Abstract
Background:
NYC taxi/for-hire vehicle (FHV) drivers have occupational and demographic characteristics associated with food insecurity (low income, comorbidities, minority race/ethnicity).
Objective:
To analyze food insecurity rates in a sample of NYC drivers and to identify associated factors.
Methods:
At health fairs, we recruited a cross-sectional sample of licensed taxi/FHV drivers willing to receive study text messages. Most lacked a primary care provider. Food insecurity prevalence and associations with health and economic indicators were analyzed.
Results:
Of 503 participants who completed a 2-item food security screener, 39.2% were food insecure. Significantly fewer food insecure than food secure drivers reported a doctor visit within the past year (48% vs 25%; P<.001). Food insecure drivers had greater weekly traffic ticket expenditure ($34 vs $24; P=.02) and were more likely to report insufficient household income (61% vs 39%; P<.001) and history of depression (14% vs 7%; P=.02), to have elevated (>200) measured total cholesterol (50% vs 37%; P=.02), and to have Perceived Stress Scale scores indicating greater stress than food secure drivers (14 vs 11, P=.002). In a binary logistic regression analysis, drivers who reported that their total household income was enough to meet their basic needs had significantly lower odds of being food insecure (0.695 odds ratio; P=.016).
Conclusions:
Food insecurity was high in this group of taxi/FHV drivers. Food insecurity interventions are needed and could be occupationally based, with worksite screening and resource navigation. Policies should address improving wages and healthcare access.
Keywords: Workplace, Stress, Psychological, Food Insecurity, Primary Health Care, Ethnicity
INTRODUCTION
Food insecurity, a lack of access to adequate food for an active and healthy lifestyle, affected 10.5% of United States (U.S.) households in 2019 [1]. It is lower among non-Hispanic White (7.9%) than non-Hispanic Black (19.1%) and Hispanic (15.6%) households and is comparatively high in low-income households overall (27.6%) [1]. Food insecurity is a social determinant of health, one of a range of nonmedical and socioeconomic and demographic factors, such as income, structural racism, and education, that powerfully influence health [2]. It is associated with obesity, depression, diabetes, hypertension, hyperlipidemia, and poor sleep [3]. People with food insecurity may delay needed medical treatment and obtaining prescription medications to purchase food [3].
New York City’s (NYC’s) 185,000 taxi/for-hire vehicle (FHV) drivers are comprised of a large, predominantly immigrant and minority population [4]. Despite often working shifts of up to 12 hours a day, 6/7 days a week [5], their incomes are low [6], and NYC food costs are among the nation’s highest [7]. NYC taxi/FHV drivers have higher rates of overweight and obesity than New Yorkers in general (77% vs 56%) [8, 9]. Their occupation is sedentary with numerous stressors [10, 11]. Both NYC and Chicago studies have reported that drivers often have energy-dense, nutrient-poor diets lacking in fruit and vegetables [6, 8, 12, 13]. Taxi drivers have a matrix of intersecting socioeconomic/demographic characteristics associated with ill health and food insecurity, including low income, frequent experiences of racism, limited healthcare access, and limited knowledge of and access to resources, such as government benefits and food banks.
The U.S. taxi/FHV driving population more than doubled from 305,000 in 2016 to 703,000 in 2019 [14–16]. There was well-documented overpricing of NYC taxi medallions, and app-based FHV driving has cut into traditional drivers’ market shares [6, 17], while creating a group of app-based drivers with unstable work and limited earnings after expenses [18–20]. Hence, taxi/FHV drivers often have marginal, unstable incomes and earnings potential. Because of the food insecurity risk factors that taxi/FHV drivers face, and their elevated risk for food insecurity-associated chronic health conditions, we investigated the prevalence of food insecurity and associated conditions among NYC taxi/FHV drivers, where food insecurity rates for the general population were 13.8% for all races/ethnicities in 2019 [21]. The results could drive targeted program and policy development to address local and national Taxi/FHV driver food insecurity [22, 23].
METHODS
Design
This was a cross-sectional study among NYC taxi/FHV drivers who attended (from April 2018-July 2019) community-based health fairs at NYC taxi garages, airport taxi/FHV holding lots, faith- and community-based organizations, and other community driver-frequented sites (described in detail in a prior publication [24]) that offered enrollment into one of three research studies. Trained bilingual staff administered the survey in drivers’ preferred languages, Bengali, English, French, or Spanish. Blood pressure, height, weight, waist circumference, and total cholesterol were measured. A physician/nurse provided relevant one-on-one health education. Drivers were then referred to needs based services, including primary care, urgent care, and health insurance navigation.
Eligible participants were men who had been licensed taxi/FHV drivers ≥3 months, worked ≥30 hours/week, spoke English/Bengali/French/Spanish, and were willing to receive text messages. Additional inclusion criteria were required for the individual study into which the drivers could enroll: study 1, no usual primary care provider (PCP) and no annual physical within the past year; study 2, high blood pressure but no history of cardiovascular disease (CVD); study 3, willingness to use a pedometer and also no history of CVD. Of 1478 drivers screened, 508 were eligible to participate in one of the three studies (n=315 from study 1, n=67 from study 2, and n=126 from study 3). 503 completed the food security screener and were included in the present analysis. Women were excluded (NYC taxi/FHV drivers are 96% male) [4]. Participants were given $50 as time compensation.
These studies were reviewed and approved by Memorial Sloan Kettering Cancer Center Institutional Review Board/Privacy Board and participants provided written informed consent.
