Article Text

Download PDFPDF

Original research
Relationship between food insecurity and smoking status among women living with and at risk for HIV in the USA: a cohort study
  1. Lila A Sheira1,
  2. Edward A Frongillo2,
  3. Judith Hahn1,
  4. Kartika Palar1,
  5. Elise D Riley1,
  6. Tracey E Wilson3,
  7. Adebola Adedimeji4,
  8. Daniel Merenstein5,
  9. Mardge Cohen6,
  10. Eryka L Wentz7,
  11. Adaora A Adimora8,9,
  12. Ighovwerha Ofotokun10,
  13. Lisa Metsch11,
  14. Janet M Turan12,
  15. Phyllis C Tien13,14,
  16. Sheri D Weiser1,15
  1. 1 Division of HIV, Infectious Diseases, and Global Medicine, University of California, San Francisco, California, USA
  2. 2 Department of Health Promotion, Education, and Behavior, University of South Carolina System, Columbia, South Carolina, USA
  3. 3 Community Health Sciences, School of Public Health, State University of New York Downstate Medical Center, Brooklyn, New York, USA
  4. 4 Department of Epidemiology and Population Health, Albert Einstein College of Medicine, Bronx, New York, USA
  5. 5 Family Medicine, Georgetown University Medical Center, Washington, District of Columbia, USA
  6. 6 Medicine, John H Stroger Hospital of Cook County, Chicago, Illinois, USA
  7. 7 School of Public Health, Johns Hopkins University, Baltimore, Maryland, USA
  8. 8 School of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA
  9. 9 UNC Gillings School of Global Public Health, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA
  10. 10 School of Medicine, Emory University, Atlanta, Georgia, USA
  11. 11 Sociomedical Sciences, Columbia University Mailman School of Public Health, New York City, New York, USA
  12. 12 Department of Health Care Organization and Policy, School of Public Health, University of Alabama at Birmingham, Birmingham, Alabama, USA
  13. 13 Department of Medicine, University of California, San Francisco, California, USA
  14. 14 Department of Veteran Affairs Medical Center, San Francisco, California, USA
  15. 15 Center for AIDS Prevention Studies, University of California, San Francisco, California, USA
  1. Correspondence to Lila A Sheira; lila.sheira{at}ucsf.edu

Abstract

Objectives People living with HIV (PLHIV) in the USA, particularly women, have a higher prevalence of food insecurity than the general population. Cigarette smoking among PLHIV is common (42%), and PLHIV are 6–13 times more likely to die from lung cancer than AIDS-related causes. This study sought to investigate the associations between food security status and smoking status and severity among a cohort of predominantly low-income women of colour living with and without HIV in the USA.

Design Women enrolled in an ongoing longitudinal cohort study from 2013 to 2015.

Setting Nine participating sites across the USA.

Participants 2553 participants enrolled in the Food Insecurity Sub-Study of the Women’s Interagency HIV Study, a multisite cohort study of US women living with HIV and demographically similar HIV-seronegative women.

Outcomes Current cigarette smoking status and intensity were self-reported. We used cross-sectional and longitudinal logistic and Tobit regressions to assess associations of food security status and changes in food security status with smoking status and intensity.

Results The median age was 48. Most respondents were African-American/black (72%) and living with HIV (71%). Over half had annual incomes ≤US$12 000 (52%). Food insecurity (44%) and cigarette smoking (42%) were prevalent. In analyses adjusting for common sociodemographic characteristics, all categories of food insecurity were associated with greater odds of current smoking compared with food-secure women. Changes in food insecurity were also associated with increased odds of smoking. Any food insecurity was associated with higher smoking intensity.

Conclusions Food insecurity over time was associated with smoking in this cohort of predominantly low-income women of colour living with or at risk of HIV. Integrating alleviation of food insecurity into smoking cessation programmes may be an effective method to reduce the smoking prevalence and disproportionate lung cancer mortality rate particularly among PLHIV.

