Use of focus groups to understand African-Americans’ dietary practices: Implications for modifying a food frequency questionnaire

Abstract

Objective

To generate information about dietary practices, food preferences and food preparation methods from African-Americans in Macon County, Alabama, as a precursor to an intervention designed to modify an existing dietary health questionnaire (DHQ).

Method

African-American males (30) and females (31) ages 20 to 75 years participated in eight focus groups in Macon County Alabama between June and July, 2007.

Results

The core topics identified were dietary practices; food preferences; food preparation methods; fast food practices; and seasonal/specialty foods. The younger focus group participants reported consuming mostly fast foods such as hamburgers for lunch. Fruits, vegetables, salads, fish, chicken and sandwiches were the most common lunch foods for the older males and females. Across the groups, rice, cornbread and potatoes were reportedly the most commonly consumed starchy foods at dinner. Frying and baking were the most common cooking methods. Fewer participants reported removing the skin when cooking chicken versus those who did not remove. Traditional foods including fried green tomatoes and cracklings were selected for addition to the modified DHQ, while those not commonly consumed, were deleted.

Conclusions

Participants described high-fat traditional food preferences, common frying and addition of salted meats to vegetables, which informed the modification of a DHQ.

Introduction

African-Americans suffer disproportionately from many chronic diseases including cancer (Wong et al., 2002). Diet is a significant, modifiable risk factor associated with cancer (WCRF/AICR, 2007). However, there is little knowledge regarding cancer-related dietary habits, food preferences and food preparation practices among African-Americans in the rural South. The semi-quantitative food frequency questionnaire (FFQ), which recalls the frequency and portion sizes of foods consumed over a specific time period, is a dietary assessment instrument for epidemiologic research (Keating et al., 2007). Although FFQs are an accepted dietary assessment instrument in a various groups, their validity for this population is uncertain. FFQs effectively measure habitual dietary intake, only if they are tailored to the target population (George et al., 2004).

Ethnic differences in dietary intake between blacks and whites were determined for low-income mothers using a FFQ (Perkin et al., 1988). Whites had low mean intakes of vitamin A and niacin; Blacks had high intakes of niacin, protein, soft drinks, chicken, hot dogs, white rice and collard/mustard greens. In an assessment of dietary fat, high intake was associated with greater insulin resistance in black and white children in Alabama (Weigensberg et al., 2005). In the high-consumption group, Blacks had 32% lower insulin sensitivity and 62% higher acute insulin response to glucose, compared to those with normal consumption. This effect was specific to black children and was not seen in whites.

Greenberg et al. (1998) computed a “healthy diet” score for Southern-born and other Blacks in Central Harlem; Southern-born Blacks had higher-risk diets than native Blacks. Shankar and Long-Dilwort (2004) reported that for college students, black females had higher fat and caloric intakes, and were more influenced by cultural traditions than Whites. Elderly Blacks in rural areas of southern states consumed fewer servings of bread, cereals, fruits, vegetables, milk and cheese, and more meat, poultry and fish than recommended (Lee et al., 1998). For African-American women, a FFQ using seven broad food categories correlated more with reference values than one using 31 individual fruits and vegetables (Warnke et al., 2001). A FFQ for the Lower Mississippi Delta (LMD) showed that regional food use patterns differed from national patterns, and between African-American and Caucasian adults (Tucker et al., 2005). African-Americans in the rural South have dietary patterns that differ from national patterns and from those in the LMD (Smith et al., 2006).

One commonly used FFQ is the dietary health questionnaire (DHQ), which is validated for the general American population (Thompson et al., 2002; Subar et al., 2001). Although focus groups are generally underutilized in diet and nutrition research, they are suitable for exploring, describing, and understanding the dietary habits of African-Americans because of the historical importance of oral communication in their culture. Elmubarak et al. (2005) used focus groups to explore perceptions regarding diet, nutrition and cancer among Sudanese-Americans. Carter-Edwards et al. (1998) used focus groups with African-Americans to generate information for questionnaire development and to provide information regarding diet and blood pressure. Vazir (2002), Begley (2002), Schlundt et al. (2003), Vuckovic et al. (2000) and James (1998) also utilized focus groups to collect dietary information from African-Americans. In our study, focus groups with African-Americans in Macon County were used to generate information about dietary practices, food preferences and food preparation methods to allow for modification of an existing DHQ. The long-term goal is to develop a modified DHQ for use in epidemiologic studies of diet and cancer in African-Americans in the rural southern U.S.

