The North Carolina BEAUTY and Health Project: Preventing Cancer in African American Beauty Salons (Cancer Disparities) Part 2

Intervention Development—Theoretical Foundations

The social ecological framework (SEF) (Stokols, 1992; Stokols, Pelletier, & Fielding, 1996) was a helpful heuristic for conceptualizing the multilevel intervention that was developed and tested in the BEAUTY study. According to the SEF, there are multiple levels of influence on each targeted customer behavior (i.e., increasing fruit/vegetable intake, reducing fat intake, and increasing physical activity). For this trial, we focused on three (of five) levels of influence within the SEF (see Figure 9.1): intrapersonal (e.g., stylist and customer knowledge, attitudes, self-efficacy, motivations); interpersonal (e.g., interactions between the stylist and customer about health messages); and organizational (e.g., access and/or support within the salon environment). According to the SEF, behavior change will be more likely if one is able to engage multiple levels of influence. We utilized constructs from both social cognitive theory (SCT) (Bandura, 2000) and the transtheoretical model of behavior change (TTM) (Prochaska, Redding, & Evers, 1997).

At the intrapersonal level, we focused on characteristics within the individual (customer or stylist); specifically, readiness to change (TTM construct) and self-efficacy to change the targeted behavior (SCT construct). Intervention strategies to move the individual along the stages of readiness to change, and to improve self-efficacy to make the change, were embedded in several components of the stylist training workshops (e.g., discussions on separating fact from fiction and "how to" sessions for making health changes) and the targeted health magazines (e.g., features on participating stylists who had made positive health changes).

At the interpersonal level, we focused on the interactions between the stylist and customer. In the stylist training workshops, we used demonstrations and role plays to increase the self-efficacy of stylists to include the key messages in conversations with customers during a typical salon visit. We did not demand that stylists deliver all messages at each visit. Instead, we told stylists that they were the "experts" about each customer, and encouraged them to assess the "readiness" of each customer to accept one or more messages on a typical visit. In the targeted health magazines, we included spotlights on successful stylists and customers as role models for success—consistent with the observational learning construct of SCT as a way of building the self-efficacy of the customer for the desired behaviors as well as building the self-efficacy of the stylist to talk with the customers. At the organizational level, we focused on the SCT construct of "reciprocal determinism," where a customer’s health behavior influences (and is influenced by) the salon environment. Thus, each campaign included a display where the key messages were highlighted in an interactive, attractive way that would allow customers to view the display while waiting in the salon, or where stylists could refer customers to pick up materials and information as they were leaving the salon.

In addition to key theoretical foundations, social marketing principles guided the development of each campaign by including themes, simple/clear health messages, and attractive, eye-catching materials for the salon, stylist, and customer that were customer focused and pretested before each quarterly campaign launch. All materials in the displays, stylist training workshops, and targeted health magazines were designed with a similar BEAUTY Project "look," including consistent colors, font style, and logo.

This intervention study did not test a single theoretical model; instead, we were conceptually organized using the social ecological framework, and responsive to calls for more theory-driven intervention studies that had speci-fied (a priori) key theoretical constructs. This multilevel intervention reinforced cancer prevention messages through a variety of educational methods expected to influence customers, stylists, owners, interactions between customers and stylists, and the salon environment itself. The conceptual model depicting how the intervention strategies at each level of the SEF work to influence study outcomes is depicted in Figure 9.1. Additional details are provided about the specific components/features of the targeted health magazines and stylist training workshops included in each campaign, as well as the intent of the educational displays.