  • HIV & AIDS
  • epidemiology
  • public health

Data availability statement

Data may be obtained from a third party and are not publicly available. You may request access to the MACS or WIHS Public Use Data Set (MWCCS public use data set will be available in 2021 at https://airtable.com/shrVDP51W5J2qcNeT), or you may submit a concept sheet in order to have access to the full MWCCS data set, which includes data not available in the PDS such as neurocognitive data, abuse data, healthy aging, etc. Having an approved concept sheet means you will receive assistance in understanding the MWCCS study and the complexity of the MWCCS data; linkage to a MWCCS liaison who can offer expertise in your area of study; support regarding study design, methodology and statistical analysis; and access to all summary files.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Strengths and limitations of this study

  • Although much of the previous literature about smoking among people living with HIV has been conducted among white men, this study was conducted among a large study sample of women and predominantly women of colour.

  • The analyses allowed the study to estimate multiple relational structures between food insecurity and smoking, including smoking status and smoking intensity.

  • The study lacked information on participant use of other tobacco products, notably e-cigarettes, precluding a more comprehensive definition of tobacco use as an outcome.

  • Smoking status varied little over time, which may have limited estimation of the relationship between changes in smoking status and in food security.

Background

Cigarette smoking is among the leading causes of excess mortality worldwide1 2 and the leading risk factor for preventable death in high-income countries such as the USA.3 The prevalence of smoking in the US general adult population was 14% in 2017.4 In 2019, the prevalence of smoking among men was 15.3% compared with 12.7% of women.5 The prevalence of cigarette smoking is also substantially higher among low-income individuals at and below the federal poverty level, among whom 41% of men and 33% of women smoke6 and higher proportions use all other forms of tobacco, including e-cigarettes.7 Women in general have a harder time quitting smoking,8 both ‘cold-turkey’9 and through other methods such as the patch,10 compared with men, leading to longer lifetime smoking duration and nicotine exposure.

The prevalence of smoking among people living with HIV (PLHIV) is 42%, similar to that among the low-income general population and more than double the general population estimates.11 Beyond the well-documented sequelae of cigarette smoking in the general population,3 PLHIV who are smokers additionally experience a higher risk of pneumonia,12 emphysema13 and other illnesses of the lung,14 compared with their HIV-seronegative counterparts who smoke. Further, PLHIV who smoke also have higher odds of a detectable viral load15 16 and faster progression to AIDS compared with non-smoking PLHIV.17 Nearly one-quarter of deaths among PLHIV can be attributed to current smoking,18 and PLHIV who smoke are also 6–13 times more likely to die from lung cancer than from AIDS-related causes.19

Despite sex differences in cigarette smoking prevalence in the general US population,4 cigarette smoking prevalence does not differ by sex among PLHIV. Smoking among women living with HIV (WLWH) may have additional reproductive and maternal health consequences, including a higher risk of pre-eclampsia than both HIV seronegative- and non-smoking WLWH counterparts,20 increased fetal morbidity compared with HIV seronegative smokers,21 and earlier onset of natural menopause compared with non-smoking WLWH.22 On the whole, these smoking-specific health consequences may have additive or multiplicative interactions with general HIV-related conditions,23 thereby reducing the immunological support of antiretroviral therapy (ART).24

Food insecurity, defined as ‘the uncertain or limited availability of nutritionally adequate or safe food or the inability to procure food in socially acceptable ways,’25 26 is prevalent in low-income households and has been linked to smoking in the general population.27 28 Food insecurity affects 12% of American households and 16% of households with a child under 18.29 Food insecurity is more prevalent among households led by women and women living alone, ethnic and racial minorities, and households with children (compared with the general US prevalence).29

