factors result in chronic stress that leads to “wear and tear” on the body that can dysregulate multiple biological systems and lead to premature illness and mortality [35]. Individuals who live in these disadvantaged neighborhoods have increased markers of tissue inflammation (e.g., C-reactive protein and interleukin–6) [36] and have shorter telomere length, which is linked to accelerated aging [37]. Some models have also hypothesized that psychological stress, acting through increased catecholamine and/or cortisol levels, can increase DNA damage and/or reduce repair mechanisms, resulting in an increased risk of DNA mutations leading to carcinogenesis [38]. These adverse neighborhood factors result in worse outcomes, particularly in NSCLC patients [12,19] and may contribute to the increased risk of lung cancer that Black women who live in these areas experience.

The burden of NSCLC (both smoking and non-smoking-related) is unequally carried by marginalized groups, like Black individuals, who experience disproportionately high incidence [39] of and mortality [19] from the disease, as well as those of low SES who also have higher incidence and mortality from NSCLC than more affluent individuals [18,40]. Prior studies have found that men and women with less than a high school education had 3 and 2 times higher lung cancer incidence rates, respectively than those with a college degree [40]. Those below the poverty level had 52–72 % higher lung cancer incidence rates than their counterparts with incomes at ≥ 600 % of the poverty level [40]. Thus, the link between socioeconomic attainment and lung cancer risk is well-established in cohorts that are predominantly smokers. However, the literature has remained quiet regarding whether neighborhood SES also impacts lung cancer incidence in never smokers. Our study was the first to directly evaluate this question in Black women.

Because there is not a gold standard for measuring neighborhood socioeconomic condition, studies use myriad variables to reflect area-based social factors. Some variables are directly captured from the US census, including factors like median household income and median educational attainment. Others are composites of various SES factors, such as the numerous deprivation and disadvantage indices that exist. While neighborhood concentrated disadvantage was associated with lung cancer incidence in our study, we did not find a significant linear association between the neighborhood SES variable and lung cancer incidence. Our finding that neighborhood disadvantage and neighborhood SES have a disparate impact on lung cancer incidence highlights the fact that neighborhood SES is a catchall term that requires context and specificity. Thus, a comprehensive evaluation of area socioeconomic attainment is required as each distinct SES factor cannot provide a full picture of the total SES of a neighborhood or area. For example, the neighborhood concentrated disadvantage index includes poverty and public assistance metrics, which the neighborhood SES variable did not. Studies have found that concentrated poverty does, in fact, link to poor oncologic outcomes in a way that, for instance, median household income may not [41,42]. This may explain why neighborhood disadvantage and not neighborhood SES is associated with lung cancer risk in Black women who have never smoked.

Because the incidence of lung cancer among never smokers is rising [3,43], it is especially important to understand the drivers of the disease [44]. Known non-smoking-related risk factors for lung cancer include occupational exposures (e.g. painting and welding fumes [45]), environmental exposures (e.g. air pollution [46] and radon [47]), preexisting lung diseases (e.g. chronic obstructive pulmonary disorder, pneumonia, and tuberculosis [48]), and family history [49]. Exposure to secondhand smoke also contributes to elevated NSCLC risk among never smokers [50]. These risk factors are now well-established, but there is a need to better appreciate whether these factors may interact with neighborhood socioeconomic attainment as they influence lung cancer incidence [8,20] and outcomes [18].

In an attempt to better understand which factors within disadvantaged neighborhoods might contribute to risk of lung cancer in Black women never smokers, we examined the association of lung cancer risk with two risk factors known to be commonly experienced by those in disadvantaged areas: air pollution and secondhand smoke exposure.

4.1. Air pollution

Epidemiological studies have suggested a link between PM_2.5 exposure and increased aggressive biology [51], lung cancer risk and mortality in smokers [52] and never smokers [53]. The literature has also revealed that predominantly Black communities have higher exposure to air pollution, despite producing less of it [54]. While these disparate exposures to airborne toxicants have been linked to a 16 % higher cancer risk among individuals living in Black-dominant areas than those in White-dominant areas (p < 0.01) [55], our study did not detect an association between PM_2.5 concentrations and incidence of never-smoking lung cancer. Most studies that have shown a link between PM_2.5 and lung cancer have had a larger number of lung cancer cases. We recommend continuing these investigations in larger cancer cohorts of never smoking Black women in order to better understand the impact of air pollution on lung cancer risk.

