"I Don't Want an App to Do the Work for Me": A Qualitative Study on the Perception of Online Grocery Shopping From Small Food Retailers

BACKGROUND

Small food retailers often stock energy-dense convenience foods, and they are ubiquitous in low-income urban settings. With the rise in e-commerce, little is known about the acceptability of online grocery shopping from small food retailers.

OBJECTIVE

To explore perceptions of the role of small food retailers (bodegas) in food access and the acceptability of online grocery shopping from bodegas among customers and owners in a diverse New York City urban neighborhood with low incomes.

DESIGN

In-depth interviews were conducted with bodega owners and adult customers between May and July 2022.

PARTICIPANTS/SETTING

Bodega owners who either had (n = 4) or had not (n = 2) implemented a locally designed online grocery system. Customers (n = 25) were recruited through purposive sampling and were eligible if they purchased at bodegas (>once per month), had low income (household income ≤130% of the federal poverty level or Supplemental Nutrition Assistance Program [SNAP] participants), and owned smartphones.

ANALYSES PERFORMED

All interviews were transcribed and analyzed in MAXQDA (Verbi Software, Berlin, Germany), using grounded theory.

RESULTS

To owners and customers, bodegas were seen as good neighbors providing culturally appropriate foods and an informal financial safety net. Their perceptions concerning food cost and availability of healthy foods in bodegas diverged. Although most perceived online grocery from bodegas as a positive community resource, they also believed it was not suited to their own community because of the bodega's proximity to customers' homes and the low digital literacy of some community members. Customers reported social norms of pride in not using online grocery shopping. Owners and customers believed the service would more likely be used if government benefits such as SNAP allowed payment for online orders. Both suggested improved outreach to increase program awareness and uptake.

CONCLUSIONS

Online grocery shopping from small food retailers may be acceptable in urban communities with low income and was perceived as a community resource. However, important barriers need to be addressed, such as social norms related to pride in not using online grocery services, digital literacy, program awareness, and allowing SNAP payment for online orders from bodegas.

Menu Labeling and Calories Purchased in Restaurants in a US National Fast Food Chain

Importance

Menu labeling has been implemented in restaurants in some US jurisdictions as early as 2008, but the extent to which menu labeling is associated with calories purchased is unclear.

Objective

To estimate the association of menu labeling with calories and nutrients purchased and assess geographic variation in results.

Design, Setting, and Participants

A cohort study was conducted with a quasi-experimental design using actual transaction data from Taco Bell restaurants from calendar years 2007 to 2014 US restaurants with menu labeling matched to comparison restaurants using synthetic control methods. Data were analyzed from May to October 2023.

Exposure

Menu labeling policies in 6 US jurisdictions.

Main Outcomes and Measures

The primary outcome was calories per transaction. Secondary outcomes included total and saturated fat, carbohydrates, protein, sugar, fiber, and sodium.

Results

The final sample included 2329 restaurants, with menu labeling in 474 (31 468 restaurant-month observations). Most restaurants (94.3%) were located in California. Difference-in-differences model results indicated that customers purchased 24.7 (95% CI, 23.6-25.7) fewer calories per transaction from restaurants in the menu labeling group in the 3- to 24-month follow-up period vs the comparison group, including 21.9 (95% CI, 20.9-22.9) fewer calories in the 3- to 12-month follow-up period and 25.0 (95% CI, 24.0-26.1) fewer calories in the 13- to 24-month follow-up period. Changes in the nutrient content of transactions were consistent with calorie estimates. Findings in California were similar to overall estimates in magnitude and direction; yet, among restaurants outside of California, no association was observed in the 3- to 24-month period. The outcome of menu labeling also differed by item category and time of day, with a larger decrease in the number of tacos vs other items purchased and a larger decrease in calories purchased during breakfast vs other times of the day in the 3- to 24-month period.

Conclusions and Relevance

In this quasi-experimental cohort study, fewer calories were purchased in restaurants with calorie labels compared with those with no labels, suggesting that consumers are sensitive to calorie information on menu boards, although associations differed by location.