Methods

Question development

The questions used by the facilitator in the focus groups were developed through review of the literature, brainstorming sessions with faculty, staff, students and community members at Tuskegee University (TU). These interactions led to the development of a preliminary database of foods commonly eaten by African-Americans in the rural south. Questions from the preliminary database were pre-tested in two pilot focus groups consisting of 19 African-American males and females.

Design

A qualitative research design involving focus group methodology was utilized (Krueger, 1994; Betts et al., 1996; Greenhalgh and Taylor, 1997). Focus group research includes four basic designs: tightly structured, phenomenological, loosely structured and exploratory (Morgan, 1988, 1998). The interview guide reflected the tightly structured design; the phenomenological process primarily involved eliciting experiences related to food preparation methods; and the exploratory design involved the use of probes to elicit food preferences. Appropriate guides for the recorders/observers guides were developed (Elmubarak et al., 2005; Mahoney et al., 2003).

Participants

Approval to conduct the study was obtained from the Human Participants Review Committees at TU and the University of Alabama at Birmingham. Invitations to participate in the focus groups were given to African-Americans from Shorter, Franklin and Tuskegee in Macon County, Alabama. Invitations were posted at various locations throughout Macon County, Alabama. Flyers were posted at local banks, churches, grocery stores, convenience stores, beauty salon and barber shops. Verbal invitations were also given to community members. To be included in the study, participants had to be African-American, 20–75 years old, able to speak and write English and willing to participate. The response rate for the study was 95.3%. Participants attended one of eight focus groups, which varied by age and gender. For each group, the purpose of the study was explained; participants were assured the information they provided would be confidential, and a brief demographic questionnaire was completed.

Procedure

Two pilot and eight formal focus groups were held in June and July 2007. Meetings lasting approximately 100 min were conducted by a trained African-American facilitator because the ethnicity of the moderator was a variable that could be controlled (Sattler, 1970). She was selected for the study because of her education, experience in focus group moderation, residency in Macon County; and cultural and linguistic appreciation/understanding of the participants. Notes relating to body language, facial expressions, levels of enthusiasm and other nonverbal responses were taken by three recorders/observers. Six to 10 persons participated in each focus group, and although the remuneration was not advertised during the recruitment phase, each participant was paid a modest stipend of $15.00 per hour.

Data collection

Focus groups followed the structured guide shown in Table 1. Although the focus groups were gender-specific, the topic did not warrant gender-matching of the moderator, since she has experience in moderating all-male groups. It is not always necessary to gender-match the moderator to the group (Day, 1993). At the start of each session, reconnaissance exercises were accomplished. The questions were open-ended, neutrally presented, conversational, clear and concise (Krueger and Casey, 2000). Probe questions were used to gain specific and detailed information. Discussions were audio-taped and transcribed. Responses were also recorded on flip charts, and the transcripts were compared to these written notes (Neumark-Sztainer et al., 1999).

Table 1.

Core topics and structured interview guide used in the focus group meetings held in Macon County, Alabama in June and July, 2007.