Measures
Sociodemographic Information
We used the 2 item food security screening tool that was adapted from the U.S. Household Food Security Survey Module (USDA) 18-item food security scale [25, 26], which has >97% sensitivity and >70% specificity when used in clinical settings with time demands that compel screening efficiency [3, 25, 27–29]. It asks whether the following 2 statements were “often true,” “sometimes true,” or “never true” in the past 12 months: “We worried whether (my/our) food would run out before (I/we) got money to buy more,” and “The food that (I/we) bought just didn’t last and (I/we) didn’t have money to get more.” An affirmative response to either question was considered indicative of food
Demographic questions elicited drivers’ gender, education level, marital status, household size, race, ethnicity, and level of English language proficiency, assessed with the question “How well do you speak English?” and the response options “Very well,” “Well,” “Not well,” and “Not at all” [30, 31]. Participants who did not speak English “Very well” were considered limited English proficient. Drivers were also asked whether they drove a taxi or FHV.
Drivers reported their work shift (day/night) and days/week and hours/day worked. Financial questions included average weekly taxi-related expenses, household monthly and yearly income, and average weekly/monthly driving income. Yearly household income was calculated using self-reported total household income, adjusted by household size. Drivers also reported their weekly/monthly expenditures, including home/apartment rent, mortgage, money sent home, food, and traffic tickets. Finally, drivers were asked “Is your total household income enough to meet the basic needs of you and/or your family?,” with the response options “No, it is often difficult to cover all household expenses,” “Yes, it is just enough to cover all expenses,” and “Yes, it is enough to cover expenses and to save the extra.”
Healthcare profile and medical history
Drivers were asked to identify their healthcare provider and health insurance status and whether they had a doctor’s office visit in the past year, using questions from the Medical Expenditure Panel Survey and the Centers for Disease Control and Prevention’s 2013 Behavioral Risk Factor Surveillance System Questionnaire [32–34]. The validated Self-administered Comorbidity Questionnaire elicited self-reported history of conditions, including diabetes, hypertension, cancer, anxiety, and depression [35–37].
Perceived Stress Scale
We used the Cohen’s Perceived Stress Scale (PSS-10), a 10-item scale that surveys individuals’ stress-related thoughts/feelings over the previous month [38], and utilizes a 5-point Likert-like scale, with responses ranging from “Almost never” to “Very often,” Scores are obtained by reverse-coding answers to the positive items and then summing the items’ scores (0-13=low stress, 14-26=moderate stress, 27-40=high stress).
Biometric measurements
Height, weight, and waist circumference were measured. Body mass index was calculated using Centers for Disease Control guidelines [39]. A Cholestech LDX Analyzer was used to analyze blood capillary samples for total cholesterol.
Data Analyses
Descriptive statistics were used to summarize sociodemographic data. Chi-square tests with continuity adjustment were used to examine the association between food insecurity and categorical covariates. Point-biserial correlation was computed to assess the relationship between food insecurity and continuous covariates, such as PSS-10 scores. Covariates with a statistically reliable univariate association were entered into a multivariate logistic regression to examine the extent to which each variable was associated with food insecurity. Only complete cases were included in the analyses. Statistical analyses were conducted using SPSS version 24.
RESULTS
Sixty-three percent of participants identified as Black, 29% Asian, 7% Caucasian, and 1% other. Thirty-six percent identified their ethnicity as Hispanic or Latino. Birth regions included Sub-Saharan Africa (36%), South Asia (24%), and Latin America (17%). They had an average of 19 years’ of U.S. residence (range,1-47).
Six percent had completed 11th grade or less, 33% were high school graduates, and 62% had received higher education. On average, participants worked 56 hours/week, lived in 3-person households, and had $20,086 adjusted yearly household income per household member. Just over half (52%) of the drivers had health insurance, and 40% had a PCP. A total of 39.2% of study taxi/FHV drivers were found to be food insecure (Table 1).
Table 1.
| Driver Characteristics | Total Drivers | Food Secure | Food Insecure | |||||
|---|---|---|---|---|---|---|---|---|
| No. | (%) | No. | (%) | No. | (%) | P Valuec | ||
| Overall (N=508) | 508 | 100 | 306 | (60.8) | 197 | (39.2) | ||
|
| ||||||||
| English proficiency (n=502) | Very well | 417 | (83) | 259 | (85) | 158 | (81) | .35 |
| Well/Not well/Not at all | 85 | (17) | 47 | (15) | 38 | (19) | ||
|
| ||||||||
| Education level (n=500) | 11th grade or less | 30 | (6) | 18 | (6) | 12 | (6) | .92 |
| High school graduate | 163 | (33) | 98 | (32) | 65 | (33) | ||