Short relative leg length is associated with overweight and obesity in Mexican immigrant women

Background

Prior research suggests that undernutrition and enteric infections predispose children to stunted growth. Undernutrition and infections have been associated with limited access to healthy diets, lack of sanitation, and access barriers to healthcare – all associated with human rights. Stunting has also been documented to be a major determinant of subsequent obesity and non-communicable diseases. Short leg length relative to stature during adulthood seems to be a good proxy indicator tracking such barriers, and has been reported to be associated with adverse health effects during adulthood. Our objective was to examine the association between relative leg length (as measured by the leg length index, LLI) and measures of adiposity – based on body mass index (BMI) and waist circumference (WC) – in a population of recent Mexican immigrant women to the New York City Area.

Methods

The analysis was based on a cross-sectional survey of 200 Mexican immigrant women aged 18 to 70 years, whose data were collected between April and November 2008; although for purposes of the current study we restricted the sample to those aged 18 to 59 years. The dependent variables were BMI and WC, both transformed into categorical variables. The main independent variable was LLI, and other correlates were controlled for (i.e. age, education, having had children, characteristics of the community of origin, acculturation, chronic conditions, sedentary behaviors, access to fresh fruits and vegetables). Two probit models were estimated: the first one analyzed the effect of LLI on BMI categories and the second one estimated the effect of LLI on WC.

Results

The probit assessing the effect of LLI on overweight/obesity suggested that having a short LLI increased the probability of overweight/obesity by 21 percentage points. Results from the probit model estimating the effect of LLI on WC indicated that having a short LLI increased the probability of having abdominal adiposity by 39 percentage points. Both results were statistically significant at p < 0.05.

Conclusion

The study found an association between having shorter legs relative to one’s height and increased risk of overweight/obesity and abdominal adiposity. Findings support the epidemiological evidence regarding the association between short leg length, early life socioeconomic conditions (i.e. limited access to basic rights), and increased risk of adverse health effects later in life.

Electronic supplementary material

The online version of this article (10.1186/s12939-019-0988-0) contains supplementary material, which is available to authorized users.

Transportation to clinic: findings from a pilot clinic-based survey of low-income suburbanites

Health care policymakers have cited transportation barriers as key obstacles to providing health care to low-income suburbanites, particularly because suburbs have become home to a growing number of recent immigrants who are less likely to own cars than their neighbors. In a suburb of New York City, we conducted a pilot survey of low income, largely immigrant clients in four public clinics, to find out how much transportation difficulties limit their access to primary care. Clients were receptive to the opportunity to participate in the survey (response rate = 94%). Nearly one-quarter reported having transportation problems that had caused them to miss or reschedule a clinic appointment in the past. Difficulties included limited and unreliable local bus service, and a tenuous connection to a car. Our pilot work suggests that this population is willing to participate in a survey on this topic. Further, since even among those attending clinic there was significant evidence of past transportation problems, it suggests that a population based survey would yield information about substantial transportation barriers to health care.

Transportation to clinic: findings from a pilot clinic-based survey of low-income suburbanites

Health care policymakers have cited transportation barriers as key obstacles to providing health care to low-income suburbanites, particularly because suburbs have become home to a growing number of recent immigrants who are less likely to own cars than their neighbors. In a suburb of New York City, we conducted a pilot survey of low income, largely immigrant clients in four public clinics, to find out how much transportation difficulties limit their access to primary care. Clients were receptive to the opportunity to participate in the survey (response rate = 94%). Nearly one-quarter reported having transportation problems that had caused them to miss or reschedule a clinic appointment in the past. Difficulties included limited and unreliable local bus service, and a tenuous connection to a car. Our pilot work suggests that this population is willing to participate in a survey on this topic. Further, since even among those attending clinic there was significant evidence of past transportation problems, it suggests that a population based survey would yield information about substantial transportation barriers to health care.

How Residential Mobility and School Choice Challenge Assumptions of Neighborhood Place—Based Interventions

Purpose.

Explore the importance of residential mobility and use of services outside neighborhoods when interventions targeting low-income families are planned and implemented.

Design.

Analysis of cross-sectional telephone household survey data on childhood mobility and school enrollment in four large distressed cities.

Setting.

Baltimore, Maryland; Detroit, Michigan; Philadelphia, Pennsylvania; and Richmond, Virginia.

Subjects.

Total of 1723 teens aged 10 to 18 years and their parents.

Measures.

Continuous self-report of the number of years parents lived in the neighborhood of residence and city; self-report of whether the child attends school in their neighborhood; and categorical self report of parents' marital status, mother's education, parent race, family income, child's age, and child's sex.

Analysis.

