EVALUATING NUTRITIONAL QUALITY IN MEALS PROVIDED BY
HOMELESS SHELTERS
CHAPTER I INTRODUCTION
According to the U.S. Department of Housing and Urban Development, in 2014 almost
1.49 million people in the U.S. were homeless and stayed in a shelter.1 Of those who were
sheltered, 60.6% were adults ages 25 and older.1 Types of shelters available to people who are
homeless include emergency or temporary shelters, transitional housing that typically provides
housing up to 24 months, and safe havens that provide temporary housing and services to hardto-
serve individuals.1
Compared to the general population, the homeless consume less food and have diets
lower in nutritional quality. In a study by Burt and Cohen,2 the homeless reported eating fewer
meals per day than other low-income Americans, and more than a third reported eating nothing
for one or more days in the week prior to their interview. A study conducting 24-hour recalls for
319 homeless adults in Rhode Island found over 94% of the participants were food insecure, with
64% of those who were food insecure experiencing hunger.3 The mean food intake of participants
did not meet USDA recommendations for the number of daily servings for vegetables, fruit,
dairy, and meats/beans.3
Many shelters provide food to their residents in addition to housing and other services. A
1992 national study of 1,704 homeless individuals found that the primary sources of food for
homeless adults were soup kitchens (63%) and shelters (51%).4 However, there are no existing
regulations on the food provided unless the shelter is receiving financial assistance from a state or
federal program. Even so, some assistance programs have minimal regulations. For example,
2
the federal Emergency Shelter Grants Program has only one minimal requirement for emergency
shelters.5 The requirement merely states food preparation areas must have suitable space and
storage and that equipment and food served must be safe and sanitary.5 On the state level, the
state of Illinois’s Emergency Food and Shelter Program requires shelters to serve at least one
meal per day in facilities that meet local health and safety code requirements.6 Some programs
available to shelters do, however, go beyond the minimal requirement, such as the USDA Child
and Adult Food Program (CACFP), which requires shelters to meet nutritional guidelines.7 The
majority of shelters also do not set their own internal nutrition regulations or standards. A recent
study of Boston shelters found that only 1 of the 9 shelters had any of their own internal nutrition
standards in place, and only 2 of the 9 had consulted with a nutritionist.8
For residents of homeless shelters, food served on-site is a main source of nutrition.
Residents especially rely on the meals provided by the shelters when their shelter rooms do not
have cooking or food storage facilities.9,10 There is limited research examining the nutritional
quality of the meals served at these facilities. The available research on the diet quality of young
children living in homeless shelters found the food served did not meet the nutritional
requirements for young children.11,12 A study on the dietary intake of adult women living in a
transitional shelter found the food served did not meet USDA recommendations.9 The
participating residents felt the food had little variety, poor taste, poor nutritional quality, and did
not help them manage their chronic illnesses.9 Studies on dietary intake of shelter residents often
mention financial constraints and individual shelter food policies as barriers to meeting
nutritional needs.9,10,12
3
There are many shelter polices that could possibly affect residents’ diets and the
nutritional quality of food provided. These policies regulate meal times, the availability of food
outside of standard meal times, food storage, cooking facility availability, food budget, and
whether volunteers plan the meals. Numerous shelters rely on volunteers to plan, purchase, and
prepare the meals served to residents. If volunteers can plan and prepare the meals using their
own recipes, the nutritional quality of meals could vary between volunteers.
Importance of the Study
People who are homeless suffer from more health problems than non-homeless and live
with unmet health care needs.13-15 The most prevalent health problems identified include
respiratory, dermal conditions, injuries, and digestive issues.13 Due to low diet quality, they are at
risk for developing nutrition-related disorders including malnutrition, obesity, and cardiovascular
diseases.16-18 There is a genuine need for this population to receive proper nutrition during their
stay at shelters, due to their pervasive health issues.
Nutrition professionals can be of use to this neglected population. In a survey of 259
family shelters, 87% reported willingness to integrate recommendations from medical providers
and nutritionists for meal planning.19 Professionals can aid not only in meal planning but in food
acquisition, food preparation and education of staff. Lending their expertise to shelters providing
meals would not only increase nutritional quality of the food but also variety, taste and overall
quality. There is also a void in assistance to homeless people in the management of
nutritionrelated diseases. Professionals can develop nutrition education programs in homeless
shelters as well as provide specialized menus or food options. To correctly assist shelter staff and
4
residents, nutrition professionals need to understand budget constraints, food policies and
practices of the shelter.
Purpose
The purpose of the study was to evaluate the nutritional quality of food provided by
homeless shelters and examine the effect of shelter meal policies on the nutritional quality of
meals provided. Specifically, the study aimed to determine if the nutritional quality of meals
differ between shelters that have volunteers provide meals and shelters that provide meals
themselves. It also aimed to determine if the nutritional quality of meals differed in shelters that
set nutritional standards for their meals and those that do not. The nutritional quality of meals
provided was evaluated by comparing it to the Dietary Reference Intakes (DRIs) for adult men
and women, considering relatively few studies have been done in recent years with this age
group.
Research Questions
Question 1
The first question was: How do the nutrients in meals served at homeless shelters
compare to the Dietary Reference Intakes (DRIs) for adult males and females?
The variable of interest was satisfaction of DRI criteria for each nutrient. The nutrients
measured include protein, carbohydrate, fiber, total fat, saturated fat, cholesterol, sodium, iron,
calcium, folate, and vitamins A, C, and D. Calories were also measured. Each measured nutrient
was defined as meeting the DRI if its measured amount was at least 33% of the DRI.
5
Question 2
The second question was: Is there a difference in how well meals meet Dietary Reference
Intakes (DRIs) between homeless shelters that have outside volunteers provide meals and
shelters that provide meals themselves?
The independent variable for this question was method of meal provision, and the
dependent variable was meeting Dietary Reference Intakes (DRIs). The methods of meal
provision were either volunteers or shelter staff purchased and planned a meal to be served at a
shelter.
Question 3
The third question was: Is there a difference in how well meals meet Dietary Reference
Intakes (DRIs) between homeless shelters that have nutrition standards for meals served and
shelters that do not?
The independent variable for this question was nutrition standards, and the dependent
variable was meeting Dietary Reference Intakes (DRIs). Nutrition standards were defined as any
followed rules or guidelines used by shelters that directed what food was to be served to make
meals nutritious.
Operational Definitions
1. A homeless shelter is a building or program providing temporary residence to homeless
individuals. This includes the wide-spread PADS program, in which the shelter moves
locations every night.
6
2. A shelter director is a staff member who manages the overall operation of a homeless
shelter. For this study, this may include an individual who oversees the food service
operation but not the overall operation.