| Some College | 108 | (22) | 64 | (21) | 44 | (23) | ||
| College graduate | 161 | (32) | 98 | (32) | 63 | (32) | ||
| Post-college | 38 | (8) | 27 | (9) | 11 | (6) | ||
|
| ||||||||
| Race (n=398) | Caucasian | 30 | (7) | 21 | (8) | 9 | (6) | .67 |
| Black | 256 | (63) | 158 | (62) | 98 | (64) | ||
| Asian | 109 | (29) | 74 | (29) | 45 | (29) | ||
| Other | 3 | (1) | 1 | (1) | 2 | (1) | ||
|
| ||||||||
| Region of birth (n=497) | East Asia | 6 | (2) | 3 | (1) | 3 | (1) | .92 |
| Latin America | 85 | (17) | 51 | (17) | 34 | (17) | ||
| North Africa and Middle East | 35 | (7) | 19 | (6) | 16 | (8) | ||
| South Asia | 117 | (24) | 75 | (25) | 42 | (22) | ||
| Sub-Saharan Africa | 178 | (36) | 108 | (36) | 70 | (36) | ||
| Other | 76 | (15) | 46 | (15) | 30 | (15) | ||
|
| ||||||||
| Ethnicity (n=254) | Hispanic/Latino | 178 | (36) | 36 | (12) | 26 | (13) | .66 |
| Non-Hispanic/Latino | 76 | (15) | 264 | (88) | 169 | (87) | ||
|
| ||||||||
| Marital status (n=497) | Married/partnered | 290 | (58) | 184 | (61) | 106 | (55) | .53 |
| Not married/partnered | 207 | (42) | 121 | (39) | 86 | (45) | ||
|
| ||||||||
| Primary care provider (n=503) | Yes | 201 | (40) | 132 | (43) | 69 | (35) | .07 |
| No | 302 | (60) | 174 | (57) | 128 | (65) | ||
|
| ||||||||
| Doctor visit in last year (n=503) | Yes | 181 | (36) | 147 | (48) | 49 | (25) | .008 |
| No | 322 | (64) | 159 | (52) | 148 | (75) | ||
|
| ||||||||
| Health insurance (n=481) | Yes | 249 | (52) | 159 | (53) | 90 | (49) | .45 |
| No | 232 | (48) | 140 | (47) | 92 | (51) | ||
|
| ||||||||
| Taxi type (n=488) | Taxi | 179 | (37) | 106 | (34) | 73 | (38) | .63 |
| FHV | 309 | (63) | 189 | (66) | 120 | (62) | ||
|
| ||||||||
| Shift (n=496) | Day | 212 | (43) | 137 | (45) | 75 | (39) | .12 |
| Night | 113 | (23) | 60 | (20) | 53 | (27) | ||
| Varies | 171 | (34) | 106 | (35) | 65 | (34) | ||
| Mean | (SD) | Mean | (SD) | Mean | (SD) | |||
|
| ||||||||
| Time in the U.S., years (n=499) | 19 | (11.898) | 19 | (11.110) | 19 | (12.395) | .64 | |
|
| ||||||||
| Time worked per week, hours (n=501) | 56 | (12.752) | 55 | (12.798) | 56 | (12.687) | .37 | |
|
| ||||||||
| Household size, no. (n=497) | 3 | (1.755) | 3 | (1.727) | 3 | (1.804) | .91 | |
|
| ||||||||
| Adjusted yearly household income, $ | 20,086 | (16,804) | 20,085 | (16,952) | 20,088 | (16,620) | .99 | |
Abbreviations: FHV, for-hire vehicle; No., number; U.S., United States
Missing data were excluded from the analyses
508 participants eligible, 503 completed food security screener
Our predetermined statistical significance level was P≤.05.
The only significant difference in food security status by demographic and health access characteristics was that food secure drivers were more likely to have had a doctor’s visit in the past year than food insecure drivers (48% vs 25%; P<.001). There were other trends: a larger percentage of food secure than insecure drivers spoke English very well, were Caucasian, were married or partnered, drove FHVs, and had PCPs and health insurance.
Food insecure drivers had greater weekly expenditures on traffic tickets ($34 vs $24/week; P=.02) and were more likely to report that their household income could not meet their needs (61% vs 39%; P<.001) than food secure drivers (Table 2).
Table 2.
Financial characteristics and food security statusa
| Financial Condition | Food Secure | Food Insecure | P Valuec | |||
|---|---|---|---|---|---|---|
| Mean | (SD)b | Mean | (SD)b | |||
| Overall, No. (%) (n=503) |
306 | (60.8) | 197 | (39.2) | ||
|
| ||||||
| Weekly income from driving, $ (n=501) |
785 | (400.611) | 726 | (452.203) | .15 | |
|
| ||||||
| Weekly spending on rent, $ (n=498) |
269 | (159.085) | 283 | (159.696) | .41 | |
|
| ||||||
| Weekly spending on mortgage, $ (n=325) |
484 | (206.719) | 488 | (174.221) | .41 | |
|
| ||||||
| Average money sent back home weekly, $ (n=367) |
99 | (95.802) | 107 | (107.692) | .46 | |
|
| ||||||
| Weekly spending on food, $ (n=421) |
179 | (129.876) | 184 | (127.285) | .69 | |
|
| ||||||
| Weekly spending on traffic tickets, $ (n=494) |
24 | (28.284) | 34 | (48.673) | .02 | |
|
| ||||||
| Average weekly spending on taxi-related expenses, $ (n=493) |
606 | (429.806) | 608 | (380.508) | .97 | |
|
| ||||||
| Whether total household income can meet needs, no. % (n=477) |
Yes, no. (%) (n=250) |
178 | (61) | 72 | (39) | .001 |
|
| ||||||
| No, no. (%) (n=227) |
115 | (39) | 112 | (61) | ||
Abbreviations: No., number; SD, standard deviation.
Missing data were excluded from the analysis
Unless otherwise indicated
Our predetermined statistical significance level was P≤.05
History of diabetes was reported among 13% of food secure versus 8% of food insecure drivers (P=.009) and history of depression among 7% of food secure versus 14% of food insecure drivers (P=.02) (Table 3). Total cholesterol was elevated (>200) among 37% of food secure and 50% of food insecure drivers (P=.02). Mean PSS scores were 11 (low) for food secure and 14 (moderate) for food insecure drivers (P=.002). Body mass index and self-reported history of CVD/heart disease/stroke, hypertension, cholesterol problems, cancer, and anxiety did not significantly differ between the two groups.
Table 3.