Estimates of food insecurity among PLHIV range from 20% to 50%,30 31 with higher prevalence among WLWH compared with their male counterparts.32 33 Food insecurity among PLHIV is associated with decreased mental and physical health status,34 suboptimal adherence to ART,35 use of illicit substances,36 and increased HIV-related morbidity and HIV mortality.37 Important limitations of the studies investigating smoking among PLHIV include that they were conducted among predominantly male populations living with HIV,38–40 lacked granularity in the smoking variable38 40 or did not examine smoking severity.39 40 Given that tobacco use is the most important preventable cause of excess mortality worldwide,1 2 with increased health consequences to PLHIV and WLWH specifically, expanding our understanding of the role of food insecurity as a potentially modifiable factor among WLWH is vital to reducing these health disparities. Therefore, we conducted an analysis of data from the Women’s Interagency HIV Study (WIHS) to understand the associations between food insecurity and smoking over time. We hypothesised that: (1) greater severity of current food insecurity would be associated with higher odds of being a current smoker, (2) change in food security status over time would be associated with change in smoking status and (3) greater severity of current food insecurity would be associated with higher intensity of smoking.

Methods

Data

Data for this study originated from the WIHS, a longitudinal cohort study of WLWH and demographically similar HIV-seronegative controls in the USA that began in 1993 and enrolled women over four recruitment waves in 1994–1995, 2001–2002, 2011–2012 and 2013–2015. The recruitment wave of 2001–2002 prioritised the recruit of younger participants, and in those with HIV, participants who were ART-naive, while the 2011–2012 wave was to replace participants who had died in the interim. From 2013–2015, four new sites in the Southern USA were added to enrol women representative of the distribution of the HIV epidemic in the USA.41–43 Recruitment methods are described in detail elsewhere;42 in brief, women were recruited from HIV care clinics, churches, HIV community organisations and social service agencies. For the duration of this study, the WIHS collected sociobehavioural, biological and clinical data from all participants during semiannual visits using interviewer-administrated standardised instruments, physical examinations and standard phlebotomy. The physical examination includes standard anthropometry, weight and a gynaecological examination. Immunological and virological biomarker measurements included current CD4 count and HIV RNA viral load.

Beginning in 2013, the Food Insecurity Substudy added data on comprehensive measures of food security, dietary intake, household savings and food support among all WIHS women at nine sites: Bronx, New York; Brooklyn, New York; Washington, DC; Chicago, Illinois; San Francisco, California; Chapel Hill, North Carolina; Miami, Florida; Birmingham, AL/Jackson, Mississippi and Atlanta, Georgia. During the substudy period, there were 12 464 person-visits in total in the WIHS among 2613 unique women. Of these person-visits, 608 were abbreviated visits during which women only contributed laboratory specimens. During the substudy period, 317 women were deactivated or disenrolled from the WIHS, mainly due to death (110, 34.4%) or unenrolment due to a site losing funding (Brooklyn only; 114, 36.6%). The rest were due to participant’s decision (10%), site’s decision (3.4%) or travel reasons (16.7%). The analyses in this paper used data on 11 692 person-visits from 2553 unique women in total (1803 living with and 750 without HIV collected from April 2013 to March 2016 at every semiannual visit). Of these women, 1689 had been recruited prior to the newest recruitment wave and could therefore contribute up to six visits in total during the substudy period; the median number of visits attended was 6 (IQR: 5–6). The remaining 864 women were recruited or transferred to the Southern sites during the latest recruitment wave occurring contemporaneously with our substudy. These women could contribute between two and five visits; the median number of visits attended was 4 (IQR: 3–4). Participants provided written informed consent and were compensated for participation.