Chi-square and multivariate logistic regression.

Results.

In this sample, 85.2% of teens reported living in the city where they were born. However, only 44.4% of black teens lived in neighborhoods where they were born, compared with 59.2% of white teens. Although 50.3% of black teens attended schools outside of their current neighborhoods, only 31.4% of whites did. Residential mobility was more common among black than white children (odds ratio = 1.82; p < .001), and black teens had 43% lesser odds of attending school in their home communities.

Conclusions.

Mobility among low-income and minority families challenges some assumptions of neighborhood interventions premised on years of exposure to enriched services and changes in the built environment.

Lifting boats without closing gaps: child health outcomes in distressed US cities from 1992-2002

OBJECTIVES

We compared cause-specific mortality and birth rates for children and youths aged younger than 18 years in 100 US cities from 1992 through 2002.

METHODS

We used 5 census indicators to categorize the 100 most populous US cities in 1990 as economically distressed or nondistressed. We used Poisson regression to calculate rate ratios for cause-specific mortality and birth rates, comparing distressed cities to nondistressed cities overall and by race/ethnicity from 1992 through 2002. We also calculated rates of change in these variables within each city over this period.

RESULTS

Despite improvements in health for the study population in all cities, disparities between city groups held steady or widened over the study period. Gaps in outcomes between Whites and Blacks persisted across all cities. Living in a distressed city compounded the disparities in poor outcomes for Black children and youths.

CONCLUSIONS

A strong national economy during the study period may have facilitated improvements in health outcomes for children and youths in US cities, but these benefits did not close gaps between distressed and nondistressed cities.

Disaster in context: the effects of 9/11 on youth distant from the attacks

Although an increasing amount of community mental health research has investigated the deleterious effects of disasters and the targeting and efficacy of treatment in their aftermath, little research has sought to identify preexisting characteristics of the social environment that are predictive of post-disaster distress. A national US telephone survey fielded before and after September 11, 2001, was used to investigate the psychological distress among American adolescents related to the attacks, and to identify environmental and other characteristics that predisposed youth to experience higher or lower levels of post-disaster distress. The study found that widespread characteristics of children's school environments-school disorder and physical threats-were at least as strongly associated with a proxy for psychological distress as exposure to the events of 9/11. Further, children exposed to physical threats at school appeared to be more vulnerable to the psychological effects of disasters than children in safer school environments.

Does the racial composition of the school environment influence children's body mass index?

PURPOSE

This study investigates the degree to which the racial composition of the school environment may influence the body mass index (BMI) of children aged 10 to 18 years. This research may be viewed as extending prior work that has found that the prevalence of risk behaviors among nonwhite adolescents is influenced by exposure to white adolescents.

METHODS

This research used data from the Survey of Adults and Youth, which was conducted as part of the evaluation of the Robert Wood Johnson Foundation's Urban Health Initiative. The study population for this analysis is comprised of parent and child respondents in the 2004 to 2005 survey wave who lived in one of the five program cities: Baltimore, Detroit, Oakland, Philadelphia, and Richmond. We constructed two-level school random effects models and added school and census tract-level variables that describe the racial composition of the residential community and the school attended.

RESULTS

Black and Hispanic adolescent girls who attend schools with a mostly nonwhite student body have higher BMIs than do their white counterparts. However, black girls in predominately white schools do not have higher BMIs than white girls. Further, black and Hispanic girls whose schoolmates are predominately white have significantly lower BMIs than black and Hispanic girls in schools where fewer than half the students are white. These associations are not found among boys, and are net of a broad variety of individual, household, and group level characteristics.

CONCLUSIONS

Our findings suggest that the BMI of minority adolescent girls is influenced by the norms of the social environment.

Race, segregation, and physicians' participation in medicaid

Many studies have explored the extent to which physicians' characteristics and Medicaid program factors influence physicians' decisions to accept Medicaid patients. In this article, we turn to patient race/ethnicity and residential segregation as potential influences. Using the 2000/2001 Community Tracking Study and other sources we show that physicians are significantly less likely to participate in Medicaid in areas where the poor are nonwhite and in areas that are racially segregated. Surprisingly-and contrary to the prevailing Medicaid participation theory--we find no link between poverty segregation and Medicaid participation when controlling for these racial factors. Accordingly, this study contributes to an accumulating body of circumstantial evidence that patient race influences physicians' choices, which in turn may contribute to racial disparities in access to health care.

