Consumption of Foods Away from Home Is Associated with Lower Diet Quality Among Adults Living in Puerto Rico

Abstract

Background

Consuming foods away from home (FAFH) is ubiquitous, yet, it is unclear how it influences diet in diverse populations.

Objective

The study aimed to evaluate the association between frequency and type of consumption of FAFH and diet quality.

Design

The study had a cross-sectional design. Participants self-reported the frequency of consuming FAFH as “rarely” (≤1 time per week) vs “frequently” (≥2 times per week) at various commercial establishments or noncommercial FAFH (ie, friends’ or relatives’ homes).

Participants/setting

Participants were adults (aged 30 through 75 years) from the PRADLAD (Puerto Rico Assessment of Diet, Lifestyle, and Diseases) study conducted in San Juan, Puerto Rico metro area (n = 239) in 2015.

Main outcome measures

A validated food frequency questionnaire captured dietary intake. The Alternate Healthy Eating Index-2010 defined diet quality. Secondary outcomes included whether participants met 2015-2020 Dietary Guidelines for Americans recommendations for sodium, added sugars, saturated fat, dietary fiber, total energy, and alcohol.

Statistical analyses performed

Linear or logistic regression models adjusted for age, sex, employment, income, education, and food insufficiency tested differences in mean Alternate Healthy Eating Index-2010 scores or odds of meeting (vs not meeting) intake recommendations by FAFH type and frequency.

Results

Overall, 54.4% and 37.2% of participants reported consuming commercial FAFH and noncommercial FAFH “frequently,” respectively. Consuming FAFH “frequently” (vs “rarely”) was associated with lower mean Alternate Healthy Eating Index-2010 scores for both commercial FAFH (57.92 vs 63.58; P = .001) and noncommercial FAFH (56.22 vs 62.32; P < .001). Consuming commercial FAFH “frequently” (vs “rarely”) at any type of food establishment was associated with lower odds of meeting the dietary fiber Dietary Reference Intakes (odds ratio 0.43; 95% CI 0.23 to 0.81). Consuming noncommercial FAFH “frequently” was associated with lower odds of meeting recommendations for sodium (odds ratio 0.30; 95% CI 0.11 to 0.79) and added sugars (odds ratio 0.41; 95% CI 0.18 to 0.93).

Conclusions

Frequent consumption of FAFH is associated with lower diet quality and lower adherence to dietary recommendations in Puerto Rico. Future studies should explore whether diet quality can be improved by prioritizing healthy at-home meals and reformulating the quality of commercial FAFH.

ANNUAL SPENDING ON FOODS CONSUMED AWAY from home (FAFH) by individuals in the mainland United States has risen dramatically over the past few decades.¹ The market share of FAFH purchases surpassed purchases of foods for consumption at home in 2010 (50.2%), up from 44% in 1987.² Eating away from home is generally associated with higher energy intake and lower diet quality than eating meals at home.^3-7 Although approximately two-thirds of US residents perceive FAFH at commercial establishments (ie, fast-food, casual restaurants, and food trucks) to be unhealthy,⁸ most tend to underestimate the fat, energy, and sodium in these foods.³

Given that environmental and lifestyle factors are important drivers of the global increase in overweight and obesity,⁹ there is a need to rigorously understand the relationship between FAFH consumption and diet quality. These factors are particularly relevant to adults living in Puerto Rico, a US territory. Although eating patterns of Hispanic and Latino populations have been studied in the mainland United States,^10-13 differences in environmental and lifestyle factors exist between the island and the mainland United States. Current evidence suggests that poor dietary habits and prevalence of chronic disease are higher in Puerto Rico than in the mainland, including for hypertension (42% vs 31%), high cholesterol (39% vs 36%), diabetes (16% vs 10%), and coronary heart disease (9% vs 6%).^14-17 A recent assessment of food intake among adults in Puerto Rico found that the top energy sources were sugary beverages (11.8%), followed by sweets and desserts (10.2%), dairy (8.5%), mixed dishes and soups (7.6%), starchy vegetables (6.3%), fast foods (5.5%), and rice (4.9%).¹⁸ Considering the unhealthy diet quality that currently characterizes the Puerto Rican diet, there is a need to understand the potential contributions to dietary intake from different establishments and contexts.

Fast-food chains have been in Puerto Rico since 1959, and their concentration has increased since 2005, with 59 fast-food establishments per 100,000 citizens.¹⁹ A supplement to the Behavioral Risk Factor Surveillance System in 2013 showed that 73.8% of the adult population in Puerto Rico consumed FAFH at fast-food establishments approximately 6.7 times per month, with lunch most commonly consumed at fast-food establishments.¹⁹ Overall, results of the survey showed no significant differences in several cardiometabolic conditions between those who consume any FAFH and the total sampled population in Puerto Rico. However, the survey did not collect detailed information on the types of food and diet quality consumed in FAFH establishments.

In addition to nationally recognized fast-food and casual restaurant chains, Puerto Rico has other types of eating establishments.²⁰ Locally owned restaurants abound in Puerto Rico, serving both local and international cuisines and offering table service. Cafeterias–smaller establishments where customers can select from various meal items–typically serve Puerto Rican food in an informal setting with limited, or without, table service. Food trucks and food carts are standard in urban areas and offer novel food varieties. These commercial sources of FAFH (C-FAFH) in Puerto Rico resemble alternatives elsewhere in the United States that adversely influence energy intake, diet quality, and health outcomes.^21-25 Yet, little is known about how C-FAFH influences dietary intake in Puerto Rico.