Measures

The exposure was food insecurity, measured using the US Department of Agriculture’s Household Food Security Module (HFSSM).44 The HFSSM has been validated in high-resource settings among both vulnerable populations45 46 and those living with HIV, and is the reference measure of food security in the USA.47 The HFSSM includes 18 items about insufficient food quantity, low diet quality, uncertainty about food and food affordability.45 The HFSSM uses recall periods of 12 months and 30 days; the WIHS module was worded to ask about food security over the previous six months or since the last WIHS visit. The HFSSM scoring algorithm categorises individuals as having high, marginal, low or very low food security.47 The HFSSM was available for 98.6% of respondents who were offered the food insecurity substudy (ie, did not have abbreviated visits), thus, no missing data methods were used for the exposure variable. Cronbach’s alpha for the HFSSM in this sample was 0.91, indicating high internal consistency.

The primary outcomes were (1) current cigarette smoking status (smoker vs non-smoker) and (2) smoking intensity (number of cigarettes/day). Both outcomes were assessed by self-report at each visit during the interviewer-administered interview. Participants were asked ‘Since your study visit on [previous study visit date], have you smoked cigarettes?’ Those who responded yes were further prompted to recall how many cigarettes on average they smoked per day. Given the non-normal distribution of cigarettes smoked per day, the values above zero for this variable were transformed by the natural logarithm. Those who reported zero cigarettes per day were not transformed and were retained as being left-censored. Complete smoking data were available for 99.5% of respondents of the food insecurity substudy module, thus, no missing data methods were used for the outcome. Per WIHS protocol, all participants who reported current smoking were subsequently referred to a smoking cessation programme.

Covariates: Covariates were selected a priori based on review of the literature regarding food security and smoking. Covariates were HIV status (seropositive or seronegative (reference group)), age at visit (per year), race/ethnicity (non-Hispanic white (reference group), non-Hispanic African-American/black, Hispanic and non-Hispanic other), annual household income as collected by WIHS (≤US$12 000 (reference group), US$12 001–US$24 000, US$24 001–US$36 000, US$36 001–US$75 000 and ≥US$75 001), employment status (employed (reference group) or unemployed), marital status (partnered (reference group), divorced/widowed/separated, never married or other), educational attainment (less than high school education (reference group) or greater than a high school education or equivalent), and if they had child dependents under age 18 in the household (none (reference group) or yes). Response options were in reference to the previous 6 months. Covariate data were missing from 373 unique person-visits (3.2% of person-visits), thus, no missing data methods were used.

Given that food insecurity is associated with several mental health outcomes including depression,34 anxiety and stress,48 and illicit substance use,36 and there is likely a bidirectional association between smoking and mental health,49 50 mental health may be a mediator rather than a confounder on the path from food insecurity to smoking outcomes. Therefore, we did not adjust for mental health variables, as these could potentially be on the causal path from food insecurity to smoking outcomes.

Analysis

Summary statistics were obtained for food insecurity, the smoking outcomes and all covariates at study baseline (ie, the first visit during the Food Insecurity Sub-Study). We used a logistic regression model with one cross-sectional sample at the first measure of food insecurity (food insecurity substudy baseline) to assess the association between food security status and the odds of being a current smoker (hypothesis 1). We also modelled an interaction term between food security and HIV status to assess whether food security in the presence of HIV-seropositivity was associated with differential odds of smoking compared with food secure, HIV-seronegative women. Next, we used a longitudinal logistic regression model with fixed effects for individuals to assess the association between changes in food security status and the odds of becoming a current smoker compared with becoming a non-smoker (hypothesis 2). This model ruled out potential confounding by unobserved or observed time-invariant characteristics. The fixed-effects model removed all individuals who did not have a change in smoking status over the study visits, allowing us to examine just those who had a change in smoking status and leaving a sample of 344 women (comprising 1700 person-visits). Given that this model removes person-to-person variability, it allows for the interpretation of change as effects are generated only by those who experience any change. The coefficients from this model are interpreted as adjusted ORs (AORs). Finally, we used longitudinal Tobit regression to model the association between food security and smoking intensity (natural logarithm of cigarettes/day) (hypothesis 3). Tobit models allow for censoring and were thus implemented given that a large proportion of the values for cigarettes per day were left-censored as over half of women in the sample were non-smokers. Given that cigarettes per day was transformed to the logarithmic scale, the results from the Tobit model are presented as exponentiated coefficients and interpreted as a relative difference (ie, multiplicative factor) compared with the reference category. All analyses were conducted using Stata V.15 (StataCorp).