Health conditions and food security statusa
| Health Condition | Food Secure | Food Insecure | P Valueb | |||
|---|---|---|---|---|---|---|
| No. | (%) | No. | (%) | |||
| Overall (n=503) | 306 | (60.8) | 197 | (39.2) | ||
|
| ||||||
| Objective measures, recorded at health fair | ||||||
|
| ||||||
| Body mass index (n=483) |
Normal | 77 | (26) | 49 | (26) | .94 |
| Overweight | 128 | (43) | 78 | (42) | ||
| Obese | 91 | (31) | 60 | (32) | ||
|
| ||||||
| Total cholesterol level (n=369) | Normal | 139 | (63) | 74 | (50) | .02 |
| Elevatedc | 83 | (37) | 73 | (50) | ||
|
| ||||||
| Self-reported comorbidity questionnaire | ||||||
|
| ||||||
| CVD/Heart Disease/stroke (n=476) |
No | 286 | (99) | 175 | (99) | .41 |
| Yes | 4 | (1) | 1 | (1) | ||
|
| ||||||
| Diabetes (n=457) |
No | 245 | (87) | 118 | (92) | .009 |
| Yes | 38 | (13) | 10 | (8) | ||
|
| ||||||
| Hypertension (n=453) |
No | 192 | (31) | 132 | (76) | .11 |
| Yes | 87 | (69) | 42 | (24) | ||
|
| ||||||
| Problem with cholesterol (n=439) |
No | 209 | (77) | 136 | (81) | .34 |
| Yes | 62 | (23) | 32 | (19) | ||
|
| ||||||
| Cancer (n=472) |
No | 290 | (99) | 180 | (99) | .73 |
| Yes | 1 | (1) | 1 | (1) | ||
|
| ||||||
| Depression (n=473) |
No | 271 | (93) | 154 | (86) | .02 |
| Yes | 22 | (7) | 26 | (14) | ||
|
| ||||||
| Anxiety (n=470) |
No | 269 | (92) | 158 | (89) | .35 |
| Yes | 24 | (8) | 19 | (11) | ||
|
| ||||||
|
PSS Score, mean (SD) n=(496) |
11 | (0.422) | 14 | (0.561) | .002 | |
Abbreviations: CVD, cardiovascular disease; No., number; PSS, Perceived Stress Scale.
Missing data were excluded from the analyses
Our predetermined statistical significance level was P≤.05.
Measured total cholesterol >200
All significant factors in the univariate analyses were analyzed in a binary logistic regression to examine the relative association of each health and socioeconomic condition and food insecurity. Drivers who reported that their total household income was enough to meet their basic needs had significantly lower odds of being food insecure than drivers who reported that their total household income was not enough to meet their basic needs (0.695 odds ratio; P=.016).
DISCUSSION
The food insecurity rate in our pooled sample of NYC taxi/FHV drivers, who had enrolled in studies aimed at improving health outcomes, was a staggering 39.2%. This was approximately three times the 2019 NYC, NY State, and national averages of 13.8%, 10.5%, and 10.8%, respectively [1, 21]. With food insecurity rates that are potentially so much higher than in the general population, population-based research is urgently needed to determine the rates across all taxi/FHV drivers and to address the causes and implications of food insecurity in the growing and vulnerable national taxi/FHV driver population.
Taxi/FHV driver food insecurity was significantly associated with the likelihood that total household income could not meet needs. Industry disruption has resulted in downward pressure on taxi drivers’ wages [17], and for-hire vehicle drivers contend with low earnings after vehicle-related expenses (such as fuel, vehicle payments, and maintenance) [40]. At the time of this study, many NYC drivers had large debts for taxi Medallions that had drastically decreased in value since they were purchased, and some Medallion drivers have reported spending most of their income on debt repayments [17]. Food insecure drivers in our study had significantly higher traffic ticket expenditure than food secure drivers. NYC drivers have mentioned the difficulty of serving customers without incurring traffic tickets, such as when picking up or dropping off in restricted parking areas [41]. San Diego drivers reported fearing fines, and drivers in San Diego and Los Angeles reported feeling powerless to affect industry change to alleviate such financial burdens [42, 43].
Consistent with reports in other populations, food insecurity was associated with history of depression and higher average Perceived Stress Scale score, indicating higher stress among food insecure than food secure drivers [44–46]. Taxi/FHV drivers contend with a range of occupationally related stressors, including discrimination, long working hours, and financial strain [5, 6, 17–20]. Studies have found that having food insecurity is positively associated with risk of depression and stress [47]. Food insecurity could be a further stressor for drivers. Research indicates that individuals faced with one stressor will intensify their negative reaction to additional stressors [48]. Interventions to address the possibly compounding effects of drivers’ stressors are indicated, as are linkages with mental health services [49].
Significantly fewer food insecure than food secure drivers had a doctor visit within the past year, a finding that is consistent with the literature associating food insecurity with delayed medical care [3]. Significantly higher proportions of food insecure than food secure drivers in our study had elevated measured total cholesterol, yet fewer food insecure than food secure drivers reported a history of high cholesterol, indicating the possibility of underdiagnosis of high cholesterol among food insecure drivers. Facilitating PCP access is indicated among food insecure taxi/FHV drivers who, based on this evidence, may require CVD risk reduction intervention.
Self-reported history of diabetes was significantly associated with food security status, with proportionately fewer food insecure than food secure drivers reporting history of diabetes. At the population level, food insecurity has been associated with increased likelihood of having diabetes [50], and our findings were unexpected. The rate of self-reported diabetes among food insecure taxi/FHV drivers (8%) was below the population rate (13%) [51], and underdiagnosis may have been a factor in our results, given the lack of PCP access among food insecure drivers in our sample. This may further emphasize the need to facilitate PCP access in this population.