Stylist Training Workshops

Stylists and owners in salons randomized to the STW or BOTH conditions were invited to attend a 4-hour training workshop for each of the six quarterly campaigns. The workshops were held at local public libraries and on Mondays, when salons were typically closed for stylist convenience. The workshops were organized and conducted by the research team, with assistance from Advisory Board members (stylists, owners who were credible sources of information for participating stylists), with occasional outside experts (nurses or physicians), and/or cancer survivors. The workshops were designed to increase stylist/owner knowledge about the core health messages, and help them develop ways to share the health messages with their clients during normal hairstyling appointments. The workshops covered the same key components for every campaign, including: (1) an update on BEAUTY Project timeline and upcoming events; (2) sharing success stories from past campaigns; (3) a review of the targeted health messages specific to the campaign, along with an expert presentation on the topic addressed by the health messages (e.g., nutrition, exercise, cancer screening); (4) the introduction of new in-salon display materials; (5) role-playing on initiating conversations and delivering targeted health messages; (6) a problem-solving session, including a group discussion around barriers and facilitators; and (7) making personal promises as a goal-setting exercise. The stylists/owners were provided with written training materials covering the information they needed for each campaign, along with a three-ring binder in which to store the materials. If a salon did not have at least one stylist at a workshop, then a project staff member would conduct a make-up training session with participating stylists from that salon within 2 weeks of the missed workshop. Stylists attending the workshops (but not those who received make-up trainings) received stipends to cover their time and travel costs, and were provided lunch.

Targeted Health Magazines

Enrolled customers of salons randomized to the TM and BOTH conditions received six quarterly, targeted health magazines, delivered to their homes. In addition, a copy of the magazine was delivered to the salons randomized to the TM and BOTH conditions, so that customers, stylists, and owners would be able to read and discuss the magazine contents in the salons. The topical content of the quarterly magazines corresponded to the six health campaigns of the BEAUTY Project. Each eight-page magazine featured educational articles, interactive quizzes, and photographic images of African American women who were actual participating customers and stylists. Similar to the in-salon displays, the magazines had an overall

BEAUTY Project design and format, as well as a consistent structure across the six campaigns, including: (1) a project update; (2) a customer success/ feature story; (3) "Ask Your Stylist" BEAUTY and health tips; (4) a feature story on one health behavior/screening guideline; (5) an interactive game, puzzle, or trial skill (e.g., a healthy recipe); (6) community resources—where to go for more information; (7) a personal promise pledge; and (8) a medical disclaimer. A new targeted "Beauty From the Inside Out" magazine was delivered to each participating customer in the TM or BOTH arm, on the same timetable as each stylist training workshop that was held for the campaign launch.

In-Salon Educational Displays

All participating salons received a tri-fold educational display with key messages highlighted for the six health campaigns. Salons in the CONTROL arm received displays with the same format and design, but with information unrelated to the primary study outcomes. The displays included the following components for every campaign: (1) a set of targeted health messages specific to the campaign; (2) print materials or handouts; (3) an interactive quiz, or "try it out" tips, related to the targeted health messages; (4) suggestions for getting more information or related resources; and (5) an "Ask Your Stylist" cue to action, with at least one picture of an enrolled stylist. Depending on the campaign, additional print materials were provided to the salons, such as brochures, flyers, or stickers for the booth mirrors with the targeted health messages (another cue to action for stylist-customer interaction). Project staff delivered materials and set up displays in the salons at the beginning of each campaign. A total of six displays with associated print materials were offered quarterly over the intervention period, in all study conditions.

Salon Recruitment

Salons were eligible to participate in the study if they: (1) were located within a 75-mile radius of Chapel Hill; (2) were not part of a franchise; (3) served primarily African American customers; and (4) served at least 75 customers. Once salons were approached, eligibility was verified, and an initial interest in participating was established, a project team member brought a 13-minute recruitment video that included extensive interviews with Advisory Board members, who extolled the benefits of joining the study and encouraged salon owners and stylists to enroll. Interested salons that were able to achieve customer recruitment standards (at least 55) were enrolled.