Furthermore, to comprehensively assess the nutritional implications of FAFH in Puerto Rico, the study of eating-out behaviors needs to be extended to other sources of FAFH that predominate in the food environment. Noncommercial FAFH (NC-FAFH; ie, food prepared by others for sharing) may be an important source of FAFH in Puerto Rico because preparing food to share is a traditional cultural value.^26,27 The concept of “social facilitation of eating” describes that people eat more in groups than alone, especially when the group is composed of friends or family.²⁸ Furthermore, a study among adults in the United States showed that consuming unhealthy snacks, sugar-sweetened beverages, and nonfruit dessert at someone else’s home is reported more frequently than at home.²⁹ Family meals have been associated with healthy dietary outcomes, particularly for children.³⁰ In contrast, meals away from home meals with relatives have been associated with unhealthy dietary outcomes for Hispanic and Latino children^30,31 and adults.¹⁰

Therefore, this study aimed to determine relationships between frequency and type of FAFH with diet quality and meeting dietary intake recommendations among adults in the San Juan metropolitan area of Puerto Rico, controlling for important sociodemographic, sociocultural, and lifestyle confounders. To assess whether there are differences in associations by type of meal consumed as FAFH, a secondary analysis evaluated the association of FAFH consumption by meal occasion (eg, breakfast, lunch, dinner, and snacks) with diet quality and recommended intake. It was hypothesized that higher C-FAFH and NC-FAFH frequency would be associated with lower diet quality and less adherence to dietary recommendations compared with lower FAFH consumption. Furthermore, diet quality and recommended intake would vary by type of C-FAFH establishment.

METHODS

Study Population, Setting, and Design

Data were obtained from 380 adults participating in a cross-sectional study, the Puerto Rico Assessment of Diet, Lifestyle, and Diseases (PRADLAD), conducted in 2015 to assess lifestyle risk factors and health among adults from the San Juan metropolitan area. The study design and methodology have been described in detail previously.³² Briefly, participants were patients waiting for a medical appointment or visitors recruited from 3 primary care clinics selected for their strategic locations, facilities, and sociodemographic representation in the San Juan metropolitan area. Eligible individuals were living in Puerto Rico at the time of the study for at least 10 months of the previous year and were between 30 and 75 years of age and able to answer questions in Spanish without assistance. All participants provided written informed consent. The Institutional Review Board at Harvard T.H. Chan School of Public Health, Ponce Health Sciences University, University of Massachusetts Lowell, and Northeastern University approved the study.

Data Collection

Trained, Spanish-speaking interviewers administered all questionnaires used in the study in a private room in the clinic where the participant was recruited. Data were collected and managed using the secure, web-based electronic data capture tool Research Electronic Data Capture.³³ Questionnaires used to collect general self-reported background characteristics were based on the Boston Puerto Rican Health Study³⁴ and the National Health and Nutrition Examination Survey.³⁵ Demographic and socioeconomic questions included sex at birth, age, ethnicity, household composition, educational attainment, marital status, work history, household income, health insurance coverage, food insufficiency (classified as having to skip a meal during the previous month due to lack of financial resources either “many times,” “sometimes,” or “never”), and participation in government-funded food-assistance programs, such as the income-based Programa de Asistencia Nutricional [Nutritional Assistance Program] and Special Supplemental Nutrition Program for Women, Infants, and Children.

Consumption of FAFH

A comprehensive questionnaire was developed to assess dietary behaviors and attitudes. Questions used in this analysis were adapted from the Food Attitudes and Behaviors Survey of the National Cancer Institute (Cronbach’s α coefficient ≥.68),^36-39 a validated dietary behaviors questionnaire for Latinos (Cronbach’s α coefficients .47 to .48),⁴⁰ and a questionnaire on FAFH for US-residing adults participating in the Hispanic Community Health Study/Study of Latinos.¹⁰ Participants were asked to indicate the frequency of consuming FAFH (ie, times per week) by meal occasion (eg, breakfast, lunch, dinner, and snacks), and context (eg, commercial, or noncommercial locations). Participants were first asked about C-FAFH. The section started by asking, “On average, in a usual week, how many times do you eat commercial foods prepared outside your household for breakfast?” Interviewers repeated this question for lunch, dinner, and snacks. A separate question was specific to the type of food establishment at which commercial meals (regardless of meal occasion) were consumed, “On average, in a usual week, how many times do you eat commercial meals prepared outside your household in each of these establishments …?” with specific questions asked for fast-food restaurants (ie, franchised fast-food restaurants); chain restaurants (ie, franchised sit-down restaurants); other restaurants (ie, independent sit-down restaurants); take-out (ie, ordering food for consumption elsewhere); cafeterias (ie, nonfranchised establishments where food is served according to customer’s order as opposed to a served meal); or food trucks (ie, mobile units where foods are precooked and meals are assembled at time of order). For all questions, potential answers were: “rarely or never (≤1 time per week),” “sometimes (2–4 times per week),” “many times (5–6 times per week), and “all of the time (every day).”

Participants were also asked about NC-FAFH, such as at a friend’s or relative’s home, where food is prepared by others but not sold. This section started by asking, “On average, in a usual week, how many times do you eat noncommercial foods prepared outside your household (eg, at your parent’s, friend’s, or relative’s house) for breakfast?” Interviewers repeated this question for lunch, dinner, and snacks. For all questions, potential answers were: “rarely or never (≤1 time per week),” “sometimes (2–4 times per week),” “many times (5–6 times per week),” and “all of the time (every day).”

Dietary Assessment and Outcomes

A semi-quantitative food frequency questionnaire (FFQ) adapted and validated for this population was used to assess usual dietary intake of 193 different foods over the previous 12 months.^41,42 The FFQ was administered by trained, Spanish-speaking interviewers and documented in Research Electronic Data Capture³³ and, for food and nutrient analyses, the file was linked with the Minnesota Nutrient Data System for Research software, version 5.0_35.^43-46 Participants with energy intakes <600 or >4,800 kcal/d or with 2 or more sections of the questionnaire left blank were excluded from dietary analyses. Food groups were created using reported serving equivalents of individual foods (or disaggregated mixed dishes). Nutrient intake excluded supplement use.