Studies should examine taxi/FHV driver food security nationally. The U.S. government’s Supplemental Nutritional Assistance Program (SNAP) provides cash assistance to purchase food for households that meet income requirements [52], but 30% of food insecure households nationally do not meet income requirements, including net income ≤100% of the poverty level [7, 53]. Furthermore, the SNAP eligibility criteria do not account for individual, household, and environmental factors that have been found to impact need, such as the time needed to prepare nutritious food and the local costs of food and rent [54]. Occupation-associated groups (FHV companies, driver guilds, unions, taxi garages, base stations) and community and faith-based organizations with large taxi driver memberships could screen drivers for SNAP eligibility to ensure that drivers who qualify are accessing SNAP benefits.
Community resources, including food banks and soup kitchens, can also help to alleviate food insecurity, although their resources have been put under pressure by 2013 SNAP cuts [55, 56], which left NYC residents with $1 billion less food through the program [55], and COVID-19 pandemic-related demand [57]. Additionally, food pantries are often inaccessible to taxi/FHV drivers because of their limited locations and opening hours. Help with non-food expenditures, such as medical and legal expenses, could help food insecure households [7].
Health fairs at taxi bases, airport holding lots, and community centers frequented by taxi/FHV drivers have proven to be effective for reaching this population [24, 58]. Gany et al. (2015) used this approach to screen taxi drivers for health risks, link them with health insurance, and navigate them into resources and healthcare [58]. Results were encouraging, with 65% of the drivers who required urgent follow-up seeking medical care at least once.
LIMITATIONS
Because of the eligibility criteria, drivers lacking access to healthcare, having high blood pressure, and/or interested in using pedometers were oversampled, which could affect the generalizability of these results. Nevertheless, our dataset allowed us to identify issues in a taxi/FHV driving population with numerous vulnerabilities that intersect with food insecurity, including working in an area with high food costs and contending with declining income and profound industry disruption [6, 7]. This population could not be isolated and analyzed using population-level surveillance data, such as the Current Population Survey, which conflates taxi drivers with other types of drivers and does not collect all of the data that our survey collected, such as health-related and detailed birth country and language data [15].
This study used the 2-item USDA food security screener. Although it has been shown to have high levels of sensitivity, its specificity is lower than the 10-item screener, and it cannot measure the severity of food insecurity [27, 59]. Hence, the level of food insecurity might have been overstated. However, the 2-item screening tool was more feasible to use than the 10-item tool under the time constraints of taxi/FHV drivers, who are often in a rush to get back to work. Future studies could utilize the 2-item screener to detect food insecurity, and then, for greater specificity, the 10-item screener for food insecure participants.
A further limitation was the use of total cholesterol as a measure of CVD risk. Future studies should consider measuring participants’ HgB A1C (glycosylated hemoglobin level) to identify potential type 2 diabetes and prediabetes and measuring lipid ratios as better indicators of CVD risk [60]. Another limitation was relying on self-reported comorbidities in a population with low healthcare access, among whom awareness of health conditions may have been low, potentially resulting in underreporting. Furthermore, depression and anxiety should be determined upon examination by a physician or (at least) by a validated self-administered questionnaire (e.g., Beck Depression Inventory) [61].
PUBLIC HEALTH IMPLICATIONS
Food insecurity was highly prevalent in this sample of NYC taxi/FHV drivers, highlighting a dire need for further investigation and intervention in this vulnerable population. Our findings illustrate how the social determinants of health in a low-wage, high-stress, largely immigrant and minority occupational group that receives little advocacy intersect with health indicators. By addressing the social determinants of taxi/FHV driver health, including food insecurity, low wages, high expenses, and healthcare access, identified in this study and others, taxi drivers’ physical and mental health could potentially be improved [5, 6, 8, 12, 42, 43].
The high prevalence of food insecurity among our sample of taxi/FHV drivers should compel policymakers and regulators to develop strategies to ease drivers’ financial strain, including through the facilitation of higher earnings and increased access to public benefits. An ongoing driver screening program, for food insecurity and other health needs/risk factors, could be implemented at driver-frequented sites to facilitate uptake [24, 58, 62]. Clinicians and advocates could provide education on healthy eating on the road and stretching the food dollar and could navigate drivers and their family members to targeted programs and resources, such as SNAP and food banks. Community-based legal clinics could assist drivers with contesting traffic tickets. These measures could lessen drivers’ occupational and financial stress, allowing them to bring healthier food to the table [22, 23].
Our results could inform research and interventions in other metropolitan areas with concentrations of taxi/FHV drivers, including San Francisco, Chicago, Los Angeles, and Miami [12, 16, 42, 43]. Our research could also be relevant to other low-wage occupational groups with immigrant and minority overrepresentation, such as the 1.5 million U.S. delivery truck drivers [22]. Additionally, service workers constituted 2.3 million of the nation’s working poor in 2017, earning less than poverty-level wages [23], making them highly vulnerable to food insecurity [7].
Since our data were gathered, NYC taxi/FHV drivers have been hit hard by the COVID-19 crisis, which reduced business by as much as 80% in May 2020 and likely exacerbated food insecurity [63]. Considering the enormous impact of COVID-19 on the taxi-driving industry [64], the need for immediate and sustained action to target driver food insecurity is more urgent than ever.
Table 4.