3. The nutrient content of meals is determined using the Nutritionist Pro Diet Analysis
Software.20
4. The method of meal provision is how the shelter serves meals to its clients.
5. Nutrition standards refers to any followed rules or guidelines used by shelters that direct
what food is to be served to make meals nutritious.
6. Dietary Reference Intakes (DRIs) are sets of nutrient intake values recommended by the
Institute of Medicine (IOM) to meet the nutrition needs of healthy people (Appendix C).
7. A nutrient is considered to meet the DRI if it is found in amounts that are at least 33% of
the DRI listed for adult men and women. This percentage was derived from the
assumption that an individual is consuming three meals per day. This standard was used
in the 1995 USDA regulations for the National School Lunch Program, which states
meals must contain one-third of the 1989 RDA for lunch and one-fourth of the RDA for
breakfast.21
CHAPTER II
METHODOLGY
Study Design
The purpose of this study was to evaluate the nutritional quality of food provided by
homeless shelters and examine the effect of shelter meal policies on the nutritional quality of
meals provided. For the first research question, the variable of interest was satisfaction of DRI
criteria for each nutrient. For the second research question, the independent variable was method
of meal provision, and the dependent variable was meeting Dietary Reference Intakes (DRIs).
The methods of meal provision were either volunteers or shelter staff purchased and planned a
meal to be served at a shelter. For the third research question, the independent variable was
nutrition standards, and the dependent variable was meeting Dietary Reference Intakes (DRIs).
This study was initiated as a correlational research design, with a purposive sampling
method, but ended up being a mixed-methods study with a qualitative portion. Data were
originally collected as planned through an online survey and self-reporting. However, due to a
low response rate, there were not enough participants to complete the statistical analysis required
to answer the second and third research questions. The qualitative assessment was subsequently
added to help enrich the study results.
Participants
Participating shelters were selected through a purposive sampling method. To be eligible
shelters must have served at least one dinner per week on-site and have adult residents. Targeted
8
shelters were also all located in the Midwest. All shelters found to meet these criteria were
recruited, and recruitment began after the Northern Illinois University’s Institutional Review
Board (IRB) approved the study. The recruitment letter was sent by mail or email to the directors
or staff members in charge of food service operations at the shelters (see Appendix D). The
recruitment letter included a link to an online survey through Qualtrics (see Appendix B). The
first page of the survey included a letter of informed consent. Directors gave their consent by
completing the online survey. It was expected that the study would have a 30% response rate,
with at least 50 shelters participating in the study after expecting to target 170 shelters. A
minimum of 30 shelter participants was needed, including at least 15 with volunteer-provided
meals and at least 15 with shelter staff-provided meals. These 30 shelter participants also needed
to include at least 15 with meals from shelters with nutrition standards and at least 15 with meals
from shelters with no nutrition standards. A shelter’s method of meal provision and nutrition
standards overlap when categorizing meals to meet these minimums. These minimum numbers of
participants were calculated to ensure that a large-enough sample size would be available to
produce reliable results when using one-way MANOVA to answer the second and third research
questions.
9
I
nstruments
Online Survey
Participants were given a link to an online survey on the Qualtrics website. The survey
questions asked participants for basic information on the shelters, including shelter type, number
of beds, number of people served for each meal, and if the shelters had male and/or female
clients. To confirm eligibility for the study, participants were asked if their clients were 18 years
or older and the meal types served at the shelter. Questions 9 and 10 measured the independent
variables for the second and third research questions (method of meal preparations and use of
nutrition standards, respectively). Questions 11 and 12 obtained qualitative data by asking the
participants to describe their guidelines regarding food or explain why they do not have
guidelines. Participants were asked to describe their guidelines to see if their guidelines include
nutrition standards. This data was also used to identify common nutrition standards among
shelters. If a shelter did not have guidelines regarding food, they were asked to explain why in
order to identify common reasons shelters do not have nutrition standards. Questions 16-22
collected demographic information on the participant filling out the online survey. The full list of
questions from the online survey can be found in Appendix B.
Data Preparation
After completion of the online survey, shelter directors were notified of which dates to
send dinner information to the researcher for analysis. Dinner was chosen because it is the most
consistently served meal across shelters, and it is most likely to be a complete meal. For each
participating shelter, the nutrient content of three dinners was assessed individually. The dinner
menu, all available recipes, and portion sizes were requested to be sent electronically or by mail
in a stamped and addressed return envelope provided to them. If recipes or portion sizes were not
10
available or provided, applicable recipes and portion sizes were determined by the researcher.
These were determined using the Nutritionist Pro Diet Analysis Software,20 which has recipes
and standard portion sizes in its database. The nutrients of the three dinners were then averaged
to represent the average nutrient content of one dinner served at each shelter. The nutrients
analyzed included protein, carbohydrate, fiber, total fat, saturated fat, cholesterol, sodium, iron,
calcium, folate, vitamin A, vitamin C, and vitamin D. Calories was also calculated. These
nutrients were chosen because they are either nutrients that have been overrepresented in shelter
meals, such as fat,22 or nutrients known to be deficient in the socioeconomically disadvantaged
populations such as iron.23 The Nutritionist Pro Diet Analysis Software20 was used to determine
the nutrient content of each meal. Additional questions for qualitative data collection were sent
via e-mail to participants who reported dinner information.
Additional Qualitative Questions
Due to a low response rate, additional questions were asked of participants to provide
more data. The following additional questions were asked:
1. Please list any barriers to following your listed guidelines regarding food when
planning meals.
2. Would you describe your shelter as having adequate staff for meal planning?
3. Would you describe your shelter as having adequate staff for cooking?
4. Would you describe your shelter as having adequate cooking space for providing
meals to your guests?
5. Please list any ways you believe a registered dietitian could help your shelter with
providing meals to your guests.
CHAPTER III
RESULTS
Characteristics of the Participating Homeless Shelters
Recruitment letters were sent to 195 homeless shelters who met eligibility requirements
in the Midwest, and 21 shelters agreed to participate. This represented a 10.77% response rate.
Of those who participated, all served clients over the age of 18, which was an eligibility
requirement. One shelter had only male clients and two shelters had only female clients, with the
remaining having both male and female clients (85.71%). For type of shelter, 11 (52.38%) were
emergency/temporary/overnight, 4 (19.05%) were transitional housing and 6 (28.57%) were
both. The number of beds available in the shelters ranged from 21 to 400 beds, with an average
of 105 ± 89.62 beds. The average number of people served in a meal ranged from 18 to 250
people, with an average across all participating shelters of 91.24 ± 61.73. Out of the 21 shelters
that participated, 20 served breakfast, 20 served lunch, 21 served dinner, and 11 served snacks to
their clients. This meant that all participating shelters served three meals per day, except for one
shelter that served only dinner. The breakfast times ranged from 5am-11am, lunch times from
11am-2:30pm (with some offering bag lunches in the morning), and dinner times ranged from
4pm-9pm. Snack times varied. These characteristics of the homeless shelters that participated in
the study are reported in Table 1.