Patient and public involvement

There was no patient or public involvement in the development of the research questions or in the analyses.

Results

At the Food Insecurity Sub-Study baseline, 42% of women reported being current smokers and 44% reported any category of food insecurity (table 1). Among current smokers, the median number of cigarettes smoked per day was 5 (IQR: 3–10). The median age of women was 48 years (IQR: 40–54) and women were predominantly HIV-seropositive (71%) and of African-American/black race (72%), followed by Hispanic women (15%). Nearly two-thirds of the women reported an education equivalent to or greater than high school (67.4%) and were unemployed at baseline (65%). Over half had annual household incomes less than US$12 000.

Table 1

Sociodemographic characteristics of the sample at first visit in the food insecurity substudy, women’s Interagency HIV Study (n=2553 unique women)

In the unadjusted model which included 2228 women, marginal, low and very low food security was associated with 1.64 (95% CI: 1.26 to 2.14), 1.90 (95% CI: 1.46 to 2.48) and 2.44 (95% CI: 1.84 to 3.23) times greater odds of being a current smoker (table 2). In the adjusted model among 2133 women, current marginal, low and very low food security was associated with 1.52 (95% CI: 1.14 to 2.04), 1.70 (95% CI 1.28 to 2.27) and 1.91 (95% CI 1.42 to 2.59) times greater odds of being a current smoker compared with those with high food security (all p<0.01; table 3). HIV-seropositivity was associated with lower odds of being a current smoker (AOR: 0.66; 95% CI 0.53 to 0.81, p<0.001). The association between food security status and current smoking was not modified by HIV status (not shown).

Table 2

Unadjusted associations between food security (FS) and smoking outcomes

Table 3

Adjusted associations between food security (FS) and smoking outcomes

In the longitudinal individual fixed-effects model, becoming of marginal, low and very low food security status was associated with 1.49 (95% CI: 1.04 to 2.13), 1.77 (95% CI: 1.17 to 2.67) and 1.62 (95% CI: 1.01 to 2.58) times greater odds of becoming a current smoker, respectively, compared with becoming a non-smoker (table 2). In adjusted models, becoming marginal, low and very low food security were associated with 1.56 (95% CI: 1.08 to 2.25), 1.88 (95% CI: 1.23 to 2.87) and 1.66 (95% CI: 1.02 to 2.81) times greater odds of becoming a current smoker, respectively, compared with becoming a non-smoker (table 3). In the adjusted model, none of the other time-varying variables were significantly associated with becoming a current smoker.

In the final model for smoking intensity, all study participants were included with non-smokers censored at zero. The unadjusted relative differences of intensity of smoking (cigarettes/day) were 1.16 (95% CI: 1.07 to 1.26), 1.22 (95% CI: 1.11 to 1.33) and 1.17 (95% CI: 1.06 to 1.30) times higher for women with marginal, low and very low food security status, respectively, compared with those with high food security (table 2). The adjusted relative differences of intensity of smoking were 1.17 (95% CI: 1.07 to 1.27), 1.21 (95% CI: 1.10 to 1.32), and 1.16 (95% CI: 1.04 to 1.29) times higher for women with marginal, low and very low food security status, respectively, compared with those with high food security (table 3).

Discussion

In this longitudinal study of WLWH and demographically similar women without HIV, food insecurity was associated with greater odds of being a current cigarette smoker, with higher odds of smoking as food insecurity severity worsened. Furthermore, any change in food security status was associated with a change in smoking status, and food insecurity was positively associated with smoking intensity. HIV status did not modify these associations, but HIV-seropositivity was associated with lower odds of being a current smoker and lower intensity of smoking compared with HIV-seronegative participants.