Inconsistencies in place definition: How different operational place definitions affect estimates of adolescent smoking and drinking risk

We find that estimates of the prevalence of teenage smoking and drinking in "urban," "suburban," and "rural" areas vary with different definitions of these types of geographic units. Given the salience of youth risk behavior to the public debate, we urge researchers to purposefully choose their definitions of geographic areas and to be explicit about those choices.

Teen pregnancy and urban youth: competing truths, complacency, and perceptions of the problem

PURPOSE

To compare and contrast perceptions of community leaders, adults, and youth about the extent of the teen pregnancy problem in five American cities: Baltimore, Detroit, Oakland, Philadelphia, and Richmond.

METHODS

In the five cities from late 1998 through early 2000, semi-structured telephone interviews were conducted with 79 key informants (leaders influential in children's policy issues) to ascertain their perceptions of the most pressing problems facing youth in their cities. Structured, computer-assisted interviews on a range of issues, including teen childbearing and sexual activity, were conducted with 7716 randomly selected adults and 2768 youth aged 10-18 years. The key informant interviews were transcribed and coded; reviewers were paired to validate the coding. The surveys were analyzed using SPSS.

RESULTS

Among the key informants, teen pregnancy was cited as a big problem by only 15%; other issues, such as crime and schools, were seen as more pressing. However, 58% of the adults in the general population thought that teen pregnancy was a big problem. Although almost 3/4 of youth in these cities believed their parents would be upset if they had sex, 87% reported that teen sexual activity before age 18 years was acceptable to their peers, 53% said that teen parenthood was considered acceptable, and 51% had at least one friend who was a teen parent. There were statistically significant differences in the adult and youth responses by race, income, and educational attainment.

CONCLUSIONS

Although few leaders see teen pregnancy as a pressing problem, adults remain deeply concerned, and youth indicate that the problem is prevalent and accepted.

Social networks and homelessness among women heads of household

To examine possible bidirectional relationships between homelessness and deficient social networks, we compared the networks of 251 mothers before, and approximately 5 years after, their families entered shelters with networks of 291 consistently housed poor mothers. At Time 1, more women on the verge of homelessness than housed women reported that they had mothers, grandmothers, friends, and relatives but fewer believed these network members were housing resources. At Time 2, after homeless women were rehoused, these network differences between consistently housed and formerly homeless women had largely disappeared. Contrary to prior research findings, formerly homeless mothers did not report smaller networks, more children or fewer partners. However, formerly homeless women did report fewer positive functions. Because of city policies, homeless mothers were frequently rehoused far from network members.

Which "broken windows" matter? School, neighborhood, and family characteristics associated with youths' feelings of unsafety

Young people's fears of victimization and feelings of unsafety constitute a serious and pervasive public health problem and appear to be associated with different factors than actual victimization. Our analysis of a population-based telephone survey of youths aged 10-18 years in five economically distressed cities and their suburbs reveals that a substantial minority of youths feel unsafe on any given day, and that an even greater number feel unsafe in school. While some traditional predictors of victimization (such as low socioeconomic status) were associated with feeling unsafe, perceived school disorder was the major factor associated with such feelings. Disorderliness may thus be the school's version of "broken windows," which serve to signal to students a lack of consistent adult concern and oversight that can leave them feeling unsafe. We suggest that fixing the broken windows of school disorderliness may have a significant, positive impact on adolescents' feelings of safety.

Low response rate schools in surveys of adolescent risk taking behaviours: possible biases, possible solutions

STUDY OBJECTIVE

To examine the potential biases introduced when students in low response rate schools are dropped from classroom based surveys of adolescent risk taking behaviour.

DESIGN

Self administered confidential surveys were conducted in classrooms, with follow up visits to each school to survey students absent during the initial survey administration. Data on students in schools that achieved a 70% response rate are compared with data on students in schools that did not achieve this level of response.

SETTING

New York City, United States.

PARTICIPANTS

1854 10th graders in 13 public (state supported) high schools.

MAIN RESULTS

Students in schools with low response rates resulting from high rates of absenteeism have different demographic characteristics and engage in more risk behaviours than students in schools with low absenteeism and high response rates. Excluding schools with low rates of response can have an effect on estimates of risk behaviour, even after data are weighted for individual absences. The potential for bias is greatest when, in sampling schools, the proportion of schools with low response rates is large, and when such schools represent a large share of the students in the area under study.