Warm-up/reconnaissance	Introduction and rapport building exercises. Participants used adjectives to describe African-Americans and to define Soul Food
In-depth interview
Dietary practices	Participants were asked to describe some common breakfast, lunch, dinner and snack foods and how they prepared or cooked fruits and vegetables.
Food preferences	Participants discussed the most commonly eaten fruits, vegetables, meats, cereals, dairy foods and desserts. Participants were given food lists to discuss whether they ate those foods, how often and how they prepared them.
Food preparation methods	Participants were asked to describe the most common cooking methods used for pork, beef, other meats, chicken, and other poultry, whether they removed the skin from chicken, whether they trimmed fats when meats are cooked, and what was added to vegetables when they cooked them.
Fast food practices	Participants were asked to describe their favorite fast food restaurants, favorite meals and frequency of consumption at fast food restaurants
Seasonal/specialty foods	Participants discussed foods eaten during the different seasons (winter, spring, summer, fall). Participants discussed the consumption of specialty foods, such as venison (deer), rabbit, ‘possum’.

Data analysis

Transcribed data were placed into qualitative text analysis software (QRS NUD*IST 6, Victoria, Australia), which assisted entry, coding, filing and organizing the discussions into themes. Each recurrent theme was coded and discussed until consensus was reached. Notes regarding body language, facial expressions and other nonverbal response cues were used in interpretation of the transcripts. As a validity check, a summary report of the major findings was prepared and reviewed by the research team (Neumark-Sztainer et al., 1999).

Results

Pilot focus groups

Two pilot focus groups, subsequently excluded from the formal groups, consisted of nine and 10 African-American males and females, respectively, between the ages of 20 and 75 years. The pilot groups served to confirm the content and cultural sensitivity of the foods identified in the structured interview and to elicit traditional, culturally important foods, which may have been overlooked. Since the findings indicated that most participants routinely consumed fast foods and beans, the structured interview was modified to reflect questions about their consumption.

Formal focus groups—participants’ characteristics

Eighty African-American males and females were contacted for the formal focus group; 61 agreed to participate; their characteristics are in Table 2.

Table 2.

Selected characteristics of participants (n=61) in the focus group meetings held in Macon County, Alabama in June and July, 2007.

Age group	Gender		Total	Range of occupations in each age group
Years	Male	Female
18–35	9	10	19	Students, Officer, Musician, Self-employed, Hotel Clerk
36–50	9	6	15	Groundsman, Police Officer, General Manager, Fire Medic, Personal Trainer, Housekeeper, Housewife, Cashier, Caregiver
51–65	6	8	14	Food Service Worker, Homemaker, Retired Teacher, Senior Center Manager, Hostess, Retired Program Director, Lawn Care, Retired
66–75	6	7	13	Retired, Homemaker, Retired Roofer
Total	30	31	61

Overall themes

In the reconnaissance section, there was substantial overlap in the defnitions for “Soul Food”. Consistently included were okra, chitlins, chicken feet, ham hocks, greens (turnip, collard, mustard) and fried chicken.

Dietary practices

Breakfast foods described among individuals and across groups were similar. They included: cold beverages (orange juice, milk); starchy items (biscuits, pancakes, toast, French toast, hash browns); protein items (eggs, bacon, salmon croquettes, sausage, ham, cheese); cereals (grits, oatmeal, white rice); and hot beverage (tea, coffee). Participants described their preferences for cold breakfast beverages as: “Milk!!! I love milk. Love orange juice too.”

Three different lunch food emerged. i) younger participants mentioned fast foods such as hamburgers, French fries, hot dogs, pizza, wings and Big Macs. Most expressed consuming “fast foods twice or thrice weekly”; one female described consumption of vegetables daily; two said that they consumed salad 2–5 times weekly. ii) For the 36–50-year-old women, foods included sandwiches of ham and cheese, bologna, turkey, tuna, banana and peanut butter/jelly and hogshead cheese. Fast foods included hamburgers, French fries and hot dogs. Most women in this group indicated consumption of “fast food fried chicken twice/week.” Of the nine men in this age group, four reported consuming chicken four times/week. iii) The lunch food pattern for older males and females included hot dogs and hamburgers, described as “only once/week.” Fruits, fish, vegetables, salads, baked chicken, pork chops and turkey sandwiches were frequently consumed. Dinner foods, which were similar among individuals and across groups, included meats, dairy, desserts, vegetables and starchy foods (Table 3). Rice, cornbread and potatoes were the commonly consumed starchy foods; peas and beans were also reported. About cornbread, participants stated, “I love fried cornbread, I eat it every chance I get” and “I eat buttermilk cornbread, I grew up on it, I eat it every weekend.”