Our findings are consistent with literature on the association between food insecurity and smoking status among the general US population,28 and among women living with and at risk for HIV who experience housing instability.51 Our findings are not consistent with the literature among WLWH, who have a higher prevalence of smoking than their HIV-seronegative peers.52 While the prevalence of smoking in the cohort is relatively high, it is lower among WLWH (39%) than demographically similar HIV-seronegative women (49%, p<0.001). Due to the nature of healthcare in the USA, WLWH may have access to more consistent healthcare compared with their demographically similar HIV-seronegative counterparts, allowing for more opportunities for smoking cessation referrals, which may in part explain the lower prevalence of smoking among WLWH in this study compared with the controls. In the WIHS, all women who reported smoking received information at each visit referring them to external smoking cessation programmes. Economically disadvantaged people and ethnic minorities typically have less access to smoking cessation treatment,53 54 and women in general have less success in smoking cessation in the long-term than men.8 Given that this study was predominantly composed of these populations who are understudied among PLHIV in the USA, our study fills an important gap in the smoking and PLHIV literature.

Being food insecure was associated with 1.16–1.21 times higher (ie, 16%–21% higher) smoking intensity compared with being food secure, even after controlling for income. Food insecurity may drive higher smoking intensity through two mechanisms. First, food insecurity is a profound stressor that leads to poor mental health48 which in turn is associated with cigarette smoking.49 50 Second, tobacco acts as an appetite suppressant via the hypothalamus,55 56 and food insecure individuals may smoke to cope with hunger.57 58 Given that measures of food insecurity and smoking intensity were obtained at the same study visit, and time between each visit (six months) was too long to assess directionality by staggering food insecurity and smoking intensity, we cannot determine which was driving the other. The consistency in linking food insecurity with smoking in low-income women living with and at risk for HIV here and in prior studies51 indicates the importance of addressing this issue.

Smoking intensity, however, did not have a dose response relationship with the severity of food insecurity. Two plausible mechanisms may explain this lack of a dose response relationship. First, being food insecure may coincide with volatility in financial resources, which in turn may affect one’s ability to afford highly taxed cigarettes and subsequent smoking intensity but not affect smoking status. That is, people of low socioeconomic status are more likely to respond to increased cigarette taxation by reducing intensity of smoking but not by eliminating consumption of cigarettes compared with those of higher socioeconomic status.59 60 Second, if food-insecure individuals have access to programmes that alleviate but not eliminate food insecurity (allowing them to go from a previously very low food security status to low or moderate) via food banks, pantries or food stamps, then resources previously dedicated to food may be available for non-essential goods including cigarettes. Nevertheless, the relationship between food security and smoking intensity underscores the importance of integrating food security alleviation programmes with smoking cessation programming.

This study was unable to assess food insecurity nor changes in food security status as a predictor of smoking cessation success. Becoming food insecure (compared with maintaining a food secure status) is associated with lower odds of smoking cessation among smokers and with smoking initiation in non-smokers.28 Yet, we do not know if a change from being food insecure to food secure may prompt a reduction or cessation in cigarette smoking, presumably as the appetite-suppressing effects of cigarettes are no longer needed. These valuable data would shed light on the effectiveness of the integration of food insecurity alleviation in smoking cessation programming and could be used for policy development and scaled programming.