CONCLUSIONS

Excluding schools with poor response rates from survey samples using a classroom based approach does not improve, and may, under some circumstances, underestimate risky behaviour among adolescent populations.

Mother-child separations among homeless and housed families receiving public assistance in New York City

We examined the incidence, characteristics, and predictors of separations of children from mothers in 543 poor families receiving public assistance, 251 of whom had experienced homelessness during the previous 5 years. Forty-four percent of the homeless mothers and 8% of housed mothers were separated from one or more children. A total of 249 children were separated from 110 homeless families and 34 children from 23 housed families. Children were placed with relatives and in foster care but were rarely returned to their mothers. Maternal drug dependence, domestic violence, and institutionalization predicted separations, but homelessness was the most important predictor, equivalent in size to 1.9 other risk factors. We infer that policies regarding child welfare and substance abuse treatment should be changed to reduce unnecessary placements. Studies of homeless children who remain with families may be biased if separated children are excluded.

Classroom-based surveys of adolescent risk-taking behaviors: reducing the bias of absenteeism

OBJECTIVES

This investigation examined the effectiveness of intensive efforts to include frequently absent students in order to reduce bias in classroom-based studies.

METHODS

Grade 10 students in 13 New York City high schools (n = 2049) completed self-administered confidential surveys in 4 different phases: a 1-day classroom capture, a 1-day follow-up, and 2 separate 1-week follow-ups. Financial incentives were offered, along with opportunities for out-of-classroom participation.

RESULTS

Findings showed that frequently absent students engaged in more risk behaviors than those who were rarely absent. Intensive efforts to locate and survey chronically absent students did not, however, significantly alter estimates of risk behavior. Weighting the data for individual absences marginally improved the estimates.

CONCLUSIONS

This study showed that intensive efforts to capture absent students in classroom-based investigations are not warranted by the small improvements produced in regard to risk behavior estimates.

The relationship of residential instability to medical care utilization among poor mothers in New York City

OBJECTIVES

This study examines the relationship between residential instability, including mobility and previous homelessness, and the use of medical care among previously sheltered and never-sheltered mothers in New York City. The study represents one of the first efforts to follow up on families after they are no longer homeless.

METHODS

Mothers from 543 welfare families in New York City were interviewed, once in 1988 (Time 1) and again beginning in 1992 (Time 2). The sample included 251 families who first entered shelters after their 1988 interview, and 292 families who spent no time in shelters before or after that point. Mothers were asked about the source and volume of medical care used in the year before follow-up.

RESULTS

Bivariate and multivariate analyses showed that previously sheltered mothers had a greater reliance on emergency departments (EDs) and weaker ties to private physicians or health maintenance organizations (HMOs) than did mothers who never used shelters. Mobility before the Time 1 interview was associated with greater reliance on EDs and absence of a usual source of care. More recent mobility was not associated with a usual source of care. Current residential stability reduced the likelihood of using an emergency department or having no regular source of care. None of the measures of residential instability were related to the volume of outpatient care used by mothers.

CONCLUSIONS

A history of residential instability, particularly previous shelter use, strongly predicts where poor mothers currently seek health care. Further research is needed to determine whether these patterns of health care use existed before mothers entered shelters. The study provides evidence that upon leaving shelters, mothers are not being well integrated into primary care services.

Predictors of homelessness among families in New York City: from shelter request to housing stability

OBJECTIVES

This study examined predictors of entry into shelter and subsequent housing stability for a cohort of families receiving public assistance in New York City.

METHODS

Interviews were conducted with 266 families as they requested shelter and with a comparison sample of 298 families selected at random from the welfare caseload. Respondents were reinterviewed 5 years later. Families with prior history of shelter use were excluded from the follow-up study.

RESULTS

Demographic characteristics and housing conditions were the most important risk factors for shelter entry; enduring poverty and disruptive social experiences also contributed. Five years later, four fifths of sheltered families had their own apartment. Receipt of subsidized housing was the primary predictor of housing stability among formerly homeless families (odds ratio [OR] = 20.6, 95% confidence interval [CI] = 9.9, 42.9).

CONCLUSIONS

Housing subsidies are critical to ending homelessness among families.

Access to hospitals with high-technology cardiac services: how is race important?