Table 3.

Common dinner foods and food preferences reported by African-Americans in eight focus group meetings held in Macon County, Alabama in June and July, 2007.

Fruits	Vegetables and starchy foods	Meat and products	Dairy	Desserts
Apples	Asparagus	Bacon	Yogurt	Cakes—chocolate, red velvet, cheese, pineapple upside down, pound, cheese pound cake
Banana	Beans—green, string/snap, Lima, baked, pinto, butter	Barbecue ribs	Cream—whipped, sour	Ice cream
Blueberries	Peas—early, black-eyed, field, sweet, purple hull	Beef—roast, steak, ground, stew with potatoes	Soymilk	Pies—apple, blackberry, mulberry, blueberry, cherry, coconut pie, cream cheese, pecan, buttermilk, lemon meringue, sweet potato, pumpkin, peach cobbler
Cantaloupe	Beets		Butter	Puddings—banana, rice, bread
Grapefruit	Cornbread	Bison	Ice cream	Cookies and Brownies
Grapes	Greens—collard, turnip, mustard	Bologna	Cheese—cream, cottage	Fruits
Honey dew melons	Turnip with pigtails	Camp stew—a stew containing corn, tomatoes, beef, chicken, early peas and other vegetables, which is usually made from leftovers.	Milk	Smoothies
Nectarine	Bell peppers	Chicken—fried, smothered, baked	Eggs	Bars—nutty, granola
Oranges	Broccoli	Chitlins	Eggnog
Peaches—canned	Brussels sprouts	Crab legs	Buttermilk
Pears	Carrots	Duck	Margarine
Pineapple	Cabbage—smothered	Fish—salmon, whiting, haddock
Strawberries	Cauliflower	Frog’s legs
Tangerine	Celery	Gumbo
Watermelon	Corn—fried, corn and tomatoes	Ham, Ham hocks
	Cucumbers	Ham hocks with string beans and potatoes
	Frozen mixed vegetables	Hotdogs
	Lettuce	Hamburger
	Okra—boiled, fried	Lamb
	Onion	Liver and gravy
	Potatoes	Meatloaf
	Rutabagas	Neck bones
	Rice	Oxtail
	Spinach	Pork—fried, baked, chops, roast, ribs
	Sweet potato	Sausage—corn and sausage, pinto beans and sausage
	Squash—fried, casserole	Pig—tail, feet, ears
	Tomatoes	Quail
	Zucchini	Rabbit
		Shrimp
		Turkey—necks, wings, legs—smoked, regular
		Turtle

For the 36–50-year-old males, snack foods were fruits and potato chips. Snacks of choice for the older males were potato chips, sherbet and baby carrots; their female counterparts preferred fried pork skins (cracklings). Young females reported eating fruits and cookies daily; most reported snacking on ice cream at least twice weekly. For the males and females, snacks included fruits, candy bars, cookies, popcorn with butter, cheesecake, puddings, peanuts and cashew nuts.

Food preferences

The participants discussed 11 food lists, which were compiled as described earlier. Based on consensus, the foods commonly eaten, and those not usually consumed were selected to be added and deleted, respectively, from the modified Dietary Health Questionnaire (MDHQ) (Tables 3, 4). Comments about the added and deleted foods included: “Turnips greens and pot likker (liquor), all of us love that.” “Fried okra, fried corn I love it.” “Would not know what a millet look like.”

Table 4.

Food preferences reported by African-Americans through discussion of the food lists in the eight focus group meetings held in Macon County, Alabama in June and July, 2007.