CUSTOMER RECRUITMENT

Salon customers were recruited for the BEAUTY Project during enrollment events held in the enrolled salons. As described above, each salon was required to recruit at least 55 customers during these events to participate in the project, and only those salons that met this requirement were eventually enrolled. A customer was eligible to join the study if she was an African American woman, at least aged 18, and she was a regular customer of the salon (e.g., had visited the salon more than twice). If she met these initial criteria, then she was asked to sign an informed consent agreement form and complete a Physical Activity Readiness Questionnaire (PAR-Q). If she answered "Yes" to any of the PAR-Q questions, she was asked to have a physician provide medical consent for her to participate in the study. Eligible customers were asked to complete the baseline BEAUTY and Health survey (BHS) questionnaire, and return it either in a postage-paid envelope to the research team or drop it off back at the salon, in a sealed envelope; telephone surveys were made available to participants unable to return the paper version of the questionnaire. Once the completed BHS questionnaire was returned and received, then the customer was considered fully enrolled in the study. The initial 40 salons recruited a total of 1, 209 customers, while 1,123 customers were enrolled from the 37 salons that remained in the study. Figure 9.2 shows study participation at the salon and customer levels over time.

Measurement Schedule

To evaluate the primary (percentage of calories from fat, servings of fruits/ vegetables, physical activity) and secondary (weight and screening behaviors) outcomes at the customer level, we compared baseline and follow-up responses to the BEAUTY and Health customer survey. The 19-page BHS was distributed to each enrolled customer. Dietary habits were measured at baseline and follow-up using the 60-item version of the NCI Health Habits and History Questionnaire (Block et al., 1986), which has been validated with a low-income Black population (Coates et al., 1991). From this measure, we calculated daily fruit and vegetable servings and the percentage of calories from fat at both time points for each cus-tomer.We measured physical activity using a single item, "I currently engage in regular physical activity," with respondents answering either "Yes" or "No." Prior to this item, the questionnaire included an explanation of what constitutes physical activity and what it means for physical activity to be "regular." Body mass index (BMI) was calculated using self-reported estimates of height and weight. In determining adherence to cancer screening guidelines, we used recommendations by the American Cancer Society (ACS) for breast, cervical, and colorectal cancer (American Cancer Society, 2001). For breast cancer, we classified women as adherent if they reported a mammogram within the past year at follow-up (among women 40 years of age and older). Women reporting a Pap smear test within the past 3 years at follow-up were considered adherent. For colorectal cancer, we considered women age 50 years and older adherent if they reported a fecal occult blood test (FOBT) in the past year or a flexible sigmoidoscopy or colonoscopy within the past 5 years (note that data were not available for colonoscopy tests within the past 10 years). We also collected information on customer demographics, health status, and salon behavior.

Note: B = Baseline; F = Follow-up.

FIGURE 9.2 The BEAUTY CONSORT Diagram

To evaluate interactions between customers and stylists, we assessed changes from baseline to follow-up from customer self-report of health talk on the BHS. In addition, we recorded conversations (20) during the salon observations at multiple time points during the intervention period, using established protocols. A 2-hour observation was conducted in each salon by a trained female African American observer on a "busy" day as determined by the salon owners, with "conversation" as the unit of analysis. Stylist "reach" was defined as the proportion of all licensed cosmetologists per participating salon that enrolled in the BEAUTY Project, calculated as the total number of enrolled stylists divided by the total number of licensed stylists working at the salon (as reported by the owner).

RESULTS

Participating Salons

Of the 12,319 licensed salons in the state, a total of 5,119 were found to be located within a 75-mile radius of Chapel Hill. Salons known to be part of franchises and those whose telephone numbers could not be confirmed were excluded. Staff identified 2,628 salons and they were assessed for eligibility: that they catered mostly to African American customers; had more than 75 "regular" customers; and had at least one stylist willing to sign on to the study. From those approached, 62 salons were interested and eligible to participate (see the CONSORT diagram; Figure 9.2). Of those 62 salons, 40 were ultimately able to meet the customer recruitment goals and were randomized into one of the four intervention arms. However, following randomization, three salons that were randomized into the CONTROL arm withdrew from the study, making the final number of participating salons 37. Forty-two owners (several salons had more than one owner) and an additional 27 licensed stylists were enrolled from the 37 participating salons.

From interviews with all participating salon owners at baseline (100% response rate, n = 40), we learned that the typical salon in our study was open for 10.3 years; had an average of three employees who worked in the salon for an average of 5.9 years; and most owners (63%) reported being in the salon daily. Sixty-seven percent of owners believed that participation in the BEAUTY Project would improve the reputation of the salon in the community; and 92.5% would recommend participation in a study like the BEAUTY Project to other salon owners.