The Alternate Healthy Eating Index-2010 (AHEI) incorporates food and nutrient components associated with chronic disease risk.⁴⁷ AHEI-2010 scores were calculated from each completed FFQ. A continuous score was created for each of 11 food groups or nutrient components ranging from 0 for minimal observance of the recommended intake of the component to 10 points for maximal observance; intermediate values were prorated. To calculate an AHEI-10 score, scores from each component were added. Scores ranged from 0 (lowest diet quality) to 110 (highest diet quality). It has been reported previously that higher scores of diet quality based on AHEI-2010 are strongly associated with lower risks of chronic diseases, including among Puerto Ricans living in the mainland United States.^18,48

The 2015-2020 Dietary Guidelines for Americans (DGA)⁴⁹ were used to establish thresholds for adequacy or meeting recommended intakes of dietary fiber, sodium, added sugars, saturated fats, dietary energy intakes, and alcohol (thresholds shown in Table 1; available at www.jandonline.org).

Statistical Analysis

Descriptive statistics for demographic, socioeconomic, and lifestyle characteristics were evaluated by C-FAFH and NC-FAFH frequency, meal occasion, and type of food establishment using χ² tests for categorical variables and analysis of variance for continuous variables. Polynomial logistic regression models were used to determine associations of C-FAFH or NC-FAFH frequency and type of food establishment with meeting intake recommendations (vs not). Linear regression models were used to determine the association between C-FAFH or NC-FAFH frequency and type of food establishment with mean AHEI scores. Main analyses were for C-FAFH or NC-FAFH at combined meal occasions. In secondary analysis, similar models for C-FAFH or NC-FAFH were run by meal occasion, with mean AHEI scores or DGA as outcomes. Applying methodology similar to that used for the primary analysis described in this article and a previous study,¹⁰ models were adjusted for age, sex, educational attainment, employment status, income level, and food insufficiency. AHEI models were also adjusted for energy intake.

To improve statistical power, responses for frequency of consuming FAFH by meal and establishment type were aggregated into binary categories; 0 if “rarely or never” (≤1 time per week) and 1 if “sometimes” (2 to 4 times per week), “many times” (5 to 6 times per week), or “all of the time” (every day). The aggregated categories were renamed as “rarely” with a value of 0 and “frequently” with a value of 1. This was done for both C-FAFH and NC-FAFH at all combined meal occasions (ie, breakfast, lunch, dinner, and snacks), and for individual and combined C-FAFH establishment types (ie, fast food, restaurants, food truck, and cafeteria). To compare characteristics across all types of C-FAFH meal occasions, a total binary variable was created as 0 for “rarely or never” consuming C-FAFH for breakfast, lunch, dinner, and snacks, or 1 for consuming C-FAFH for at least 1 of these meal occasions. The same procedure was applied to develop a total binary variable across meal types as NC-FAFH.

Results were aggregated as described to improve sample size and power due to low response frequencies for frequent FAFH consumption (ie, “many times” and “all of the time”) across meal occasions and establishment types. Analysis conducted using 3 categories (rarely or never, sometimes, or many times plus all of the time) resulted in similar results regarding magnitude and significance, albeit slightly attenuated. From the 380 PRADLAD participants, individuals with missing FFQ data (n = 132) were excluded from analysis. Participants without FFQ data had sociodemographic characteristics similar to those with FFQ data, except that fewer participants with FFQ data were not employed.¹⁸ In addition, 9 participants were excluded due to missing values for meal occasion and/or FAFH establishment type. The final sample size for these analyses was 239 adults. All analyses were conducted using Stata software, version 16.0.⁵⁰ Significance level was set at P < .05.

RESULTS

Participant Characteristics

C-FAFH by Establishment Type.

A majority of participants (54.4%) reported “frequently” consuming C-FAFH at any establishment type and 45.6% reported “rarely” consuming C-FAFH at the evaluated establishment types (Table 2). “Frequently” consuming C-FAFH at any establishment type was more often reported by individuals who were aged 30 through 59 years (vs 60 through 75 years), identified as female (vs male), reported educational attainment of eighth grade and beyond, were not employed (vs employed), lived in households with annual income ≤$20,000 (vs income >$20,000), and were not participants in the Programa de Asistencia Nutricional (vs participants). C-FAFH were mostly from fast-food restaurants (38.9%), followed by cafeterias (35.2%), and take-out (28.9%). Income differed significantly by frequency of consuming C-FAFH at various establishment types. For example, fewer individuals living in households with annual household income >$20,000 per year reported consuming C-FAFH “frequently” at fast-food locations (14.6%), chain restaurants (17.2%), other restaurants (19.6%), and cafeterias (23.7%) compared with those with income ≤$20,000. For most establishment types, the proportion of employed individuals consuming C-FAFH was considerably higher than those not employed, except for fast-food establishments. Those who “frequently” consumed C-FAFH at any establishment type were less likely be in the highest AHEI-2010 tertile and meet the recommendations for added sugars, dietary fiber, and saturated fats (Table 3). Compared with those who “rarely” consumed C-FAFH, individuals who “frequently” consumed C-FAFH had higher energy intake; lower AHEI-2010 score; and higher intake of sodium, added sugars, and saturated fats for each establishment type. Alcohol intake was higher among individuals who frequently consumed C-FAFH at chain restaurants, other restaurants, take-out, cafeterias, and food trucks.

Table 2.