Food list item	Foods added to the DHQ	Foods deleted from the DHQ
Fried cornbread	x
Hot biscuits	x
Cornbread and buttermilk	x
Fried green tomatoes	x
Turnip greens with ‘pot likker’ (the liquid from boiled greens eaten as a type of gravy or as a drink in and of itself)	x
Stewed okra with tomatoes	x
Fried okra	x
Succotash	x
Rutabagas	x
Fried cabbage
Fried corn
Lima beans with ham hocks	x
Ham hocks/ham bone	x
Salt pork	x
Pig’s feet	x
Pig’s ears	x
Tripe		x
Hog jowl	x
Neck bones	x
Breaded fried pork chops	x
Cracklings (fried pork skins)	x
Gumbo	x
Dodgers, “Philpy”		x
Sauerkraut		x
Cauliflower	x
Brussels sprouts	x
Catsup (called ketchup)	x
Bagels		x
English muffin		x
Sweet muffins		x
Dessert breads		x
Turkey nuggets		x
Soy burgers		x
Liverwurst		x
Tacos	x
Burritos	x
Tostados, Tamales, Fajitas, Enchiladas Quesadillas		x
Chimichangas		x
Soybeans, Nut butter		x
Bulgur, Cracked wheat, Millet, Tofu		x
Stuffed shells, Stuffed manicotti, Tortellini		x
Ravioli	x

Food preparation methods

The most common cooking methods were frying, baking, grilling, boiling, steaming and microwaving. For poultry and pork, frying was most common followed by baking, grilling or barbecuing, boiling and broiling. Across the eight focus groups, fewer participants said that they removed the skin when cooking chicken. Approximately half of the participants reported trimming fats when they cooked meats. Table 5 shows preparation methods for vegetables; for vegetables, boiling and frying were common.

Table 5.

List of food items usually added to vegetable when cooking as reported by African-Americans in the eight focus group meetings held in Macon County Alabama in June and July, 2007.

Meat, poultry, meat-related	Spices and seasonings	Fats, oils, oil-related	Miscellaneous
Salt pork	Spices—cinnamon	Cheese	Sour cream
Pig’s feet	Peppers—black, bell	Oil	Okra
Ham hocks	Sauce	Bacon	Lemon
Meat	Herbs	Butter	Sugar
Fatback	Seasonings—smoked, Mrs. Dash	Shortening	Vegetable broth
Neck bones	Celery
Turkey—smoked, parts, neck	Thyme
	Seasoning salt
	Salt—sea, Kosher light, low sodium
	Onions
	Garlic
	Curry powder

Fast food practices

The most commonly frequented fast food restaurants were MacDonald’s, Church’s Fried Chicken, Wendy’s, Subway, The Coop, Burger King, Checker’s, Captain D’s and In and Out. The most common meals included bacon cheese burgers, Big Macs, triple Whoppers with extra cheese, French fries, fried chicken and fish, meatball subs, and thin crust pizzas.

Seasonal and specialty foods

The participants reported the seasonal and specialty foods shown in Tables 6 and 7. The 18–36-year-old females reported chitlins, crabs, deer, duck, quail, guinea birds and turtles. Their male counterparts reported deer and turtle but added raccoon, frog legs, squirrel, alligator, crayfish and ‘possum’.

Table 6.

Seasonal foods as described by African-Americans in the eight focus group meetings held in Macon County, Alabama in June and July, 2007.

Foods	Season/occasion
Gumbo	Winter
Chitlins (all participants, all focus groups)	Winter, Thanksgiving, New Year
Crabs	Fall
Veggies	When in season
Turkey, turkey dressing	Thanksgiving, Xmasa, New Year
Ham	Holidays, Thanksgiving, Christmas
Hen	Holidays
Watermelon (all participants, all focus groups)	4th July, summer
Hog jowl	New Year’s
Sweet potato pie	Thanksgiving
Beef stews	Fall
Pumpkin pie	Easter
Boiled eggs	Easter, Summer
Sugarcane	Fall, winter
Cranberry	Late summer
Muscadine	Late summer
Supadines	Winter

^aChristmas.

Table 7.