Table 1
Characteristics of Homeless Shelters (N=21)
Variable Frequency (Percent)
Client Gender
Male Only
1 (4.76%)
Female Only
Both
2 (9.52%)
18 (85.71%)
Housing Type
Emergency/Temporary/Overnight
11 (52.38%)
12
Transitional Housing 4 (19.05%)
Both 6 (28.75%)
Mean Beds (±SD) 105 ± 89.62
Mean Clients Served at Meals (±SD) 91.24 ± 61.73
Meal Types Served
Breakfast
20
Lunch 20
Dinner 21
Snacks 11
Meal Time Ranges
Breakfast
5am-11am
Lunch 11am-2:30pma
Dinner 4pm-9pm
Snacks Varied
aSome offered bag lunches in the mornings
Characteristics of Participating Shelter Directors
A shelter director is a staff member who manages the overall operation of a homeless
shelter. For the purpose of this study, the shelter director included individuals who oversaw the
food service operation, but not the overall operation. Staff members designated as in charge of
the food service operation were better able to provide dinner information than a shelter director
who oversaw an entire shelter operation. However, not every shelter had this designated staff, or
their contact information was not as readily available. Therefore, shelter directors were contacted
if a food service director was not identified. There were 9 (42.86%) male and 12 (57.14%)
13
female participating shelter directors. Ages ranged from 24 to 69 years old, with a mean age of
45.05 ± 13.5 years. The education level of the directors ranged from high school diploma to a
doctorate degree, with the majority (38.1%) holding an undergraduate degree. Only 2 out of the
21 (9.52%) completed a culinary degree. As for employment at the homeless shelter, 18
(85.71%) were full-time employees. Fourteen of the directors (66.67%) were involved with
planning meals. The characteristics of shelter directors who participated in the study are reported
in Table 2. The shelter directors were also asked to list any nutrition education they had received.
Nine reported to having completed at least one nutrition course, including one participant who
was a registered dietitian and one who received their Master in Public Health degree.
Table 2
14
Characteristics of Shelter Directors (N=21)
Variable Frequency (Percent)
Gender
Male
9 (42.86%)
Female 12 (57.14%)
Age (±SD) 45.05 ± 13.5
Education Level
High School Diploma
3 (14.29%)
Some College 5 (23.81%)
Associate Degree 2 (9.52%)
Undergraduate Degree 8 (38.1%)
Master’s Degree 2 (9.52%)
Doctorate Degree 1 (4.76%)
Culinary Degree
Yes
2 (9.52%)
No 19 (90.48%)
Employment
Full-Time
18 (85.71%)
Part-Time 2 (9.52%)
Other 1 (4.76%)
Plans Meals
Yes
14 (66.67%)
No 7 (33.33%)
Nutrient Adequacy/Quality of Dinners Meals
To answer the first question: “How do the nutrients in meals served at homeless shelters
compare to the Dietary Reference Intakes (DRIs) for adult males and females,” the dinner menus
provided were analyzed for nutrient content and compared to the DRI. Out of the 21 shelters that
agreed to participate in the study, 11 (52.38%) sent dinner information, with a total of 31 dinners
reported. Ten out of the eleven shelters reported information for three dinners while one shelter
reported information for only one dinner. The nutrients of each dinner were analyzed, then
averaged across meals to represent the average nutrient content of one meal served at each
15
shelter. Then those representative meals were averaged to find the overall mean nutrients in a
shelter meal. If the overall mean nutrient was less than 1/3 of the Dietary Reference Intakes
(DRI), then it was categorized as nutritionally inadequate. Overall mean nutrients that did not
meet 1/3 of the DRI were calories for men only, carbohydrates, total fiber, calcium, and vitamin
D for both men and women. The average level of sodium was over twice the recommended
amount. The nutrient analysis of these dinners is shown in Table 3. The most frequent foods
served for these dinners are shown in Table 4. The foods are listed by food group and are listed
from most frequently served to least frequently served.
Table 3
Overall Nutrition Quality of Homeless Shelter Dinners (N=31*)
Males 19-50 Females 19-50
16
Nutrient Mean ± SD 1/3 of DRI Percentage 1/3 of DRI Percentage of 1/3
DRI of 1/3 DRI
*A total of 31 dinners were reported by 11 homeless shelters.
Table 4
Most Frequently Served Foods by Food Group
Food Group Frequent Foods Served
Dairy Milk, Cheese (cheese slices, shredded cheese, cheese sauce)
Grains
Bread (bread slices, dinner rolls, garlic bread, buns, bread sticks,
baguette, submarine), Noodles, White Rice, Wheat Tortilla,
Tortilla Chips, Cornbread
Fruit, Fruit Juices
Fresh Fruit Salad, Orange Juice, Canned Fruit Cocktail,
Watermelon, Canned Peaches, Canned Pears, Mandarin Oranges,
Strawberries, Grapes, Cantaloupe, Apple Juice, Grape Juice,
Cinnamon Applesauce, Pineapple Tidbits
Calories 767.68 ± 266 866.67 88.58% 666.67 115.15%
Protein (g) 35.17 ± 11.56 18.67 188.38% 15.33 229.42%
Carbohydrates (%
of calories)
43.63 ± 8.9 45-65% of
calories
67.12-
96.96%
45-65% of
calories
67.12-
96.96%
Fat (% of
calories)
38.22 ± 8.7 20-35% of
calories
109.2-
191.1%
20-35% of
calories
109.2-
191.1%
Saturated Fat (g) 11.2 ± 5.77 0 - 0 -
Cholesterol (mg) 127 ± 81.59 0 - 0 -
Total Fiber (g) 7.32 ± 3.04 12.67 57.78% 8.33 87.88%
Sodium (mg) 1086.95 ±
394.57
500 217.39% 500 217.39%
Iron (mg) 7.1 ± 7.87 2.67 265.92% 6 118.33%
Calcium (mg) 269.42 ± 188.72 333.33 80.83% 333.33 80.83%
Folate (µg) 167.24 ± 68.97 133.33 125.43% 133.33 125.43%
Vitamin A (RAE) 423.88 ± 255.82 300 141.29% 233.33 181.67%
Vitamin C (mg) 37.84 ± 22.25 30 126.13% 25 151.36%
Vitamin D (µg) 2.18 ± 3.4 5 43.6% 5 43.6%
17
Vegetables Potatoes (mashed, baked, fried, potato chips, tater tots), Mixed
Salad Greens, Green Beans, Carrots, Canned Corn, Canned
Mixed Vegetables, Canned Peas, Onion, Broccoli, Cole Slaw,
Marinara Sauce, Salsa, Asparagus, Green Onion, Pickles
Protein
Ground Beef (hamburger patty, meatloaf, meatballs, meat sauce),
Baked Chicken, Beans (baked, refried, black), Ham, Bacon,
Lunch Meat, Hot Dogs, Fish Sticks, Tuna, Ribs, Turkey Roast
Slices
Fats, Oils Salad Dressing (Ranch, Italian), Butter, Margarine, Mayonnaise,
Gravy, Creamer, Vegetable Oil, Sour Cream
Sweets, Baked Goods
Cookies, Pie, Brownies, Cake, Cupcake, Jell-O
Sweetened Beverages
Lemonade, Sprite, Coke
Combined Foods
Pizza, Lasagna, Taco Pie, White Chicken Chili, Chicken and Rice
Casserole, Potato Soup, Cream of Mushroom Soup, Quiche
Method of Meal Provision on Nutrition Quality of Meals
The second question: “Is there a difference in how well meals meet Dietary Reference
Intakes (DRIs) between homeless shelters that have outside volunteers provide meals and
shelters that provide meals themselves,” was answered by comparing the average nutrient
content of meals prepared by volunteers to those prepared by shelter staff. Of the participating
shelters, 18 had staff-planned meals and 3 had volunteer-planned meals. Of those that provided
dinner information, nine shelters had staff-planned meals and two had volunteer-planned meals.