Binary logistic regression on associations with food insecurity (n=503)a
| Variables | Coefficient | Odds Ratio | 95% CI | P valueb | |
|---|---|---|---|---|---|
| Having diabetes | Yes | 1.149 | 3.154 | (0.772-12.885) | 0.11 |
| No | ref | ref | ref | ||
|
| |||||
| Having a PCP | Yes | −0.561 | 0.57 | (0.861-3.567) | 0.122 |
| No | ref | ref | ref | ||
|
| |||||
| PSS score | 0.043 | 1.044 | (0.998-1.093) | 0.063 | |
|
| |||||
| Weekly spending on traffic tickets | 0.007 | 1.007 | (0.997-1.017) | 0.162 | |
|
| |||||
| Total cholesterol level | Normal | −0.558 | 0.572 | (0.295-1.112) | 0.099 |
| Risk level | ref | ref | ref | ||
|
| |||||
| Having depression | Yes | −0.363 | 0.479 | (0.272-1.780) | 0.449 |
| No | ref | ref | ref | ||
|
| |||||
| Total income enough to meet the basic needs? | Yes | −0.807 | 0.695 | (0.232-0.859) | 0.016 |
| No | ref | ref | ref | ||
Abbreviations: PCP, primary care provider; PSS, perceived stress scale; ref, reference.
“Ref” denotes the reference group, which is the comparison group in the analysis.
Our predetermined statistical significance level was P≤.05.
Funding:
This research was supported by funding from the National Institutes of Health (R01NR015265, R24MD008058, U01MD010648, U54CA137788, P30CA008748) and the Aetna Foundation.
Role of the funder/sponsor:
The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Editorial support:
The authors would like to thank Sonya J. Smyk, Memorial Sloan Kettering Cancer Center, for writing and editorial support. She was not compensated beyond her regular salary.
Footnotes
Ethical approval: This study was reviewed and approved by the Memorial Sloan Kettering Cancer Center’s Institutional Review Board (IRB approval nos. 15-163, 16-1659, 16-038).
Informed consent: Participants provided their written informed consent.
Research team: This study would not have been possible without the contributions of the research team. The authors would like to thank the clinicians, community outreach assistants, investigators, and researchers who supported this study, including Memorial Sloan Kettering’s Abraham Aragones, M.D., Nicole Roberts-Eversley, M.P.H., Julia Ramirez, M.A., Anuradha Hashemi, M.P.H., Caroline Sturm Reganato, R.N., Jessica Llamozas, R.N., Redwane Gatarny, Katherine Leopold, Muksha Jingree, Chanel Martinez, Khaliq Sanda, and Brooke Shawcross, the South Asian Council for Social Services’ Rehan Mehmood and Sudha Acharya, and the Taxi Network Community Advisory Board.
Conflict of interest: The authors report no relevant conflicts of interest.
REFERENCES
- [1].United States Department of Agriculture Economic Research Service Food Security Status of US Households in 2019: USDA; 2020. [updated September 9, 2020]. Available from: https://www.ers.usda.gov/topics/food-nutrition-assistance/food-security-in-the-us/key-statistics-graphics/#foodsecure. [Google Scholar]
- [2].Braveman P, Gottlieb L. The social determinants of health: it’s time to consider the causes of the causes. Public Health Rep. 2014;129 Suppl 2:19–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Gundersen C, Ziliak JP. Food insecurity and health outcomes. Health Aff. 2015;34(11):1830–9. [DOI] [PubMed] [Google Scholar]
- [4].New York City Taxi and Limousine Commission. Taxicab Factbook. 2018. Available from: http://www.nyc.gov/html/tlc/downloads/pdf/2018_tlc_factbook.pdf.
- [5].Gany FM, Gill PP, Ahmed A, Acharya S, Leng J. “Every disease...man can get can start in this cab”: focus groups to identify south Asian taxi drivers’ knowledge, attitudes and beliefs about cardiovascular disease and its risks. J Immigr Minor Health. 2013;15(5):986–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Mirpuri S, Gill P, Ocampo A, Roberts N, Narang B, Hwang SW, et al. Discrimination and Health Among Taxi Drivers in New York and Toronto. J Community Health. 2018;43(4):667–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Gundersen C, Dewey A, Kato M, Crumbaugh AS, Strayer M, Odeen B, et al. Map the Meal Gap 2019: A Report on County and Congressional District Food Insecurity and County Food Cost in the United States in 2017. Chicago, Illinois: Feeding America; 2019. [Google Scholar]
- [8].Gany F, Bari S, Gill P, Ramirez J, Ayash C, Loeb R, et al. Step On It! Workplace Cardiovascular Risk Assessment of New York City Yellow Taxi Drivers. J Immigr Minor Health. 2016;18(1):118–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].New York City Department of Health and Hygeine. Obesity [Web page]. Available from: https://www1.nyc.gov/site/doh/health/health-topics/obesity.page.
- [10].Mirpuri S, Ocampo A, Narang B, Roberts N, Gany F. Discrimination as a social determinant of stress and health among New York City taxi drivers. J Health Psychol. 2018:1359105318755543. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Mirpuri S, Traub K, Romero S, Hernandez M, Gany F. Cardiovascular health status of taxi/for-hire vehicle drivers in the United States: A systematic review. Work. 2021;69(3):927–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Apantaku-Onayemi F, Baldyga W, Amuwo S, Adefuye A, Mason T, Mitchell R, et al. Driving to better health: cancer and cardiovascular risk assessment among taxi cab operators in Chicago. J Health Care Poor Underserved. 2012;23(2):768–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Mirpuri S, Riley K, Gany F. Taxi drivers and modifiable health behaviors: Is stress associated? Work. 2021;69(4):1283–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Bureau of Labor Statistics. Taxi Drivers, Ride-Hailing Drivers, and Chauffeurs (Occupational Outlook Handbook; ) 2017. Available from: https://www.bls.gov/ooh/transportation-and-material-moving/taxi-drivers-and-chauffeurs.htm. [Google Scholar]
- [15].Bureau of Labor Statistics. Labor Force Statistics from the Current Population Survey Washington, DC: US Government; 2019. [updated January 22, 2020]. Available from: https://www.bls.gov/cps/cpsaat18.htm. [Google Scholar]
- [16].Griswold A, Kopf D. The number of American taxi drivers has tripled in a decade New York, NY: Quartz; 2019. [updated June 21, 2019]. Available from: https://qz.com/1647247/more-americans-are-driving-taxis-thanks-to-uber-and-lyft/. [Google Scholar]
- [17].Rosenthal BM. ‘They Were Conned’: How Reckless Loans Devastated a Generation of Taxi Drivers. New York Times. 2019. May 19, 2019. [Google Scholar]
- [18].Hartman M What makes gig economy workers anxious? : Minnesota Public Radio; 2018. [updated March 8, 2018]. Available from: https://www.marketplace.org/2018/03/08/economy/anxiety-index/gig-workers-and-economically-anxious-lifestyle. [Google Scholar]
- [19].Iqbal M Uber Revenue and Usage Statistics. 2018. Available from: http://www.businessofapps.com/data/uber-statistics/#4.