Specialty foods described by African-Americans in the eight focus group meetings held in Macon County, Alabama in June and July, 2007.

Deer	Duck	Raccoon with sweet potato	Calamari
Guinea Bird	Alligator	Quail	Buffalo
Squirrel	Crayfish	Frog legs	Rabbit
Escargot	Goat meat	Possum	Wild Turkey
	Venison	Turtle, Turtle Soup	Caviar

Discussion

Indigenous cultural and historical backgrounds overarch food preferences and food preparation methods, and these factors influence eating behaviors (Bronner et al., 2001; Pomerleau et al., 2001; Lau et al., 1984). Assessment of dietary practices among African-Americans requires culturally-specific information, not typically found in existing standard FFQs. There is a disproportionate diet-related mortality in African-Americans, and a need for culturally relevant interventions (Airhihenbuwa and Kumanyika, 1996). Knowledge of culturally-specific diet-related habits is necessary for changing dietary practices (Kristal et al., 1994). The focus groups provided relevant information useful for modification of the FFQ (DHQ). Pertinent findings from the focus groups revealed their preference for traditional foods, which represents a large part of their cultural heritage.

There was similarity between the seasonal/specialty foods and items in early diets of African-Americans. For example, ‘possum’ and raccoon were meats of choice among the earliest African-Americans (Browser and Eckstein, 1970; Bernstein, 2008). Chitlins and foods such as ‘pot likker’, hogshead cheese and cracklings, constituents of early African-Americans’ diets, were mentioned in every focus group. These early African-American diets largely shaped the dietary habits passed down to generations of African-American descendents. Most families in the South did not consume fresh meats (Dirks and Duran, 2001); the traditional diet was based on salt-preserved pork (Smith et al., 2006).

In our study, African-Americans described the same pork-based eating practices with high consumption of pigs’ feet, fried pork and addition of processed, salted meats when cooking vegetables. Many of these foods are rich in nutrients, but their preparation, which involves addition of fat, may contribute to the disproportionate cancer burden for this group (Patterson et al., 1995). As examples, participants reported adding butter and salted meats when cooking vegetables (Table 6). The practice of frying foods and using fats in cooking was a practice for early African-American families (Dirks and Duran, 2001). Black women prefer fried foods more than white women (Dacosta and Wilson, 1996) and African-Americans obtain much of their dietary fat from luncheon meats, bacon and fried poultry (Popkin et al., 2001). In urban, African-American adolescents, there is excessive consumption of fat (Di Noia et al., 2008). These preferences are of concern because polycyclic aromatic hydrocarbons and heterocyclic amines associated with ill-health in humans are formed when meat, chicken and fish are cooked at high temperatures (Weisburger, 2002; WCRF/AICR, 2007). Additionally, consumption of dietary fat is a risk factor for obesity (WCRF/AICR, 2007). Dairy food preferences described were not as varied as other foods. In African-American populations, dairy food consumption is low (Fulgoni et al., 2007; Jarvis and Miller, 2002).

Study limitations and strengths

Although the focus groups used open-ended questions, the topics probed were largely predetermined by the structured interview protocol (Airhihenbuwa and Kumanyika, 1996). The data collection protocol (including choice of moderator, group size and composition) was designed to improve validity of the information collected, but it is difficult to estimate the validity of the responses (O’Doutherty et al., 2006). Participants were well represented across age and gender, but they were not randomly selected and may not be representative of all African-Americans. Also, the potential for bias or misinterpretation of the results exists. Despite these limitations, this study adds to the limited data on food preferences, dietary practices and food preparation methods of African-Americans in the rural South, and will allow for culturally-specific interventions to help African-Americans reduce their cancer risk.

Conclusions

African-Americans have unique dietary practices, which should be taken into consideration when modifying or creating tools for dietary assessment. The current results provide information for the development of a modified FFQ (the DHQ) tailored to capture their dietary intake. Focus group participants articulated traditional, cultural foods, which were selected to be added to the DHQ; others, not commonly consumed, were chosen to be deleted.