18
Eight out of the nine shelters with staff-planned meals sent information for all three dinners, and
the remaining shelter sent information for one dinner. This makes a total of 25 staff-planned
dinners reported. The two shelters with volunteer-planned meals sent information for all three
dinners, making a total of six volunteer-planned dinners reported. Table 5 shows the results of the
nutrient analysis for staff-planned dinners and Table 6 reports results for volunteer-planned
dinners. For staff-planned dinners, the mean nutrients that did not meet 1/3 of the DRI were
calories for men only, carbohydrates, total fiber, iron for women, calcium, and vitamin D for both
genders. Sodium was over twice the recommended amount for both genders (Table 5). For
volunteer-planned dinners, mean nutrients that did not meet 1/3 of the DRI were calories and
total fiber for men only, calcium and vitamin D for both genders. Sodium was also over twice the
recommended amount for both genders (Table 6).
Table 5
Nutrition Quality of Staff-Planned Dinners (n=25*)
Nutrient
Calories
Staff-Planned
Mean ± SD
95% Confidence
Interval
1/3 of DRI
for Males
age 19-50
1/3 DRI
for Females
age 19-50
760.8 ± 296.29 (638.49, 883.11) 866.67 666.67
Protein (g) 36.59 ±12.43 (31.46, 41.73) 18.67 15.33
Carbohydrates (% of
calories)
40.97 ± 9.84 (21.68, 60.26) 45-65% of
calories
45-65% of
calories
Fat (% of calories) 39.57 ± 9.78 (20.4, 58.74) 20-35% of
calories
20-35% of
calories
Saturated Fat (g) 11.38 ± 6.37 (8.75, 14.01) 0 0
19
Cholesterol (mg) 139.67 ± 85.34 (104.44, 174.9) 0 0
Total Fiber (g) 6.79 ± 2.53 (5.75, 7.83) 12.67 8.33
Sodium (mg) 1080.28 ± 438.39 (899.31, 1261.24) 500 500
Iron (mg) 4.84 ± 1.57 (4.19, 5.48) 2.67 6
Calcium (mg) 291.39 ± 185.98 (214.62, 368.16) 333.33 333.33
Folate (µg) 172.27 ± 75.89 (140.94, 203.6) 133.33 133.33
Vitamin A (RAE) 422.33 ± 221.15 (331.03, 513.62) 300 233.33
Vitamin C (mg) 33.63 ± 18.31 (26.07, 41.19) 30 25
Vitamin D (µg) 2.39 ± 3.73 (0.85, 3.93) 5 5
*Twenty-five dinners reported were staff planned.
Table 6
Nutrition Quality of Volunteer-Planned Dinners (n=6*)
Nutrient
Calories
Volunteer-Planned
Mean ± SD
95% Confidence
Interval
1/3 of DRI
for Males
age 19-50
1/3 DRI
for Females
age 19-50
798.63 ± 53.69 (742.28, 854.99) 866.67 666.67
Protein (g) 28.77 ± 0.52 (28.22, 29.31) 18.67 15.33
Carbohydrates (%
of calories)
55.02 ± 20.31 (15.21, 94.83) 45-65% of
calories
45-65% of
calories
Fat (% of calories) 32.4 ± 19.11 (-5.06, 69.86) 20-35% of 20-35% of
20
calories calories
Saturated Fat (g) 10.37 ± 2.45 (7.79, 12.94) 0 0
Cholesterol (mg) 69.98 ± 18.88 (50.17, 89.8) 0 0
Total Fiber (g) 9.72 ± 5.26 (4.2, 15.23) 12.67 8.33
Sodium (mg) 1116.97 ± 131.14 (979.32, 1254.62) 500 500
Iron (mg) 17.27 ± 18.62 (-2.28, 36.81) 2.67 6
Calcium (mg) 170.58 ± 235.77 (-76.89, 418.05) 333.33 333.33
Folate (µg) 144.61 ± 15.43 (128.41, 160.81) 133.33 133.33
Vitamin A (RAE) 430.85 ± 512.91 (-107.5, 969.2) 300 233.33
Vitamin C (mg) 56.78 ± 37.26 (17.67, 95.9) 30 25
Vitamin D (µg) 1.22 ± 1.53 (-0.39, 2.82) 5 5
*Six dinners reported were volunteer planned.
Due to the small sample size, a statistical analysis could not be performed to determine if
there was a significant relationship between method of meal provision and nutrition quality of
meals. Therefore, the researcher asked participants additional open-ended questions to further
explore shelter staff-planned meals, presented later.
Shelter’s Use of Nutrition Standards and Nutrition Quality of Meals
The third question: “Is there a difference in how well meals meet Dietary Reference
Intakes (DRIs) between homeless shelters that have nutrition standards for meals served and
shelters that do not,” was answered by examining the observed sample differences in the
nutrition quality of meals prepared using standards and those prepared without any nutrition
standards. Results of the use of nutrition standards showed that 14 (66.67%) had nutrition
standards and 7 (33.34%) had no nutrition standards. Table 7 shows the nutrition standards used
by the participating shelters. The nutrition standards consisted of food groups that must be
represented at each meal served, with some requiring specific amounts from each food group. In
most shelters (52.38%), the kitchen manager, food service manager, or lead cook was responsible
21
for ensuring the nutrition standards were being followed. Other strategies used by shelters to
ensure the standards were being followed included training the staff responsible for cooking the
food and requiring management to check the menus. Of the participants who provided dinner
information, eight (72.73%) had nutrition standards and three (27.27%) did not have any
nutrition standards. All eight of the shelters with nutrition standards sent information for three
dinners, making a total of 24 dinners planned with nutrition standards reported. Two out of the
three shelters without nutrition standards sent information for three dinners, and the remaining
shelter sent information for one dinner. This makes a total of seven dinners planned without
nutrition standards reported. Table 8 shows the results of the nutrient analysis of dinners with
nutrition standards and Table 9 shows the results of those prepared without nutrition standards.