Socioeconomic characteristics of participants in the PRADLAD^a study (2015) by frequency and type of establishment of commercial foods away from home^b

	Fast Food		Chain Restaurant		Other Restaurant		Take-Out		Cafeteria		Food Trucks		Total
Characteristic	Rarely	Frequently	Rarely	Frequently	Rarely	Frequently	Rarely	Frequently	Rarely	Frequently	Rarely	Frequently	Rarely	Frequently
n (%)
Participants	146 (61.1)	93 (38.9)	177 (74.1)	62 (25.9)	179 (74.9)	60 (25.1)	170 (71.1)	69 (28.9)	155 (64.9)	84 (35.2)	197 (82.4)	42 (17.6)	109 (45.6)	130 (54.4)
%
Age group
30-59 y	72.6**	88.2**	74.6**	90.3**	77.1	83.3	75.9	85.5	75.5	84.5	76.6	88.1	70.6**	85.4**
60-75 y	27.4**	11.8**	25.4**	9.7**	22.9	16.7	24.1	14.5	24.5	15.5	23.4	11.9	29.4**	14.6**
Female sex	68.5	67.7	73.4**	53.2**	71.5	58.3	70.6	62.3	73.5*	58.3*	70.6	57.1	75.2*	62.3*
Less than 8th grade education attainment	13.7	7.6	13.1	6.5	11.8	10	13	7.2	13.6	7.1	11.7	9.5	16.5*	7*
Employed	35.6	48.4	36.7*	51.6*	35.2**	56.7**	36.5*	50.7*	33.5**	53.6**	36.5**	59.5**	32.1*	47.7*
Annual household income >$20,000	22.8**	14.6**	20.5*	17.2*	23.5*	19.6*	19	21	17.3*	23.7*	21.8	10.3	15.5**	23.2**
Married or living with partner	66.7	72.2	65.5	76.7	67.5	71.8	65.8	76.2	65.7	74.1	68.4	70.0	66.7	72.2
Lives alone	27.4	22.6	26.6	22.6	25.1	26.7	25.9	24.6	27.7	21.4	24.4	31	28.4	23.1
Puerto Rican ethnicity	76	84.9	78.5	82.3	80.4	76.7	78.2	82.6	79.4	79.8	79.2	81	75.2	83.1
Has health insurance	78.1	75.3	74.6	83.9	75.4	81.7	74.7	82.6	76.1	78.6	76.1	81	75.2	78.5
Food insufficiency within past month c	53.8	53.8	51.7	59.7	55.6	48.3	54.4	52.2	51.3	58.3	53.6	54.8	56.5	51.5
WICd participatione	8.3	5.4	7.3	6.6	7.3	6.7	6.5	8.7	8.4	4.8	7.1	7.1	10.1	4.7
PANf participationg	52.7	45.2	54.2*	37.1*	54.7**	35**	52.9	42	58.7***	33.3***	51.3	42.9	62.4***	39.2***

^aPRADLAD = Puerto Rico Assessment of Diet, Lifestyle, and Diseases.

^bComparisons were made by frequency (“frequently” vs “rarely”) for each meal location using χ² for categorical variables and analysis of variance for continous variables. “Frequently” means consumption at that location ≥2 times per week, “rarely” means <2 times per week.

^cFood insuficiency was defined as skipping a meal due to lack of financial resources sometimes or many times during the last month.

^dWIC = Special Supplemental Nutrition Program for Women, Infants, and Children.

^eParticipation was defined as receipt of program funds by participant or a member of their household.

^fPAN = Programa de Asistencia Nutricional [Nutritional Assistance Program].

^gPAN participation was defined as receipt of program funds by participant or a member of their household.

^*P < .05.

^**P < .01.

^***P < .001.

Table 3.

Dietary characteristics of participants in the PRADLAD^a study (2015) by frequency of consumption and type of establishment of commercial foods away from home^b

	Fast Food		Chain Restaurant		Other Restaurant		Take-Out		Cafeteria		Food Trucks		Total
Characteristic	Rarely	Frequently	Rarely	Frequently	Rarely	Frequently	Rarely	Frequently	Rarely	Frequently	Rarely	Frequently	Rarely	Frequently
n (%)
Participant	146 (61.1)	93 (38.9)	177 (74.1)	62 (25.9)	179 (74.9)	60 (25.1)	170 (71.1)	69 (28.9)	155 (64.9)	84 (35.2)	197 (82.4)	42 (17.6)	109 (45.6)	130 (54.4)
kcal/d, mean (SD)
Energy intake	2,150 (836)*	2,400 (1,080)*	2,133 (852)**	2,573 (1,112)**	2,114 (843)***	2,644 (1,113)***	2,114 (847)***	2,575 (1,089)***	2,135 (836)*	2,455 (1,091)*	2,109 (835)***	2,897 (1,148)***	2,153 (841)	2,327 (1,019)
%
Meets energy recommendation	14.4	16.1	15.3	14.5	15.1	15	15.9	13	15.5	14.3	15.7	11.9	14.7	15.4
mean (SD)
AHEI-2010c score	62.7 (11.3)***	56.6 (10.1)***	62.0 (11.4)***	55.7 (9.3)***	61.8 (11.4)***	56.1 (9.5)***	62.4 (11.2)***	55.4 (9.7)***	62.5 (11.3)***	56.4 (10.0)***	61.7 (11.3)***	53.9 (8.1)***	64.4 (11.4)***	57.0 (9.9)***
%
In the highest AHEI-2010 tertile	42.5***	18.3***	38.4**	17.7**	38.0**	18.3**	40.6***	14.5***	40.6***	19.0***	38.1***	9.5***	46.8***	21.5***
mg/d, mean (SD)
Sodium intake	3,890 (1,762)*	4,505 (2,283)*	3,882 (1,774)**	4,836 (2,413)**	3,827 (1,751)***	5,030 (2,402)***	3,807 (1,752)***	4,923 (2,337)***	3,862 (1,709)**	4,623 (2,379)**	3,798 (1,705)***	5,681 (2,511)***	3,931 (1,784)	4,295 (2,156)
%
Meets sodium recommendation	18.5	14	18.1	12.9	19	10	20.6*	7.2*	17.4	15.5	19.3*	4.8*	18.3	15.4
g/d, mean (SD)
Added sugars intake	76.6 (52.3)**	95.9 (57.2)**	75.9 (49.6)***	107.4 (62.8)***	76.4 (49.4)***	107.0 (64.0)***	75.4 (49.9)***	105.4 (61.2)***	77.6 (51.8)*	96.1 (58.8)*	75.9 (50.5)***	122.8 (59.2)***	75.6 (51.4)*	91.2 (57.0)*
%
Meets added sugar recommendation	32.2**	15.1**	29.4*	14.5*	27.9	18.3	30.0*	14.5*	30.3*	16.7*	28.9**	9.5**	33.0*	19.2*
g/d, mean (SD)
Dietary fiber intake	26.1 (10.9)	24.5 (10.7)	25.6 (11.0)	25 (10.1)	25.6 (10.9)	25.3 (10.4)	26 (11.2)	24.3 (9.8)	26.0 (10.9)	24.6 (10.7)	25.3 (10.8)	26.3 (11)	27.5 (11.1)**	23.4 (10.3)*
%
Meets dietary fiber recommendation	49.3	36.6	45.2	42.9	46.4	38.3	47.6	36.2	49.0*	35.7*	45.2	40.4	56***	34.6***
g/d, mean (SD)
Saturated fat intake	24.5 (12.8)*	29.1 (16.4)*	24.6 (13.7)**	31.1 (15.6)**	24.3 (13.6)***	32.0 (15.5)***	24.0 (13.6)***	31.8 (15.0)***	24.6 (13.4)*	29.4 (15.9)*	24.3 (13.5)***	35.5 (15.2)***	24.3 (13.3)	27.9 (15.2)
%
Meets saturated fat recommendation	54.1*	39.8*	53.1*	35.5*	53.1*	35.0*	54.7**	33.3**	52.3	41.7	53.3**	26.2**	57.8**	40.8**
drink-equivalents/d, mean (SD)
Alcohol intake	0.17 (0.39)	0.36 (1.4)	0.15 (0.36)**	0.52 (1.7)**	0.13 (0.34)***	0.58 (1.7)***	0.14 (0.36)**	0.50 (1.6)**	0.13 (0.35)**	0.46 (1.4)**	0.15 (0.37)***	0.70 (2.0)***	0.14 (0.38)	0.33 (1.2)
%
Meets alcohol recommendation	95.9	93.5	96	91.9	96.1	91.7	95.3	94.2	95.5	94	95.4	92.9	95.5	94.6