The study included a large, geographically diverse sample of WLWH and without HIV with similar demographic characteristics. A rich set of information was collected for each woman and models adjusted for appropriate control variables. Nevertheless, several potentially important individual characteristics that were not measured were smoking products and smoking behaviours, family history of smoking, previous cessation attempts, social support and participation in food insecurity alleviation programmes. Given that there were few differences in the longitudinal model (compared with the cross-sectional model) when we were able to control for individual variability, as well as the fixed effects model (which removed measured and unmeasured individual characteristics that do not change over time), the role of these potential confounders may be minimal. Collecting data on smoking intensity is notoriously difficult given variabilities in recalling smoking intensity. The survey questions assessed smoking intensity since the last visit six months earlier; long recall periods present opportunities for recall bias in self-reporting of health data. Likewise, food insecurity was assessed during that same recall period. Food security status may fluctuate on a monthly or weekly basis in households experiencing food or financial scarcity; a recall period of six months may not allow us to evaluate these periodic shifts. Further, we were unable to assess other types of common tobacco use, notably vaping and e-cigarettes, whose prevalence and public health impact are increasing greatly. The development of validated questions that can evaluate tobacco exposure from traditional (ie, cigarette, chewing tobacco, cigars) as well as new delivery mechanisms (ie, e-cigarettes, vaping) will facilitate a more accurate measurement of tobacco exposure. This potential misclassification in the exposure could bias results towards the null, whereby women who are food insecure and exclusive vape users were categorised as non-smokers, yet the paths from food insecurity and tobacco use and addiction potentially remain the same. Furthermore, although the prevalence of self-reported cigarette smoking was not low in the sample, there was little change in status over the assessed study visits, which may have limited estimation of the relationship between changes in smoking status and food security. Lastly, despite recent recruitment efforts to match the demographic profile of WLWH in the USA,43 WIHS women may not match the profile of those with incident HIV in the USA, notably regarding age, which has implications for the generalisability of this study.

In conclusion, food insecurity was associated with both being a smoker and smoking intensity in this sample of WLWH and without HIV. Smoking has a high attributable risk for preventable deaths in the USA and globally, and food insecurity may exacerbate this risk. Even on its own, food insecurity significantly increases persistent morbidity across multiple populations.61 62 Food insecurity is modifiable and alleviating it should be considered in conjunction with smoking cessation programmes. Alleviating food insecurity through augmentation of resources while simultaneously reducing smoking, however, is challenging. For example, alleviating food insecurity in overweight women through resource augmentation has tended to exacerbate excess weight,63 but a recent randomised control trial has demonstrated that a programme that was carefully designed and implemented based on in-depth understanding of women’s lives in Costa Rica64 was effective in simultaneously reducing food insecurity and excess body weight.65 Therefore, experimental or observational studies of programmes to alleviate food insecurity should collect data on recent tobacco initiators or quitters to provide valuable information on the contextual milieu that is related to smoking initiation, duration, intensity and cessation. With such information, researchers may be able to better identify those at risk for tobacco use initiation as well as design and enhance policies and programmes to simultaneously reduce food insecurity and promote smoking cessation.

Data availability statement

Data may be obtained from a third party and are not publicly available. You may request access to the MACS or WIHS Public Use Data Set (MWCCS public use data set will be available in 2021 at https://airtable.com/shrVDP51W5J2qcNeT), or you may submit a concept sheet in order to have access to the full MWCCS data set, which includes data not available in the PDS such as neurocognitive data, abuse data, healthy aging, etc. Having an approved concept sheet means you will receive assistance in understanding the MWCCS study and the complexity of the MWCCS data; linkage to a MWCCS liaison who can offer expertise in your area of study; support regarding study design, methodology and statistical analysis; and access to all summary files.

Ethics statements

Patient consent for publication

Ethics approval

Institutional Review Boards at each of the following universities or institutions reviewed and approved the study procedures (per the MWCCS overall study director, approval numbers are not available): University of Mississippi Medical Center, University of North Carolina at Chapel Hill, University of Alabama at Birmingham, University of Miami, Emory University, Brooklyn WIHS site, SUNY Downstate Medical Center, Kings County Medical Center, Bronx WIHS site, Montefiore Medical Center, Beth Israel Medical Center, Mount Sinai School of Medicine, Chicago WIHS site, Cook County Health & Hospitals System, Northwestern University, Rush University Medical Center, University of Illinois at Chicago, San Francisco WIHS site, University of California, San Francisco, Alameda Health System, Sutter Health, Santa Clara Valley Medical Center, San Mateo Medical Center, DC WIHS site, Georgetown University, Montgomery County Department of Health and Human Services, Inova, Howard University, Whitman-Walker Clinic

Acknowledgments

Publication made possible in part by support from the UCSF Open Access Publishing Fund.