For shelters where nutrition standards were used, the nutrients that did not meet 1/3 of the DRI
were calories for men only, iron for women only, and carbohydrates, total fiber, and vitamin D
for both genders. The average sodium content was over twice the recommended amount (Table
8). For shelters that prepared meals without nutrition standards, the nutrients that did not meet
1/3 of the DRI were calories, total fiber, calcium, folate, vitamin A, vitamin C, and vitamin D for
both genders.
Table 7
Listed Shelter Nutrition Standards for Dinner Planning
Nutrition Standards
Must include: Protein, Starch, Vegetable, Salad, Fresh Fruit, Bread, Dessert
Must include: Protein, Vegetable, Salad, Bread, Dessert
22
Must include: Protein, Grain or Starch, Vegetable, Fruit, Bread. If using pork, there must be
another type of protein also.
Must include: 2-3oz Protein, 1 cup Starch, 1 cup Vegetables, 1 cup Fruit, 2 slices of
Bread or 1 cup Pasta, 1 serving Dessert. Approximately 1 cup Milk, 1 cup Juice, 1 cup
Coffee with sugar
Must include: 4-6oz Protein, 2 Vegetables with at least 1 Leafy Green Vegetable, Fruit and/or
Fruit Juice, no more than 1 Starch
We are part of the Child and Adult Care Food Program just like the schools are. Dinner must
include: 2oz Meat/Meat Alternative, ½ cup Vegetable, ½ cup Fruit, 2oz equivalents Grains.
At least one serving per day must be whole grain-rich.
Follow Child and Adult Care Food Program guidelines
Provide a meat protein, veggie, salads, starch, beverage, dessert, and optionally bread.
Incorporate any miscellaneous donations that are available. Use proper serving sizes,
promote healthy choices if possible. No alcoholic beverages are allowed on the premises.
Follow basic nutritional guidelines using the food groups
Serve a balanced meal: Protein, Vegetable, Starch, Grain, Limited sugar (dessert items)
Try to always serve a balanced meal: Protein, Vegetables, Fruit
Try to include: Protein, Starch, Non-Starchy Vegetable
Try to include: Protein, Starch, Vegetable, Fruit (canned or fresh)
Try to serve high calorie, depending on food available
Table 8
Nutrition Quality of Dinners Prepared with Nutrition Standards (n=24*)
23
Nutrient
Calories
Nutrition
Standards
Mean ± SD
95% Confidence
Interval
1/3 of DRI for
Males age
19-50
1/3 DRI for
Females
age 19-50
834.38 ± 274.8 (718.32, 950.43) 866.67 666.67
Protein (g) 36.84 ± 13.33 (31.21, 42.47) 18.67 15.33
Carbohydrates (%
of calories)
42.8 ± 10.1 (23, 62.6) 45-65% of
calories
45-65% of
calories
Fat (% of calories) 39.62 ± 9.98 (20.06, 59.18) 20-35% of
calories
20-35% of
calories
Saturated Fat (g) 12.58 ± 6.21 (9.96, 15.2) 0 0
Cholesterol (mg) 143.17 ± 90.01 (105.16, 181.19) 0 0
Total Fiber (g) 7.05 ± 1.81 (6.28, 7.81) 12.67 8.33
Sodium (mg) 1216.5 ± 367.17 (1061.44,
1371.57)
500 500
Iron (mg) 5.02 ± 1.35 (4.45, 5.59) 2.67 6
Calcium (mg) 344.69 ± 161.98 (276.28, 413.1) 333.33 333.33
Folate (µg) 182.04 ± 60.17 (156.62, 207.45) 133.33 133.33
Vitamin A (RAE) 512.01 ± 228.49 (415.51, 608.51) 300 233.33
Vitamin C (mg) 43.01 ± 21.71 (33.84, 52.18) 30 25
Vitamin D (µg) 2.92 ± 3.76 (1.33, 4.51) 5 5
*Twenty-four dinners reported were prepared with nutrition standards.
Table 9
24
Nutrition Quality of Dinners Prepared with No Nutrition Standards (n=7*)
Nutrient No Nutrition 95% Confidence 1/3 of DRI for 1/3 DRI for
Standards Interval Males Females Mean ± SD age 19-50 age 19-50
*Seven dinners reported were prepared with no nutrition standards.
Calories 589.8 ± 155.62 (445.87, 733.73) 866.67 666.67
Protein (g) 30.71 ± 2.32 (28.57, 32.85) 18.67 15.33
Carbohydrates (%
of calories)
46.79 ± 18.86 (9.82, 83.76) 45-65% of
calories
45-65% of
calories
Fat (% of calories) 32.95 ± 17.77 (-1.88, 67.78) 20-35% of
calories
20-35% of
calories
Saturated Fat (g) 7.51 ± 1.83 (5.82, 9.2) 0 0
Cholesterol (mg) 83.89 ± 33.08 (53.29, 114.49) 0 0
Total Fiber (g) 8.06 ± 5.81 (2.69, 13.43) 12.67 8.33
Sodium (mg) 741.47 ± 245.81 (514.12, 968.81) 500 500
Iron (mg) 12.64 ± 15.48 (-1.68, 26.96) 2.67 6
Calcium (mg) 68.71 ± 56.23 (16.7, 120.72) 333.33 333.33
Folate (µg) 127.8 ± 88.9 (45.58, 210.02) 133.33 133.33
Vitamin A (RAE) 188.86 ± 174.87 (27.12, 350.6) 300 233.33
Vitamin C (mg) 24.03 ± 20.79 (4.81, 43.26) 30 25
Vitamin D (µg) 0.18 ± 0.11 (0.08, 0.28) 5 5
25
Results of the Qualitative Assessment
Due to the small sample size, a statistical analysis could not be executed to determine if
there were any significant relationships between the shelter meal policies and nutritional content
of meals. The eleven homeless shelters that provided dinner information were contacted
electronically and asked five additional questions to help get more insight into the researcher’s
questions. Nine out of the eleven responded to these questions (81.82%). Table 10 below shows
the qualitative assessment questions and the themes identified from the participant’s responses.
These questions were asked in order to further explore potential barriers to providing a
nutritious meal, such as lack of staff or cooking space. The last question was asked to learn how
dietitians could benefit shelters because previous research has suggested dietitians could be of
use.7,8,16
Table 10
Qualitative Assessment and Identified Themes
Qualitative Assessment Shelter Group Asked
(Sample Size)
Themes
Please list any barriers to
following your listed guidelines
regarding food when planning
meals.