^aPRADLAD = Puerto Rico Assessment of Diet, Lifestyle, and Diseases.

^bComparisons were made by frequency (“frequently” vs “rarely”) for each meal location using χ² for categorical variables and analysis of variance for continous variables. “Frequently” means consumption at that location ≥2 times per week, “rarely” means <2 times per week.

^cAHEI-2010 = Alternate Healthy Eating Index-2010.

^*P < .05.

^**P < .01.

^***P <.001.

C-FAFH by Meal Occasion.

More than one-half of participants (62.3%) reported “frequently” consuming C-FAFH for any meal occasion and 37.7% reported “rarely” consuming C-FAFH (Table 4; available at www.jandonline.org). “Frequently” consuming C-FAFH was more often reported by individuals who were aged 30 through 59 years (vs 60 through 75 years), reported educational attainment of eighth grade and beyond, were not employed (vs employed), and were not participating in the Programa de Asistencia Nutricional (vs participants). Among meal occasions, lunch was most reported (54.4%) as “frequently” consuming C-FAFH, followed by snacks (32.6%), and dinner (31.4%). Reporting female sex; annual household income; marital status; household size; ethnicity; health insurance; food insufficiency; and Special Supplemental Nutrition Program for Women, Infants, and Children participation did not differ significantly by frequency of C-FAFH consumption across meal occasions. Compared with those who “rarely” consumed C-FAFH, those who “frequently” consumed C-FAFH had higher intake of energy, sodium, added sugars, and saturated fat for each meal occasion (Table 5; available at www.jandonline.org). Among participants, those who “frequently” consumed C-FAFH were less likely to meet the recommendations for dietary fiber across all meal occasions.

NC-FAFH by Meal Occasion.

The frequency of “frequently” consuming NC-FAFH for any meal type was 37.2% compared with 62.8% “rarely” consuming NC-FAFH (Table 6; available at www.jandonline.org). Among meal occasions, dinner was most highly reported (27.6%) as “frequently” consuming NC-FAFH, followed by lunch (27.2%) and snacks (17.6%). “Frequently” consuming NC-FAFH was more often reported by individuals who were aged 30 through 59 years (vs 60 through 75 years). Compared with those who “rarely” consumed NC-FAFH, individuals “frequently” consuming NC-FAFH at any meal occasion had higher energy intake; lower AHEI score; and higher intake of sodium, added sugars, saturated fat, and alcohol (Table 7; available at www.jandonline.org). Other dietary factors did not differ significantly by frequency of NC-FAFH.

Dietary Recommendation Outcomes

C-FAFH and NC-FAFH.

The odds of meeting the recommendation for dietary fiber were significantly lower for individuals who “frequently” consumed C-FAFH (OR 0.43; 95% CI 0.23 to 0.81) compared with those who “rarely” consumed C-FAFH (Figure 1, Table 8; available at www.jandonline.org). The odds of meeting recommendations for sodium were significantly lower (OR 0.30; 95% CI 0.11 to 0.79) for individuals “frequently” consuming NC-FAFH compared with those who “rarely” consumed NC-FAFH. The same pattern was observed for meeting added sugar recommendations (OR 0.41; 95% CI 0.18 to 0.93) for those “frequently” consuming NC-FAFH. There were no significant associations between the frequency of C-FAFH and meeting recommendations for dietary energy, sodium, added sugars, saturated fats, or alcohol. Similarly, there were no significant associations between consuming NC- FAFH and meeting recommendations for dietary energy, dietary fiber, saturated fat, or alcohol.

Figure 1. Adjusted odds ratios (ORs) and 95% CIs of meeting nutrient intake recommendations by consuming commercial (any location) and noncommercial FAFH “frequently” (vs “rarely”) at any meal occasion among participants in the PRADLAD (Puerto Rico Assessment of Diet, Lifestyle, and Diseases) study (2015).^a.