References

Footnotes

  • Contributors Planning: SDW, PCT and EAF. Research and study design: LAS, SDW, EAF and PCT; data analysis: LAS and EAF; data interpretation: LAS, SDW, EAF, JH and EDR; data collection: LAS, KP, TEW, AA, DM, MC, ELW, AAA, IO, ELW, LM, JMT, PCT and SDW. All authors reviewed and provided feedback on intellectual content and approved the final version of the manuscript.

  • Funding This study was funded by a Women’s Interagency HIV Study (WIHS) substudy grant from the National Institute of Mental Health, R01MH095683, a mentoring grant K24AI134326 (Weiser), and a scientist development grant K01DK107335 (Palar). Data in this manuscript were collected by the Womens Interagency HIV Study (WIHS), now the MACS/WIHS Combined Cohort Study (MWCCS). MWCCS (principal investigators): Atlanta CRS (Ighovwerha Ofotokun, Anandi Sheth and Gina Wingood), U01-HL146241; Baltimore CRS (Todd Brown and Joseph Margolick), U01-HL146201; Bronx CRS (Kathryn Anastos and Anjali Sharma), U01- HL146204; Brooklyn CRS (Deborah Gustafson and Tracey Wilson), U01-HL146202; Data Analysis and Coordination Centre (Gypsyamber D’Souza, Stephen Gange and Elizabeth Golub), U01-HL146193; Chicago-Cook County CRS (Mardge Cohen and Audrey French), U01-HL146245; Chicago- Northwestern CRS (Steven Wolinsky), U01-HL146240; Northern California CRS (Bradley Aouizerat, Jennifer Price and Phyllis Tien), U01-HL146242; Los Angeles CRS (Roger Detels and Matthew Mimiaga), U01-HL146333; Metropolitan Washington CRS (Seble Kassaye and Daniel Merenstein), U01- HL146205; Miami CRS (Maria Alcaide, Margaret Fischl and Deborah Jones), U01-HL146203; Pittsburgh CRS (Jeremy Martinson and Charles Rinaldo), U01-HL146208; UAB-MS CRS (Mirjam- Colette Kempf, Jodie Dionne-Odom, and Deborah Konkle-Parker), U01-HL146192; UNC CRS (Adaora Adimora), U01-HL146194. The MWCCS is funded primarily by the National Heart, Lung, and Blood Institute (NHLBI), with additional cofunding from the Eunice Kennedy Shriver National Institute Of Child Health and Human Development (NICHD), National Institute On Ageing (NIA), National Institute Of Dental and Craniofacial Research (NIDCR), National Institute Of Allergy And Infectious Diseases (NIAID), National Institute Of Neurological Disorders And Stroke (NINDS), National Institute Of Mental Health (NIMH), National Institute On Drug Abuse (NIDA), National Institute Of Nursing Research (NINR), National Cancer Institute (NCI), National Institute on Alcohol Abuse and Alcoholism (NIAAA), National Institute on Deafness and Other Communication Disorders (NIDCD), National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institute on Minority Health and Health Disparities (NIMHD) and in coordination and alignment with the research priorities of the National Institutes of Health, Office of AIDS Research (OAR). MWCCS data collection is also supported by UL1- TR000004 (UCSF CTSA), P30-AI-050409 (Atlanta CFAR), P30-AI-050410 (UNC CFAR) and P30-AI- 027767 (UAB CFAR).

  • Disclaimer The contents of this publication are solely the responsibility of the authors and do not represent the official views of the National Institutes of Health (NIH).

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.