Meals with Nutrition
Standards (n=7)
1.
2.
3.
Use of food donations
Lack of resources
Special dietary needs of clients
Would you describe your shelter as
having adequate staff for meal
planning?
Staff-Planned Meals
(n=7)
1. Meal planning involves
prioritizing foods that need to
be used first
26
Would you describe your shelter as
having adequate staff for cooking?
Staff-Planned Meals
(n=7)
1. Have at least one cook, but rely
on volunteers to help with
cooking and serving of meals
Would you describe your shelter as
having adequate cooking space for
providing meals to your guests?
All Shelters (n=8) 1. Adequate cooking space, fully
functioning kitchens
Please list any ways you believe a
registered dietitian could help
your shelter with providing meals
to your guests.
All Shelters (n=9) 1.
2.
Provide nutrition education and
counseling to shelter clients
Aid in meal planning to
improve meal nutrition
3. Identify food to serve clients
with specific dietary needs
Barriers to Following Nutrition Standards
To further explore shelter nutrition standards, participants were asked to list any barriers
they had to following their nutrition standards. The common themes identified from the
participants’ responses (n=7) included the use of food donations, lack of resources, and the
special dietary needs of clients. The most prevalent barrier described by shelter directors was the
use of food donations. Shelters do not get to choose the nutritional value of the food donated.
Due to limited funds, and relying mainly on donations, we do not have the luxury of
choosing our ingredients… A high degree of what we get is heavy in carbohydrates and
sugar… We never get fresh fruit! The fruit we get is expired or old.
Participants also described how they had to use donated items first because the foods donated
were often close to expiring or expiration date. Using these donated items first often changed the
planned meal that had taken into consideration the nutrition standards.
27
My biggest barrier is I never know what’s going to be donated and when the expiration
date is. From seven pallets of fries to 160lbs of chicken. Fresh product is even harder to
count on. Potato salad that needs to be eaten today, bananas that have at most two days
of shelf life, or day-old bagels that have four days before they are not serviceable.
Planning is usually not the issue. The issue comes in when we have donations that we
must use quickly. At that point, we have to serve those items.
Another identified theme, the use of volunteer provided meals, was a barrier because
shelters could only encourage, not force, volunteers to provide healthy meals.
We offer "soft" guidelines; I encourage volunteers to consider low-sodium foods and to
incorporate fresh fruits and vegetables. We are working on creating a meal of healthy
recipes. However, we cannot force volunteers to purchase healthy foods, as they are
volunteers. Our shelter does not have the capacity to purchase healthy food every night,
so we need volunteer-provided food.
Since many are untrained, they desire to give and give as much as they can, and some
give more than others, unfortunately resulting in overeating and discouraging eating
fruits and vegetables. We need somebody to stand by the serving line and enforce basic
rules, since even our volunteers ignore them.
Other themes included the lack of resources to make a nutritious meal, such as funding
and staffing (n=2), and the special dietary needs of clients altering meal plans, such as food
allergies, needing low sodium for heart disease, or not eating pork for religious reasons (n=2).
A major barrier that is common are those with religious beliefs that do not allow them to
eat specific kinds of meat such as pork. However, our shelter and program have a series
of backup plans put in place so that we are able to still provide a hot meal to those in
need…We also have concerns with those who need special diets for medical or health
concerns. We do our best to provide for those who may have high blood pressure or
diabetes. As a whole we try to limit adding salt and sugar when possible. We cater to
those who have food allergies by trying to accommodate through the use of substitution
from our premade backups.
If using pork for the protein, I make a different protein choice as some do not eat pork for
religious reasons.
28
Staff and Cooking Space
The second question asked participants if they thought their shelters had adequate staff
for meal planning. Four out of the seven participants who responded to this question felt their
shelter had adequate staff. The most common theme found from responses was prioritizing foods
that needed to be used first. Participants described that this is a main consideration when
planning meals.
Staffing the kitchen takes precedence to meal planning, and for some time, meal planning
has been simplified and more like a triage of what needs to be used first and what simple
meals to hand off to volunteer groups.
Yes [we have adequate staff]. I would, but if I miss a day I miss seeing what came in for
donations that day. I might find them two days later than the use date.
Yes [we have adequate staff], we prepare the week’s menu on Monday and will only
make adjustments based on large donations which have a small window to use the
item(s).
Participants explained that when meal planning, foods expiring soon were used first and
at the same time they had to be mindful of the donations received daily. Menus were altered to
use the newly received large donations and donations that would expire soon. The third question
asked participants if they thought their shelter had adequate staff for cooking. For this question,
only two out of seven felt their shelter had adequate staff. The majority stated they had at least
one cook but relied on volunteers to help with cooking and serving of meals.
No. I have lots of product I can’t get to in my hours at work. And definitely need help. I
think I could offer more raw foods if I had more hands.
Yes, I have a lunch cook and dinner cook for each day of the week, plus we have kitchen
volunteers through the week who are here to help with prep.
No. Staffing is always an issue here. Currently, I am short two full-time and one part-time
cook. Finding people who are willing to work the hours of a cook is hard. Due to lack of
29
cooking staff, I have to pick up some shifts, which cuts into my ability to do my job as
Food Service Administrator planning the meals.
The fourth question asked participants if they thought their shelter had adequate cooking
space for providing meals to their clients. Six out of the eight participants who responded to this
question felt they had adequate cooking space, describing their kitchens as fully functioning.
My kitchen was an afterthought, and with servicing about 400 daily, I need more kitchen
space to be a production kitchen.
Yes, we have three 6-8 feet prep tables, two convection ovens, a Rational oven which
allows for cooking either 100% steam, 100% convection or a combination, a 20-gallon
kettle and a two-burner stove top.
We have a fully functioning industrial-sized kitchen, with the use of both convection and
conventional ovens, a braising pan, industrial gas stove, multiple prep areas, multiple
sinks, and two walk in freezers, a walk-in cooler, and eight industrial reach-in coolers.
We have plenty of space to have at least eight people prepping with adequate space.
How Participants Think Registered Dietitians Could Help with Homeless Shelter Meals
Participating shelter directors (n=9) were asked to describe any ways in which they believed a
registered dietitian could help their shelters with providing nutritious meals to their shelter
clients. There were three themes identified from the directors’ responses regarding how a
registered dietitian could help homeless shelters: 1) provide nutrition education and counseling to
shelter clients, 2) aid in meal planning to improve meal nutrition, and 3) identify food to serve
clients with specific dietary needs.
Many shelter directors (n=5) identified nutrition education as a valuable service a
registered dietitian could provide to shelter clients. As they described, homeless shelter clients
could use nutrition education to improve their eating habits while at the shelter, but more
importantly once they leave the shelter and are obtaining food on their own. The clients also need
help finding affordable food and resources as well as learning how to eat healthy on a budget.