^aSignificance is indicated as *P < .05; **P < .01. Logistic regression adjusted for age, sex, employment status, income, educational attainment, and food insufficiency. For all commercial FAFH and noncommercial FAFH output variables, n = 208, except alcohol (n = 177). Cutoffs for dietary recommendations were set as follows (described for the relevant ages based on the study population): daily dietary energy intakes (women: 1,800-2,400 kcal for age 30 y; 1,800-2,200 kcal for age 31-50 y; 1,600-2,200 kcal for age 51-60 y; 1,600-2,000 kcal for age 61-75 y; and men: 2,400-3,000 kcal for age 30-35 y; 2,400-2,800 kcal for age 36-40 y; 2,200-2,800 for age 41-55 y; 2,200-2,600 for age 56-60 y; 2,000-2,600 for age 61-75 y), sodium (<2,300 mg/d), dietary fiber (women: ≥28 g/d for age 30 y, ≥25.2 g/d for age 31-50 y, ≥22.4 g/d for age 51-70 y, ≥22.4 g/d for age >70 y; men: ≥33.6 g/d for age 30 y, ≥30.8 g/d for age 31-50 y, ≥28 g/d for age 51-70 y, ≥28 g/d for age >70 y), saturated fats (<10% of daily estimated energy), added sugars (<10% of daily estimated energy), and alcohol (1 drink-equivalent or less per day for women and 2 drink-equivalents or less per day for men). ^bFAFH = foods away from home.

C-FAFH by Establishment Type.

Compared with “rarely” consuming C-FAFH, “frequently” consuming C-FAFH was associated with significantly lower odds of meeting the sodium recommendation at other restaurants (OR 0.33; 95% CI 0.12 to 0.96), take-out (OR 0.28; 95% CI 0.10 to 0.78), and food-trucks (OR 0.17; 95% CI 0.04 to 0.78) (Figure 2, Table 9; available at www.jandonline.org). “Frequently” consuming C-FAFH was significantly associated with lower odds of meeting the recommendation for saturated fat when meals were consumed at fast-food locations (OR 0.48; 95% CI 0.26 to 0.89), chain restaurants (OR 0.41; 95% CI 0.21 to 0.82), other restaurants (OR 0.27; 95% CI 0.13 to 0.57), take-out (OR 0.35; 95% CI 0.18 to 0.69), or food trucks (OR 0.28; 95% CI 0.12 to 0.65), compared with “rarely” consuming C-FAFH at these locations. “Frequently” consuming C-FAFH was significantly associated with lower odds of meeting the added sugar recommendation when meals were consumed at fast-food locations (OR 0.31; 95% CI 0.13 to 0.73), cafeterias (OR 0.44; 95% CI 0.20 to 0.99), or food trucks (OR 0.23; 95% CI 0.06 to 0.84), compared with “rarely” consuming C-FAFH at each of these locations. “Frequently” consuming C-FAFH at other restaurant locations was associated with lower odds (OR 0.20; 95% CI 0.04 to 0.92) of meeting alcohol recommendations compared with “rarely” consuming C-FAFH at other restaurant locations. No significant associations were identified among C-FAFH locations and dietary energy or fiber (Figure 2).

Figure 2. Adjusted odds ratios (ORs) and 95% CIs of meeting nutrient intake recommendations by type of establishment of commercial FAFH^b “frequently” (vs “rarely”) among participants of the PRADLAD (Puerto Rico Assessment of Diet, Lifestyle, and Diseases) study (2015).^a.

^aSignificance is indicated as *P < .05; **P < .01. Logistic regression adjusted for age, sex, employment status, income, educational attainment, and food sufficiency. n = 208 with the exception of alcohol (n = 177). Cutoffs for dietary recommendations were set as follows: daily dietary energy intakes (women: 1,800-2,400 kcal for age 30 y; 1,800-2,200 kcal for age 31-50 y; 1,600-2,200 kcal for age 51-60 y; 1,600-2,000 kcal for age 61-75 y; and men: 2,400-3,000 kcal for age 30-35 y; 2,400-2,800 kcal for age 36-40 y; 2,200-2,800 for age 41-55 y; 2,200-2,600 for age 56-60 y; 2,000-2,600 for age 61-75 y), sodium (<2,300 mg/d), dietary fiber (women: ≥28 g/d for age 30 y, ≥25.2 g/d for age 31-50 y, ≥22.4 g/d for age 51-70 y, ≥22.4 g/d for age >70 y; men: ≥33.6 g/d for age 30 y, ≥30.8 g/d for age 31-50 y, ≥28 g/d for age 51-70 y, ≥28 g/d for age >70 y), saturated fats (<10% of daily estimated energy), added sugars (<10% of daily estimated energy), and alcohol (1 drink-equivalent or less per day for women and 2 drink-equivalents or less per day for men). ^bFAFH = foods away from home.

C-FAFH and NC-FAFH by Meal Occasion.

“Frequently” (vs “rarely”) consuming C-FAFH at breakfast was significantly associated with a lower likelihood of meeting the saturated fat recommendation (OR 0.40; 95% CI 0.21 to 0.79) (Table 10; available at www.jandonline.org). C-FAFH consumed “frequently” at lunch was also significantly associated with lower odds of meeting the dietary fiber recommendations (OR 0.45; 95% CI 0.25 to 0.84). C-FAFH consumption at dinner or for snacks was not significantly associated with any of the DGA recommendations.

For NC-FAFH, consuming “frequently” (vs “rarely”) was significantly associated with lower odds of meeting the sodium recommendations at all meal occasions (breakfast: OR 0.11; 95% CI 0.01 to 0.90; lunch: OR 0.23; 95% CI 0.07 to 0.82; dinner: OR 0.24; 95% CI 0.07 to 0.74; snacks: OR 0.09; 95% CI 0.01 to 0.67), as well as for meeting the saturated fat recommendation (breakfast: OR 0.31; 95% CI 0.13 to 0.74; lunch: OR 0.48; 95% CI 0.25 to 0.95; dinner: OR 0.50; 95% CI 0.26 to 0.97; snacks: OR 0.46; 95% CI 0.22 to 0.98) (Table 11; available at www.jandonline.org). No significant associations were observed for NC-FAFH consumption and other DGA recommendations by meal occasion.