- [20].Tran M, Sokas RK. The Gig Economy and Contingent Work: An Occupational Health Assessment. J Occup Environ Med. 2017;59(4):e63–e6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].NYC Food Policy. Food Metrics Report 2019. New York, NY: NYC Mayor’s Office of Food Policy; 2020. [Google Scholar]
- [22].Bureau of Labor Statistics. Delivery Truck Drivers and Driver/Sales Workers Washington, DC: Office of Occupational Statistics and Employment Projections; 2019. [updated June 12, 2019]. Available from: https://www.bls.gov/ooh/transportation-and-material-moving/delivery-truck-drivers-and-driver-sales-workers.htm#tab-1. [Google Scholar]
- [23].Bureau of Labor Statistics. A profile of the working poor, 2017. Washington, DC: US Bureau of Labor Statistics; 2019. [Google Scholar]
- [24].Costas-Muniz R, Roberts N, Narang B, Mehmood R, Acharya S, Aragones A, et al. Colorectal Cancer Screening and Access to Healthcare in New York City Taxi Drivers. J Immigr Minor Health. 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Swindle TM, Whiteside-Mansell L, McKelvey LJ. Food Insecurity: Validation of a Two-Item Screen Using Convergent Risks. Child Fam Stud. 2013;22(7):932–41. [Google Scholar]
- [26].United States Department of Agriculture. U.S. household food security survey module: three stage design, with screeners. Washington, DC: Economic Research Service; 2008. [Google Scholar]
- [27].Hager ER, Quigg AM, Black MM, Coleman SM, Heeren T, Rose-Jacobs R, et al. Development and validity of a 2-item screen to identify families at risk for food insecurity. Pediatrics. 2010;126(1):e26–32. [DOI] [PubMed] [Google Scholar]
- [28].Gundersen C, Engelhard EE, Crumbaugh AS, Seligman HK. Brief assessment of food insecurity accurately identifies high-risk US adults. Public Health Nutr. 2017;20(8):1367–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Makelarski JA, Abramsohn E, Benjamin JH, Du S, Lindau ST. Diagnostic Accuracy of Two Food Insecurity Screeners Recommended for Use in Health Care Settings. Am J Public Health. 2017;107(11):1812–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Hadden KB, Prince LY, Rojo MO, Selig JP, McElfish PA. Screening Patients Who Speak Spanish for Low Health Literacy. Health Lit Res Pract. 2019;3(2):e110–e6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [31].Sentell T, Braun KL. Low health literacy, limited English proficiency, and health status in Asians, Latinos, and other racial/ethnic groups in California. J Health Commun. 2012;17 Suppl 3:82–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Medical Expenditure Panel Survey: Access to Care (AC) Section Rockville, MD: Agency for Healthcare Research and Quality; 2014. Available from: https://meps.ahrq.gov/survey_comp/hc_survey/2014/AC111214.htm. [Google Scholar]
- [33].Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System Questionnaire. Atlanta, GA; 2014. Available from: https://www.cdc.gov/brfss/questionnaires/pdf-ques/2014_BRFSS.pdf. [Google Scholar]
- [34].Medical Expenditure Panel Survey Rockville, MD: Agency for Healthcare Research and Quality; 2016. [updated 2016]. Available from: https://meps.ahrq.gov/mepsweb/. [Google Scholar]
- [35].Sangha O, Stucki G, Liang MH, Fossel AH, Katz JN. The Self-Administered Comorbidity Questionnaire: a new method to assess comorbidity for clinical and health services research. Arthritis Rheum. 2003;49(2):156–63. [DOI] [PubMed] [Google Scholar]
- [36].Adam S, van de Poll-Franse LV, Mols F, Ezendam NPM, de Hingh I, Arndt V, et al. The association of cancer-related fatigue with all-cause mortality of colorectal and endometrial cancer survivors: Results from the population-based PROFILES registry. Cancer Med. 2019;8(6):3227–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [37].Schwartz CE, Zhang J, Stucky BD, Michael W, Rapkin BD. Is the link between socioeconomic status and resilience mediated by reserve-building activities: mediation analysis of web-based cross-sectional data from chronic medical illness patient panels. BMJ Open. 2019;9(5):e025602. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [38].Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav. 1983;24(4):385–96. [PubMed] [Google Scholar]
- [39].Centers for Disease Control and Prevention. Healthy Weight: Assessing your Weight Atlanta, Georgia: U.S. Department of Health & Human Services; 2015. Available from: http://www.cdc.gov/healthyweight/assessing/. [Google Scholar]
- [40].Shokoohyar S Ride-sharing platforms from drivers’ perspective: Evidence from Uber and Lyft drivers. Int J Data Netw Sci. 2018;2. [Google Scholar]
- [41].RadioFreeTaxi. Driving Yellow Cab in New York City New York, NY: YellowCabNYC.com; 2018. [updated 2019 March 13]. Available from: https://www.yellowcabnyctaxi.com/uncategorized/driving-yellow-cab-in-new-york-city#comment-1250831. [Google Scholar]
- [42].Murray KE, Buul A, Aden R, Cavanaugh AM, Kidane L, Hussein M, et al. Occupational health risks and intervention strategies for US taxi drivers. Health Promot Int. 2019;34(2):323–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [43].Blasi G, Leavitt J. Driving Poor: taxi drivers and the regulation of the taxi industry in Lost Angeles Los Angeles, U.S. Project funded by the ULCA Institute of Industrial Relation. 2006. Available from: http://www.taxi-library.org/driving-poor.pdf.