30
A registered dietitian would be helpful in teaching classes on nutritional cooking to the
clients and teach them how to make and cook healthy choices when they leave the
shelter.
I believe having a registered dietician could benefit us in providing various healthy food
options as well as educating our guests in this field.
Shelter directors recognized that a registered dietitian involved in meal planning would
improve nutrition content of the meals served. They believed a dietitian could also help improve
meal variety and creativity. Specific tasks directors mentioned they would like a dietitian to do
were to reduce sodium, make better use of the foods the shelter has, and provide large-scale
healthy recipes.
We would be able to more thoughtfully prepare meals that would allow us to balance all
the nutritional needs of children and adults. It would allow us to prevent repetition of
meals, which sometimes happens here randomly—we might see pizza here three times in
a week and then not again for months!
We work with a variety of people; it can be difficult to cater to every need seeing as how
needs are different for each and every person. The best way a dietitian could assist us is
by providing more knowledge on healthy recipes that are able to be made on a large
scale.
Some homeless shelter clients have special dietary needs due to various health
conditions. Often, these special requirements go unmet because the homeless shelter is
unprepared or unable to accommodate them. Shelter directors identified that registered dietitians,
with their specific knowledge in the area, could solve this problem. One director stated the
following:
Currently, I am acting as the dietitian. Whenever the nurse has a client who is looking for
a special diet it falls to me to determine what the appropriate diet should be based on my
culinary training and research.
Meals would likely be more nutritious, more creative and make better use of the food that
we have on hand. Additionally, there are individuals with special food needs - diabetics,
31
heart disease, gastrointestinal disorders, allergies - that go unmet in most cases as there is
no one to help plan meals for them.
CHAPTER IV
DISCUSSION
Nutrient Analysis of Homeless Shelter Dinners
The first question asked was: “How do the nutrients in meals served at homeless shelters
compare to the Dietary Reference Intakes (DRIs) for adult males and females,” and from the
nutrient analysis of participating shelter dinners, it was found that the mean nutrients that did not
meet 1/3 of the DRI were calories for men only, carbohydrates, total fiber, calcium, and vitamin
D for both genders. This trend was followed throughout the different categories of shelter meals
analyzed (staff-planned meals, volunteer-planned meals, meals with nutrition standards, meals
without nutrition standards), with slight variations. Staff-planned meals showed a similar trend,
except, in addition to the other inadequacies, they did not meet 1/3 of the iron DRI for women.
Volunteer-planned meals met the carbohydrate requirements and the total fiber requirements for
women. Meals with nutrition standards met the calcium requirements but, like the staff-planned
meals, did not meet 1/3 of the iron DRI for women. Meals without nutrition standards were the
most varied from the average meal across all shelters. The mean nutrients that did not meet 1/3 of
the DRI for meals without nutrition standards were calories, total fiber, calcium, vitamin D,
folate, vitamin A, and vitamin C for both genders. Dinners without nutrition standards only met
20.61% of calcium and 3.6% of vitamin D requirements, compared to the mean across all
shelters which met 80.83% of calcium and 43.6% of vitamin D requirements. The mean sodium
level was twice the recommended amount for all shelter meals, staff-planned meals,
volunteerplanned meals, and meals where nutrition standards were used. Total fat, saturated fat,
and cholesterol were also over the recommended amounts. The nutritional inadequacies in these
meals can be attributed to the foods available to shelters for cooking. In addition, shelters ranged
33
in size, with varying levels of resources and available staff to produce nutritious meals. The lack
of nutritional guidelines for meal planning can also contribute to nutritional inadequacies, as seen
in the nutrient analysis. The low amounts of calcium and vitamin D can be explained by the
small amounts of dairy served in meals. Many meals did not contain foods from the dairy group,
and the most some contained were small servings such as a cheese slice in a sandwich. Storage
availability was not explored in this study, however; a possible explanation could be that shelters
do not have adequate storage for perishable products such as dairy.
These findings are like other available research on the nutritional content of homeless
shelter meals and the homeless population’s diet. Research on the nutritional quality of food
served has found it to be high in fat and low in whole grains, fresh fruits and vegetables.9,11,12 In
their study focusing on the homeless adult population, Silliman and Wood22 analyzed the
nutritional adequacy of meals served in rural shelters or soup kitchens in northern California and
found that the meals were low in fiber, high in fat, and provided less than 33% of the RDA/DRI
for folate, calcium, magnesium, and zinc.22 The current study did not analyze for magnesium or
zinc, but results were similar in that shelter meals were low in fiber and calcium and high in fat.
In addition to nutrition quality of homeless shelters, the research analyzing homeless individuals’
diets also found that they are inadequate in vitamins and minerals and are high in sodium,
saturated fat, and cholesterol.17,23-26 A study in New York city that collected 24-hour recalls with
41 homeless individuals found the participants’ diets were low in zinc, vitamin B6, vitamin A,
and calcium.24 Another study in Hawaii that used a similar 24-hour recall method to analyze
diets of 57 homeless families found that their diets were also low in fiber, calcium, and vitamin
A.25 Similarly, an assessment of the nutritional status of 55 homeless urban adults at a New York
34
medical center found participants’ diets were inadequate in most nutrients but high in sodium,
saturated fat, and cholesterol.23 Last, a 1992 study of homeless women and children reported
diets less than 50% of the 1989 Recommended Dietary Allowances (RDA) for iron, magnesium,
zinc, folic acid, and calcium.26 Overall, results from the study and previous research are in
agreement in showing that homeless individuals’ diets are inadequate in fiber and calcium and
are high in sodium, fat, and cholesterol.
Method of Meal Provision on Nutrition Quality of Meals
The second question asked was: “Is there a difference in how well meals meet Dietary
Reference Intakes (DRIs) between homeless shelters that have outside volunteers provide meals
and shelters that provide meals themselves.” In this study, the average nutrient content of meals
between volunteer and staff planned was similar in that they both were below 1/3 of the DRI for
calcium and vitamin D, and both types of meals provided well over twice the sodium
requirements. Surprisingly, staff-planned meals fell short of 1/3 the DRI for certain nutrients that
volunteer-planned meals did not. These included carbohydrates for all genders, total fiber for
women, and iron for women. This could be due to the fact that staff-planned meals relied on
donated foods for their meal preparation, which made it difficult to follow nutrition standards. To
date there are no other studies that have examined nutrient content of solely volunteer-planned
meals.