Diet Quality Outcomes

C-FAFH and NC-FAFH.

In adjusted linear regression models, “frequently” consuming C-FAFH or NC-FAFH was significantly associated with lower mean AHEI scores (Figure 3, Table 12; available at www.jandonline.org) compared with “rarely” consuming C-FAFH or NC-FAFH. “Frequently” consuming C-FAFH was significantly associated with mean AHEI-2010 scores that were 5.66 points lower (P = .001) compared with “rarely” consuming C-FAFH. Similarly, compared with “rarely” consuming NC- FAFH, “frequently” consuming NC-FAFH was significantly associated with 6.10-point lower mean AHEI-2010 scores (P < .001). For all commercial establishment types, “frequently” consuming C-FAFH was significantly associated with lower mean AHEI scores compared with “rarely” consuming FAFH at: fast-food restaurants (56.43 vs 62.48; P < .001); chain restaurants (55.96 vs 61.73; P = .002), other restaurant locations (54.14 vs 62.06; P < .001), take-out (55.45 vs 62.11; P < .001), cafeterias (56.26 vs 62.35; P < .001), and food trucks (54.17 vs 61.50; P = .001).

Figure 3. Adjusted mean (SE) of the Alternate Healthy Eating Index-2010 (AHEI-2010) scores by consuming C-FAFH and NC-FAFH “frequently” (vs “rarely”) at any meal occasion among participants of the PRADLAD (Puerto Rico Assessment of Diet, Lifestyle, and Diseases) study (2015).^a.

^aResults are significant at the 1% significance level. Linear regression models adjusted for age, sex, employment status, income, educational attainment, food security status, and energy intake. For all output variables n = 208. ^bNC-FAFH = noncommercial foods away from home. ^cC-FAFH = commercial foods away from home.

FAFH by Meal Occasion.

Lastly, consuming C-FAFH “frequently” vs “rarely” was significantly associated with lower mean AHEI scores at breakfast (55.56 vs 61.84; P = .001), lunch (57.08 vs 63.67; P < .001), and dinner (57.68 vs 61.21; P = .046), but not for snacks (59.98 vs 60.20; P = .899) (Table 13; available at www.jandonline.org). Significant associations were also identified for “frequently” consuming NC-FAFH and lower mean AHEI-2010 scores compared with “rarely” consuming NC-FAFH at breakfast (52.51 vs 61.61; P < .001), lunch (55.66 vs 61.73; P = .001), dinner (54.71 vs 62.32; P < .001), and snacks (56.24 vs 61.08; P = .022).

DISCUSSION

Among adults in the San Juan metropolitan area in Puerto Rico, consuming FAFH at commercial establishments was associated with lower overall diet quality and lower likelihood of following dietary recommendations for fiber, confirming the predefined hypothesis. This analysis confirmed the hypothesis that a higher frequency of C-FAFH was associated with lower diet quality. In addition, higher frequency of consuming NC-FAFH was associated with lower diet quality and lower odds of adherence to dietary recommendations for sodium and added sugars.

Findings from this study are consistent with reports for the US mainland, where FAFH was higher in saturated fat and sodium and lower in fiber than foods at home.² The same national report found that the nutritional composition of FAFH varied across establishment types,² consistent with the current study’s findings. Others have also found that FAFH consumption from establishment types such as fast-food and full-service restaurants is associated with higher dietary energy intake, saturated fat, sodium, lower fiber intake, and other essential nutrients.^13,21,22,51 Among major chain restaurant administrators,⁵² the driving factor influencing menu offerings is profit, and obstacles to including more nutritious ingredients include short shelf-life of produce, increased preparation time, low sales, and high labor costs. These driving factors may, in part, explain the scant presence of fruits and vegetables in their offerings.⁵²

The association between the frequency of FAFH and overall diet quality has been evaluated in few studies. Findings from the present study are consistent with one study of Hispanics and Latinos living in the United States, where increased FAFH consumption was associated with a lower AHEI score.¹⁰ FAFH has been associated with a lower intake of fruit and vegetables,⁵³ good sources of fiber, and a higher intake of total energy, saturated fat, alcohol, and added sugars.^54-56 Similar to the current study, FAFH from fast-food and full-service restaurants has been associated with a higher intake of dietary energy, saturated fat, and sodium.^13,21,22,51 Stronger and more consistent associations were identified for take-out, fast-food establishments, and food trucks than for other types of establishments, suggesting that these should be prioritized for further research and public health strategies. Unlike prior studies,^13,21,22,51 the present study did not find significant results across all relevant outcomes. Possible discrepancies may be due to the type of foods prepared and served at commercial establishments in the previously reported settings vs this study. Individuals in Puerto Rico may also have a generally poor intake of these nutrients, as shown p_revi_ously,¹⁸ making it difficult to see a difference by source. This provides a reason to study factors besides FAFH consumption that may contribute to the overall quality of diet and health among participants, including the overall food environment in Puerto Rico. The high concentration of fast-food establishments in the urban areas of Puerto Rico may further explain the poor diet quality of the overall sample.^19,57