OBJECTIVES

Relatively few hospitals in the United States offer high-technology cardiac services (cardiac catheterization, bypass surgery, or angioplasty). This study examined the association between race and admission to a hospital offering those services.

METHODS

Records of 11,410 patients admitted with acute myocardial infarction to hospitals in New York State in 1986 were analyzed.

RESULTS

Approximately one third of both White and Black patients presented to hospitals offering high-technology cardiac services. However, in a multivariate model adjusting for home-to-hospital distance, the White-to-Black odds ratio for likelihood of presentation to such a hospital was 1.68 (95% confidence interval = 1.42, 1.98). This discrepancy between the observed and "distance-adjusted" probabilities reflected three phenomena: (1) patients presented to nearby hospitals; (2) Blacks were more likely to live near high-technology hospitals; and (3) there were racial differences in travel patterns. For example, when the nearest hospitals did not include a high-technology hospital, Whites were more likely than Blacks to travel beyond those nearest hospitals to a high-technology hospital.

CONCLUSIONS

Whites and Blacks present equally to hospitals offering high-technology cardiac services at the time of acute myocardial infarction. However, there are important underlying racial differences in geographic proximity and tendencies to travel to those hospitals.

Impact of employment-based health insurance on home attendants

This study examines 253 newly hired home attendants to measure the degree to which employment-based health insurance can affect health status and utilization of health care services among a working poor population that has little experience with health insurance and may face other significant barriers to care. Physician contacts increased after benefits were received; attendants who had no coverage during the prior year experienced the greatest average increase. More attendants also reported using emergency rooms. Neither hospitalizations nor health status were affected. These findings indicate that insurance benefits may substantially improve access to care for many working poor persons, regardless of other barriers they may face.

Health status and health care utilization among New York City home attendants: an illustration of the needs of working poor, immigrant women

In this paper, the health needs and health care utilization patterns of home attendants and their families have been studied as an illustration of those likely to be found among working poor, immigrant women and their children. Despite tremendous growth in the number of immigrants, studies to date provide only limited information regarding the specific health needs and patterns of health care utilization among such women and their children. As part of a longitudinal study on the impact of insurance on health status and health care utilization, 387 female, immigrant home attendants were interviewed. Data were also gathered on 355 of their minor children. These women and children were found to be less likely than other Americans to make use of basic health services, despite the fact that they are more likely to indicate fair or poor health status. This is true even in comparison to poor or uninsured Americans. Immigrant attendants in fair or poor health report an average annual visit rate of 4.1 ambulatory care visits for themselves and 2.2 for their children, as compared to 8.4 for poor adults and 4.4 for poor children in national samples. These findings illustrate the likelihood that poor, immigrant women make limited use of American medical care, and face barriers to health care that appear even greater than those faced by the uninsured and the poor.

Predictors of shelter use among low-income families: psychiatric history, substance abuse, and victimization

For poor housed and homeless families in New York City, NY, we examined the degree to which psychiatric and substance-abuse problems and victimization placed the families at elevated risk of requiring emergency housing, and we documented the prevalence of such problems. These problems were infrequently reported by both groups. However, past mental hospitalization, treatment in a detoxification center, childhood sexual abuse, and adult physical abuse were associated with increased risk of homelessness.

Social relationships and vulnerability to becoming homeless among poor families

This study compares social relationships of 677 mothers in families requesting shelter with those of 495 mothers in housed families, randomly selected from the public assistance caseload in New York City. As hypothesized, women seeking shelter had experienced higher levels of a variety of childhood and adult events indicative of disruptions in social relationships. Contrary to our hypothesis, they were more likely than were housed mothers to have had recent contact with parents, other relatives, and friends, although they felt less able to draw on these resources for help with their current housing needs. More than three fourths of families seeking shelter had already stayed with members of their social network in the past year. The data suggest that they had used up potential sources of support before turning to public shelter.

Pregnancy and childbirth: risk factors for homelessness?

A comparison of 704 homeless public assistance families in New York City with 524 families on public assistance who had housing found that pregnancy and recent births were highly correlated with becoming homeless. Thirty-five percent of homeless women were pregnant at the time of the interview, and 26 percent had given birth in the past year, compared with six percent and 11 percent, respectively, of women in the housed sample. In addition, having a baby before age 18 (as had 37 percent of the homeless women and 24 percent of the housed women) was significantly related to homelessness but family size was not.