- [44].Silverman J, Krieger J, Kiefer M, Hebert P, Robinson J, Nelson K. The Relationship Between Food Insecurity and Depression, Diabetes Distress and Medication Adherence Among Low-Income Patients with Poorly-Controlled Diabetes. J Gen Intern Med. 2015;30(10):1476–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [45].Montgomery J, Lu J, Ratliff S, Mezuk B. Food Insecurity and Depression Among Adults With Diabetes: Results From the National Health and Nutrition Examination Survey (NHANES). Diabetes Educ. 2017;43(3):260–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [46].Seligman HK, Laraia BA, Kushel MB. Food insecurity is associated with chronic disease among low-income NHANES participants. J Nutr. 2010;140(2):304–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [47].Pourmotabbed A, Moradi S, Babaei A, Ghavami A, Mohammadi H, Jalili C, et al. Food insecurity and mental health: A systematic review and meta-analysis. Public Health Nutr. 2020;23(10):1778–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [48].Brondolo E, Libby DJ, Denton EG, Thompson S, Beatty DL, Schwartz J, et al. Racism and ambulatory blood pressure in a community sample. Psychosom Med. 2008;70(1):49–56. [DOI] [PubMed] [Google Scholar]
- [49].Restrepo J, Lemos M. Addressing psychosocial work-related stress interventions: A systematic review. Work. 2021;70:53–62. [DOI] [PubMed] [Google Scholar]
- [50].Seligman HK, Bindman AB, Vittinghoff E, Kanaya AM, Kushel MB. Food insecurity is associated with diabetes mellitus: results from the National Health Examination and Nutrition Examination Survey (NHANES) 1999-2002. J Gen Intern Med. 2007;22(7):1018–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [51].Centers for Disease Control and Prevention. National Diabetes Statistics Report 2020. 2020. [Google Scholar]
- [52].Food and Nutrition Service. Supplemental Nutrition Assistance Program (SNAP) Eligibility Alexandria, VA: US Department of Agriculture; 2016. Available from: http://www.fns.usda.gov/snap/eligibility.] [Google Scholar]
- [53].Coleman-Jensen AR, Matthew P. Gregory Christian A.; Singh Anita. Household Food Security in the United States in 2017. United States Department of Agriculture; 2018. September, 2018. Contract No.: ERR-256. [Google Scholar]
- [54].Committee on Examination of the Adequacy of Food R, Allotments S, Food, Nutrition B, Committee on National S, Institute of M, et al. Conclusions and Recommendations. In: Caswell JA, Yaktine AL, editors. Supplemental Nutrition Assistance Program: Examining the Evidence to Define Benefit Adequacy. Washington (DC): National Academies Press (US); 2013. [PubMed] [Google Scholar]
- [55].Food Bank for New York City. Research, Reports and Financials. New York, NY; 2018. Available from: https://www.foodbanknyc.org/research-reports/?gclid=EAIaIQobChMImIO-lciv4wIVjq_ICh12dw0oEAAYASAAEgIcEfD_BwE. [Google Scholar]
- [56].Weinfield NS, Mills G, Borger C, Gearing M, Macaluso T, Montaquila J, et al. Hunger in America. Chicago, IL: Feeding America; 2014. [Google Scholar]
- [57].Wolfson JA, Leung CW. Food Insecurity and COVID-19: Disparities in Early Effects for US Adults. Nutrients. 2020;12(6). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [58].Gany F, Bari S, Gill P, Loeb R, Leng J. Step on it! Impact of a workplace New York City taxi driver health intervention to increase necessary health care access. Am J Public Health. 2015;105(4):786–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [59].Ames AJ, Barnett TM. Psychometric Validation of the 10-Item USDA Food Security Scale for Use with College Students. J Appl Meas. 2019;20(3):228–42. [PubMed] [Google Scholar]
- [60].Glasziou PP, Irwig L, Kirby AC, Tonkin AM, Simes RJ. Which lipid measurement should we monitor? An analysis of the LIPID study. BMJ Open. 2014;4(2):e003512–e. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [61].Aalto AM, Elovainio M, Kivimäki M, Uutela A, Pirkola S. The Beck Depression Inventory and General Health Questionnaire as measures of depression in the general population: a validation study using the Composite International Diagnostic Interview as the gold standard. Psychiatry Res. 2012;197(1–2):163–71. [DOI] [PubMed] [Google Scholar]
- [62].Landsbergis PA, Schnall PL, Belkić KL, Baker D, Schwartz J, Pickering TG. Work stressors and cardiovascular disease. Work. 2001;17:191–208. [PubMed] [Google Scholar]
- [63].Long H A New York City taxi driver kept working to pay bills. Now he and his family are sick. Washington Post. 2020 April 4, 2020. [Google Scholar]
- [64].Hu W, Schweber N. Coronavirus Decimates N.Y.C. Taxi Industry: ‘The Worst It’s Ever Been’. New York Times. 2020. March 25, 2020. [Google Scholar]