Exploring the effects of FAFH in a noncommercial context is relevant to dietary choices–this includes food consumed at the home of relatives, friends, or neighbors. Why adults may eat meals of lower diet quality in someone else’s home warrants further research. Two possible reasons may be that friends and family offer comforting food as part of cultural norms of socialization, identity, and affection,^58,59 or that the person undergoes social facilitation (eats more when in groups),²⁹ or sociotropy (pleases others and maintains social harmony).⁶⁰ In Puerto Rico, relatives and friends take care of children either as childcare providers or occasionally after school.⁶¹ This dynamic gives both children and parents the opportunity to consume regular meals as NC-FAFH. Older adults are also cared for by neighbors and relatives,^62,63 extending the relevance of NC-FAFH across age groups. There may be other understudied contexts of FAFH associated with celebration and nurturing that may contribute unique food choices. One previous study explored the influence of NC-FAFH on a group of predominantly Hispanic and Latino children’s dietary intake, reporting an association with increased intake of sugar-sweetened beverages.³¹ Although the population studied here were adults, this analysis yielded similar results for lack of adherence to recommended levels of added sugars. In line with the present study’s findings, a cross-sectional analysis of Hispanic Community Health Study and Study of Latinos data evaluated FAFH consumption among Hispanic and Latino adults and found that higher FAFH consumption at a friend’s or relative’s home was associated with lower AHEI-2010 scores.¹⁰

Foods prepared away from home, particularly at fast-food establishments, are more likely to be fried, making them a common source of fat.⁶⁴ Furthermore, in the sampled population, it has been reported that sweets and desserts, along with dairy, make up more than 18% of the energy intake, contributing to saturated or trans fats in the diet.¹⁸ In the current study, NC-FAFH was positively associated with dietary intakes of sodium and saturated fat across all meal occasions. The results of this analysis agree in part with those of another study that reported that consuming breakfast away from home was associated with fewer servings of whole grains and dairy and higher saturated and solid fats, alcohol, and added sugar, and dinner away from home was associated with fewer servings of vegetables.⁵⁵

The results of this study have important implications for public health nutrition initiatives in Puerto Rico and similar settings. The current health profile in Puerto Rico is similar to that of other developed countries.⁶⁵ Consuming FAFH in Puerto Rico is not only a factor of the nutritional transition observed among developing economies⁶⁶; economic, cultural, and social factors make this a prevalent trend on the island. Notably, the food sector employed >8% of the Puerto Rican workforce at the time of this study, making it 1 of the top 5 employers on the island.⁶⁷ In May of 2020, the food sector still employed 7.5% of the local workforce, despite COVID-19 restrictions.⁶⁸ Policies to improve the nutrient quality of FAFH may help maintain this vital source of jobs, while improving the nutritional health of the population. However, regulating these types of establishments may not be practical under the current socioeconomic instability of the island.^69-71 A study that used a mathematical modeling approach determined that it may be feasible to develop healthy and affordable FAFH options within the Thrifty Food Plan developed by the US Department of Agriculture.⁷² Menu labeling at commercial establishments may reduce the energy content of food purchased in some contexts, although evidence is mixed.⁷³ Policies for voluntary restriction of sodium in the food supply, as well as government-sanctioned regulations,^74,75 may provide opportunities to reduce sodium intake from FAFH, although it has been reported that reducing sodium in the food supply may not alter individual consumption effectively,⁷⁶ and individual education should accompany policy measures.⁷⁵ To the authors’ knowledge, these promising policies have not been considered in the context of food establishments in Puerto Rico.

Finally, the results from this study showed that frequent NC-FAFH was associated with low diet quality, as much or even more than C-FAFH. Policies that build on our understanding of the determinants of food choices may provide viable incentives to promote diet quality in FAFH settings.^77-79 Family-based interventions where the effect of financial access to nutrient-dense foods on dietary outcomes is evaluated may be appropriate and effective. For example, coupling educational programs with targeted food-assistance vouchers has positively influenced food and vegetable preparation and consumption.^80,81 Efforts that target specific subgroups may be more effective at influencing dietary choices when incorporating cooking and food literacy skills.⁸² Promoting home-cooked meals may also be effective, primarily as these have been associated with better diet quality and lower adiposity in other populations.^83,84 This should also be accompanied by educational programs for healthy home cooking to reduce added sugars and sodium; studies on similar interventions have been promising but limited.⁸⁵ Lastly, future studies should consider longitudinal effects of FAFH behaviors on overall dietary and health outcomes.

This study has several limitations. Cross-sectional studies cannot establish causality. Therefore, the cross-sectional design of this study and the convenience sampling among primary health clinics in the metropolitan area of San Juan, Puerto Rico limit generalizability. However, the study demonstrated wide sociodemographic representation across the sampling sites.³² The validity of results may be limited by measurement error from self-report bias and recall bias^86,87 and by the lack of evaluation of the adapted FAFH questions. The present study was not powered to find associations between FAFH consumption and specific demographic or socioeconomic factors (ie, sex or income), which could further elucidate underlying mechanisms. Despite the limited sample size, the posited hypotheses were tested.

A strength of this study is that it provides an initial understanding of the potential influence of FAFH behaviors on diet quality among Puerto Ricans on the island. Food environments and dietary behaviors in Puerto Rico are understudied and research in this area is necessary to provide an evidence-based foundation to inform public policy.

CONCLUSIONS

The current study identified significant associations between FAFH consumption and unhealthy dietary intake at a granular level that included commercial vs noncommercial settings, various types of commercial establishments, and meal occasion. It is concluded that frequent consumption of FAFH is associated with lower diet quality and lower adherence to dietary recommendations of several key nutrients among adults from the San Juan metropolitan area of Puerto Rico. The results may encourage further research to evaluate whether shifting toward consumption of healthy at-home meals and establishing policies that reformulate the quality of commercial FAFH improves dietary intake in the Puerto Rico population.

RESEARCH SNAPSHOT.

Research Questions:

Is consumption of foods away from home associated with lower diet quality and less adherence to dietary recommendations? Do these associations vary by type (commercial vs noncommercial) of foods away from home?

Key Findings:

In this cross-sectional study of 239 adults living in the San Juan, Puerto Rico metropolitan area, “frequent” consumption of foods away from home (vs “rarely”) was associated with lower diet quality for commercial and noncommercial foods away from home. “Frequent” consumption of commercial foods away from home (vs “rarely”) was significantly associated with lower odds of meeting dietary fiber recommendations, and “frequent” consumption of noncommercial foods away from home was significantly associated with lower odds of meeting recommendations for sodium and added sugars.