35
Shelter’s Use of Nutrition Standards and Nutrition Quality of Meals
The third question was: “Is there a difference in how well meals meet Dietary Reference
Intakes (DRIs) between homeless shelters that have nutrition standards for meals served and
shelters that do not,” and in this study, 14 out of 21 shelters had nutrition standards that they used
to plan meals for their clients. The shelter nutrition standards listed by participants consisted of
food groups that must be represented at each meal, with some shelters specifying amounts of
food to be supplied in each food group. There are currently no governmental regulations or
national standards for the nutrition of food served at homeless shelters. Therefore, it is up to
shelters to create their own nutrition standards if they desire. The education of shelter directors
may play an important part in their decision to establish nutrition standards. Those who are
educated may be more aware of the need for nutrition standards when planning meals. In this
study, 85% of shelter directors had attended college while the rest had high school diplomas. In
addition, nine (42.86%) reported having completed at least one nutrition course, which would
increase their likelihood of valuing nutrition. In a 2015 study by Koh et al.,8 only one out of the
nine shelters had nutrition standards. To help improve nutrition, some cities have developed
nutrition standards for shelters at the local level. For example, in 2008, New York City
government implemented standards for foods purchased and served by city agencies, including
shelters and soup kitchens.8 The standards require a minimum number of fruits and vegetables;
requires beverages to be 100% juice, water, or low-fat milk; and recommends serving whole
grains.8 The city government also prohibits agencies from accepting donations of candy and
sugar-sweetened beverages.8 These are more specific and targeted nutrition standards than those
36
that only require certain food groups at each meal. Requiring whole grains rather than simply a
starch would greatly improve the well-established finding that homeless diets are low in fiber.
Prohibiting certain foods from being donated such as sugar-sweetened beverages would also
improve the healthfulness of meals. In the current study, 66.67% of the shelters had standards
without a government mandate.
The qualitative results provided more insight into the shelter directors’ views regarding
nutrition needs of the shelters. Themes from their responses about barriers to following nutrition
standards included the use of food donations, the use of volunteer-provided meals, lack of
resources, and the special dietary needs of clients. Shelters explained they do not get to choose
the nutritional value of food donated, and donated food must be used first, which often changes
the meal plan considering the nutrition standards. A Canadian study examining the effect of food
donations on nutrition found 14 out of the 18 participating programs used food donations, with
the majority also purchasing some food.27 The total energy contributed by donated foods varied
widely, as well as the amount of food donations from one meal to the next.27 This supports what
directors in this study stated pertaining to the unpredictable nature of the use of food donations
on nutrition quality of meals prepared. To help overcome the donations barrier, the 2015 study by
Koh et al.8 recommends shelter staff should encourage healthy food donations rather than the
common high-starch, high-sugar donations. In turn, corporate and community partners can be
more cognizant and provide these healthier options.8 Scouten et al.28 also acknowledged this,
stating shelters should ask for specific food items rather than only working with what is donated.
37
How Registered Dietitians Could Help with Homeless Shelter Meals
A 2001 research study found 87% of shelters had a willingness to integrate
recommendations from medical providers and nutritionists into meal planning.19 In the current
study, one of the themes identified was how a registered dietitian could be of help to homeless
shelters. The participants stated that one of the ways a registered dietitian could be of assistance
was to provide nutrition education and counseling to shelter clients. This need was also identified
in the 2015 study by Koh et al.,8 in which only one out of nine shelters provided food education
to guests and two out of eight provided food education to kitchen staff. While previous studies
have shown that nutrition education is an effective tool in increasing nutrition knowledge of
homeless shelter clients and staff, they have also reported that it alone does not improve quality
of the diet.11,12 In one study, the cafeteria staff’s ability to demonstrate their learning was impeded
by the constraints of food donations, thus limiting the shelter client’s ability to choose healthful
options.12 Nutrition intake does not improve unless policies are also implemented to ensure
healthful food is available for shelter clients. The second identified theme pertaining to registered
dietitian need was to aid in meal planning to improve meal nutrition, and the third theme was to
identify food to serve clients with specific dietary needs. Many shelters do not have the resources
or specific knowledge required to help clients with specific dietary needs due to various health
conditions. A study by Davis et al.9 revealed shelter clients believe the food may even contribute
to their health conditions. Many shelter residents stated they believed the shelter food contributed
to poor management of chronic illness or caused them to suffer symptoms such
as constipation, headaches, fatigue, and weight gain.9
38
Limitations
The primary limitation of this study is a small sample size, due to a 10.77% response rate.
The low response rate suggests that shelter directors who did respond were more motivated
and/or valued the role nutrition on health more than those who did not respond. Therefore, the
shelters that participated may have better nutrition or policies related to nutrition than other
shelters. This was reflected in that 85.7% of those who responded had staff-planned meals rather
than volunteer-planned meals. In addition, some shelters may have lacked the staff and/or were
too busy serving clients. Indeed, when contacted, those who did not participate stated they were
too busy, short staffed, or unwilling to coordinate with volunteers who planned meals.
Another limitation was the data collection method due to limited time and budget. The original
plan was to observe the meals and serving sizes at the participating shelters, but the plan was
revised due to a lack of time. As a result, dinner information was self-reported rather than
observed, and the amount of food consumed by shelter clients was not determined. Therefore, the
meals reported may not be representative of the actual nutrients consumed. In addition, some
shelter dinner information reported was more detailed than others, varying the level of accuracy
for each shelter.
Last, the 1/3 DRI set as the standard for comparison for the nutrient analysis was
conservative. As previous research has reported, many homeless only consume two meals a
day.22 A goal of 50% of DRI is more appropriate for the homeless population.22 It is also
conservative because as Tse and Tarasuk27 discuss in their research, the DRI estimates apply to
healthy adults, and research shows this population are in poor health with nutrition deficiencies.
CHAPTER V
CONCLUSION
This study aimed to find if the method of meal provision (staff-planned versus
volunteerplanned meals) and nutrition standards influenced nutrition quality of meals served.
Due to a small sample size, the relationship between shelter nutrition policies and nutrition
quality of meals served could not be determined. The qualitative data gathered suggests food
donations are the main factor in limiting meal planning and override nutrition standards.
Therefore, future research should focus on the nutritional value of food donations and how to
encourage healthful food donations.
The findings on the nutrient analysis of dinner meals from 11 shelters suggest that
homeless shelter dinners may be inadequate overall in calories, carbohydrate, fiber, calcium, and
vitamin D and are excessive in sodium, fat, and cholesterol. Most of the shelters were lacking in
dairy products and only provided water and/or juice with meals. Shelters could set a nutrition
standard of providing dairy products with meals to improve the calcium and vitamin D content of
meals. To improve fiber, shelters could change the common nutrition standard of providing a
starch with every dinner to providing a whole-grain product with every dinner. Shelters
could request the services of registered dietitians to volunteer or consult at shelters to help
improve the nutrition quality of meals and address the themes identified in this study. Shelters
would like help with meal planning, nutrition education of clients, and identifying foods for
clients with special dietary needs.