4.2. Secondhand smoke

Secondhand smoke has a considerable impact on risk of lung cancer in never smokers [56]. Previous research found a 25 % increased odds of lung cancer in never smokers who were chronically exposed to secondhand smoke at work [56]. Our study also found that secondhand smoke, particularly when experienced at work, was associated with risk of lung cancer in never smoking Black women. The interaction of secondhand smoke exposure at work with neighborhood concentrated disadvantage and neighborhood SES was not significant. This study may have been underpowered to detect an interaction between secondhand smoke exposure and neighborhood SES and neighborhood concentrated disadvantage; thus, we recommend further investigation into this interaction in larger cancer cohorts.

Similar to air pollution and secondhand smoke, diet may be causally linked to lung cancer risk, as consumption of fruits and vegetables has been shown to be protective against lung cancer, including in non-smokers [57]. Across the US, low SES neighborhoods have been found to have limited access to affordable fruits and vegetables, as supermarkets tend to be located in wealthier areas [27]. Therefore, low SES neighborhoods may contribute to increased lung cancer risk in Black women through low access to healthy food. Similarly, physical activity has been suggested as a preventative factor for lung cancer [58]. As with access to healthy food, residents of low SES neighborhoods tend to have less access to infrastructure that supports physical activity (e.g. gyms and sports clubs) [59] and may be less likely to be physically active due to high area crime rates [60]. Although assessing the impact of diet and physical activity on lung cancer risk in never smoking Black women is beyond the scope of this study, we certainly recommend its evaluation in larger cohorts of never-smoking lung cancer patients.

Our study had several strengths. We classified the incidence of lung cancer in a large cohort of Black women from across the US, aiding in the generalizability of our findings to this often-understudied population. We also were able to control for exposure to ambient air pollution and secondhand smoke, both of which have been inconsistently adjusted for in previous related research. This study also had some weaknesses. The characterization of histologic and molecular NSCLC subtypes is important when discussing the outcomes of never smokers with lung cancer; however, this biologic data is not currently available within the BWHS cohort. We also based the neighborhood-level factors (SES, disadvantage, PM_2.5) on the participant’s home address at the time of enrollment, without considering length of time at this residence. This could potentially lead to exposure misclassification among the neighborhood-level variables, although this misclassification is likely to be non-differential with respect to the outcome. Additionally, there was a limited number of cases among never-smoking participants, which hampered our ability to detect subgroup differences, potential interaction effects, or mediators. Furthermore, we lacked access to treatment and survival data whose inclusion certainly would have been beneficial.

The positive correlation between neighborhood concentrated disadvantage and non-smoking NSCLC risk may be signaling the existence of neighborhood-related environmental exposures, social factors and/or behaviors correlated with where one lives that influence lung cancer risk [13]. Our findings may provide clues to further research of the contribution of area-level socioeconomic attainment in lung cancer risk. This analysis demonstrates that neighborhood disadvantage is an independent predictor of non-smoking lung cancer incidence and may contribute to observed racial disparities. Social policy measures aimed at improving the broader social determinants, such as material living conditions and the social and physical environments, are needed to tackle health inequalities in cancer risk.

In conclusion, we found that Black, never-smoking women were more likely to be diagnosed with lung cancer if they lived in neighborhoods of concentrated disadvantage, which was independent of their exposure to ambient air pollution or secondhand smoke. These findings are important because they suggest that neighborhood conditions may impact lung cancer risk through a mechanism that is not dependent on smoking. We believe that these findings may aid in the understanding of why Black, never smoking women are disproportionately affected by lung cancer relative to their NHW counterparts. Future research should aim to uncover the mechanisms underlying this relationship, which will require larger cohorts of lung cancer-affected women.