Customer Description

Overall, 1,123 customers enrolled in the study from the 37 participating beauty salons and completed a baseline survey, and 559 (49.8%) completed the follow-up survey. Retention patterns overall and by study arm from baseline to follow-up are summarized in Figure 9.2. At baseline, the average customer age was 38.4 years, 55% were single, 47% were college graduates or higher, 34% had a yearly household income between $24,999 and $49,999, and 34% had a yearly household income of $50,000 or more. Among enrolled customers, 73% were overweight/obese, 36% reported getting moderate, regular physical activity (PA), 86% consumed more than 30% of their total calories per day from fat, and 20% reported eating at least five servings of fruits and vegetables per day (Table 9.3). At baseline, there were significant group differences by education level (x2 = 14.73, p < .°5) and age (F(3,1056) = 4.91, p < .01), which were adjusted for in the final modeling efforts.

Most customers (88%) reported attending the salon at least once every 7 weeks; 57% visited every 2-4 weeks; and 18% visited at least once weekly. Nearly all (98%) tried to see the same stylist at each visit; 69.7% reported spending between 1.5 and 3 hours per visit, and 18% spent more than 3 hours per visit (Linnan et al., 2007). On average at baseline, customers reported talking with their stylists "very much/a lot" about health; and only 11% reported their health as "excellent."

Approximately half of the customers at baseline completed the follow-up survey (n = 559/1,123) and those who were retained were more educated (x2 = 12.92, p < .01; HS: 15.6 vs. 14.9, some college: 28.8 vs. 37.5, college graduate: 55.7 vs. 47.6); had a higher income (x2 = 6.27, p < .05; < 24.9K: 18.4 vs. 23.5, 25-49.9K: 38.7 vs. 38.4, 50K+: 43.0 vs. 38.0); were more likely to be overweight/obese as measured by BMI (x2 = 8.03, p < .05; normal: 19.6 vs. 25.3, overweight: 31.5 vs. 28.9, obese: 48.3 vs. 44.5); and were less likely to self-report more favorable general health status (x2 = 203.51, p < .001; poor/fair: 42.4 vs. 11.3, good/excellent: 57.6 vs. 88.7) compared to the initial group at baseline.

Primary Outcomes

As shown in Table 9.4, there were no significant differences between intervention groups on our primary or secondary outcomes (i.e., daily percentage of calories from fat, daily fruits and vegetables, BMI, and regular PA). However, there were some promising modest improvements in health behaviors among participants overall. On average, the daily percentage calories from fat decreased for all respondents (F{1422) = 9.22, p < .01), with minimal (but nonsignificant) differences between the treatment arms. Increased consumption of daily fruit and vegetable servings were reported by all respondents, regardless of treatment condition. Additionally, while the majority of respondents did not report engaging in regular physical activity at either time point, there was an increase in the overall number of participants who reported being moderately physically active at follow-up, regardless of treatment condition. And at follow-up, the majority of respondents reported being adherent to age-appropriate cancer screening guidelines for breast, cervical, and colorectal cancer.

TABLE 9.3 Customer Characteristics at Baseline (N = 1,123)

	All N = 1,123		Control N = 216		Magazine N = 304		Training N= 306		Both N = 297
Demographic Characteristics	N	m (SD)	n	m (SD)	n	m (SD)	n	m (SD)	n	m (SD)
Age (years)3	1,060 Freq	38.5 (12.0) %	205 Freq	36.8 (11.9) %	290 Freq	37.3 (12.6) %	287 Freq	39.1 (11.8) %	278 Freq	40.3 (11.5) %
Weight status
Normal	275	26.6	47	23.3	75	27.1	75	26.6	65	23.8
Overweight	299	28.9	60	29.7	86	31.0	79	28.0	74	27.1
Obese	460	44.5	90	44.6	114	41.2	125	44.3	131	48.0
Education level3
High school or less	163	14.9	23	11.0	40	13.4	56	18.7	44	15.4
Some college	410	37.5	94	44.8	107	35.9	116	38.8	93	32.5
College graduate or more	520	47.6	93	44.3	151	50.7	127	42.5	149	52.1
Household income ($)
< 25k	240	21.4	41	20.8	67	23.7	74	27.3	58	21.6
25-49.9k	392	34.9	83	42.1	106	37.5	87	32.1	116	43.1
> 50k	388	34.6	73	37.1	110	38.9	110	40.6	95	35.3
Marital status
Married	489	45.0	92	44.2	137	46.0	137	43.5	123	45.0
Single	598	55.0	116	55.8	161	54.0	161	56.5	160	55.0

Health Characteristics	N	m (SD)	n	m (SD)	n	m (SD)	n	m (SD)	n	m (SD)
	Freq	%	Freq	%	Freq	%	Freq	%	Freq	%
General health
Poor/Fair	119	11.3	27	13.2	36	12.5	29	10.2	27	9.7
Good/Excellent	936	88.7	177	86.8	251	87.5	256	89.8	252	90.3
Current regular PA
No	695	64.1	136	64.8	190	68.1	183	61.6	186	65.0
Yes	390	35.9	74	35.2	102	34.9	114	38.4	100	35.0
Daily percentage from fat
< 30	134	13.9	28	14.8	29	11.3	37	14.0	40	15.7
> 30	831	86.1	161	85.2	228	88.7	228	86.0	214	84.3
Five daily fruits and
vegetables
No	878	79.8	167	78.8	241	81.4	246	81.7	224	77.0
Yes	222	20.2	45	21.2	55	18.6	55	18.3	67	23.0

Significant group differences. Totals may be less than the stated sample size due to missing data. m= mean, SD = standard deviation.

TABLE 9.4 Customer Baseline/Follow-up on Primary and Secondary Outcomes, by Arm

All N = 559		Control N = 106		Magazine N = 151		Training N = 148		BOTH N = 154
Daily percentage of calories from fata
n	m (SD)	n	m (SD)	n	m (SD)	n	m (SD)	n	m (SD)
Baseline 429	35.9 (5.5)	83	36 (5.5)	122	35.9 (5.1)	111	35.8 (5.1)	110	35.9 (5.4)
Follow-up 426	35.1 (5.9)	83	35.4 (5.7)	122	34.6 (5.3)	111	35.2 (5.6)	110	35.4 (6.6)
Difference 426	-0.8 (5.1)	83	-0.5 (5.6)	122	-1.3 (4.5)	111	-0.6 (4.8)	110	-0.6(5.5)
Daily servings of fruits and vegetables
Baseline 490	3.6 (1.9)	92	3.9 (1.7)	127	3.6(1.9)	130	3.2 (1.8)	141	3.8 (1.9)
Follow-up 490	3.7 (1.8)	92	3.8 (1.8)	127	3.7(1.7)	130	3.6 (1.9)	141	3.7 (1.8)
Difference 490	0.1 (1.9)	92	-0.8(1.7)	127	0.1 (2.0)	130	0.3 (1.9)	141	-0.1 (1.9)
Baseline 497	30.2 (7.1)	97	30.0 (6.8)	130	29.8 (6.8)	134	30.2 (7.5)	136	30.7 (7.1)
Follow-up 497	30.6 (6.9)	97	29.5 (6.9)	130	30.0 (6.7)	134	31.1 (7.4)	136	31.9 (11.2)
Difference 497	0.4 (3.6)	97	.4(2.6)	130	2 (3.3)	134	0.9(4.3)	136	0.2 (3.7)
Number who reported engaging in regular PA
Freq	%	Freq	%	Freq	%	Freq	%	Freq	%
Baseline 390	35.4	74	35.2	102	34.9	114	38.4	100	35.0
Follow-up 238	43.0	42	40.0	63	42.3	62	42.5	71	46.4

aSignificant differences from baseline to follow-up.

Totals may be less than the stated sample size due to missing data. m = mean, SD = standard deviation.

FIGURE 9.3 Observed Health Talk in Salons: Salons With Stylist Training Workshops (STW) Versus Salons With No Stylist Training Workshops (Non-STW)