Research in Social and

Administrative Pharmacy 11 (2015) 891900

Original Research

Disease-specic direct-to-consumer advertising of

pharmaceuticals: An examination of endorser type and
gender eects on consumers attitudes and behaviors
Nilesh S. Bhutada, Ph.D.a, Brent L. Rollins, R.Ph., Ph.D.b,*
a

Department of Clinical & Administrative Sciences, California Northstate University, College of Pharmacy, Elk Grove, CA
95757, USA
b
Pharmacy Practice, Philadelphia College of Osteopathic Medicine (PCOM) Georgia Campus, School of Pharmacy, 625
Old Peachtree Rd. NW, Suwanee, GA 30024, USA

Abstract
Background: Direct-to-consumer (DTC) advertising is still a controversial topic for pharmaceutical
manufacturers and researchers, and while numerous studies have examined the DTC phenomenon, little
research has examined the eect of gender, particularly gender of the endorser and consumer.
Objective: The objective of this research was to assess the impact of the endorser (celebrity vs. expert vs.
non-celebrity) and gender both gender of the endorser and gender of the consumer on consumers
attitudes and behaviors in response to a print disease-specic direct-to-consumer advertisement.
Methods: Using Qualtrics consumer panel, data were obtained for 514 US adults (age 18 years and above)
who demonstrated at least minimal symptoms of depression and need for monitoring based on the Patient
Health Questionnaire (PHQ) score. Data were analyzed using a 3 (Endorser Type: Celebrity/Expert/NonCelebrity)
2 (Endorser Gender: Male/Female)
2 (Consumer Gender: Male/Female) full factorial
between subjects multivariate analysis of variance (MANOVA) and necessary univariate analysis.
Results: Only the type of the endorser (celebrity vs. expert vs. non-celebrity) used in the ad had a signicant
main eect on the dependent variables. Further univariate analyses revealed that, of the several dependent
variables, endorser type had a signicant inuence only on attitude towards the ad, attention paid to the
ad, and endorser credibility, with gender being non-signicant in all cases.
Conclusions: Expert endorser generated signicantly more favorable levels of attitude towards the ad, and
endorser credibility compared to the non-celebrity endorser. Celebrity endorser attracted more consumer
attention towards the ad and generated favorable endorser credibility perceptions compared to the noncelebrity endorser. However, celebrity and expert endorsers did not signicantly dier from each other on
the abovementioned ad eectiveness variables. Lastly, endorser gender and consumer gender did not have
a signicant inuence on ad eectiveness.
2015 Elsevier Inc. All rights reserved.
Keywords: Direct-to-consumer advertising; Disease-specic advertising; Celebrity endorsements; Endorser eects; Pharmaceutical marketing; Gender eects

Funding support: California Northstate University College of Pharmacy; no other financial disclosures.
* Corresponding author. Tel.: 1 678 407 7352; fax: 1 678 407 7347.
E-mail address: [email protected] (B.L. Rollins).
1551-7411/$ - see front matter 2015 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.sapharm.2015.02.003

892

Bhutada & Rollins / Research in Social and Administrative Pharmacy 11 (2015) 891900

Introduction
Advertisers have continually used celebrities as
endorsers to market their products and services to
consumers. It is estimated that approximately
15% of all US advertisements feature a celebrity
spokesperson,1 costing companies additional fees
to have the celebrity appear in the various advertisements (e.g., print or broadcast media) and
endorse their product/service. The proclivity for
using celebrities among brand managers and advertisers is the perception a famous personality
will make their brand more appealing and prominent in the crowded marketplace.
Owing to erce global competition, generic
introductions, formulary restrictions, etc., pharmaceutical marketers have faced growing challenges in the marketplace and, thus have used a
variety of marketing strategies in order to gain
and/or maintain market share.2,3 Drawing from
their popularity in consumer goods advertising
and given the potential to attract consumer attention and increased return on investment,48 celebrity
endorsers
are
also
employed
by
pharmaceutical marketers. While use of celebrity
endorsers is a relatively new phenomenon in
direct-to-consumer prescription drug advertising
(DTCA), within a short period of time, their use
in prescription drug campaigns has substantially
increased. Celebrities such as Sally Field, Mike
Ditka, Paula Deen, and Phil Mickelson are just
a few who have been employed to endorse prescription pharmaceuticals.
Additionally, in recent years, pharmaceutical
marketers have shifted some of their marketing
eorts from product-specic to disease-specic, or
non-branded, advertising. This can be attributed to
the growing debate over the benet of productspecic DTCA, public scrutiny, lawsuits against
pharmaceutical manufacturers,9,10 and skepticism
about the underlying intentions of pharmaceutical
manufacturers advertising specic products.11,12 According to the U.S. Food and Drug Administration
(FDA), disease/health education and medication
compliance should be the primary focus of diseasespecic advertising without mentioning or representing a particular drug or device.13 Compared to their
use in product-specic/brand-name prescription
drug DTC advertising, the use of celebrity endorsers
in disease-specic DTCA is all the more recent, such
as Danica Patrick, a female race car driver, promoting Boehringer Ingelheims Drive 4 COPD
campaign (www.drive4copd.com) informing consumers about chronic obstructive pulmonary disease

(COPD) which is often treated by Boehringer Ingelheims Spiriva.

The primary endorser types used in DTC
advertising include celebrities, experts, or general/typical consumers. Since the beginning, pharmaceutical marketers have used typical consumers
or patients, depicting an individual tting the
demographics of a typical patient aected by
the disease state the advertised medication treats,
in their DTC ads owing to the fact that the
consumer watching the commercial can relate to
the person featured in the DTC advertisement.
Although perhaps not to the extent as celebrity
endorsers, in recent years, there has also been a
steady increase in the use of expert endorsers in
DTC advertising. Credentials and/or qualications of the expert endorser are the primary focus
and appealing factor in these advertisements
where, typically, a physician (or at least an individual portraying a physician) recommends a
particular brand-name product in a DTC advertisement. Prior research suggests increased source
credibility associated with expert endorsements
enhances the ad believability.14
Considerable evidence exists in the literature
about how men and women process promotional
information dierently. According to the Selectivity Hypothesis theory,15,16 men, because they
mostly rely on salient cues, have been characterized as selective processors, and women,
because they seek all the relevant information
before forming any impressions, have been characterized as comprehensive processors. Widgery and McGaugh17 reported that women can
be more easily persuaded than men, suggesting
a discrepancy in how the two sexes respond to
persuasive messages. Berney-Riddish and
Areni18 found that, generally, compared to
men, women were more accepting of the advertising claims in the ad, with their reactions
turning more negative towards the ads compared
to men when the ads used hedges such as may
and probably, or pledges such as denitely
and absolutely, in their advertising claims.
While men do not prefer objective claims in advertisements, women tend to equally prefer subjective and objective advertising claims, under
low risk conditions.19
Ohanian8 found no signicant eect of consumers gender on their likelihood to purchase
the celebrity endorsed product or their evaluation
of the celebritys eectiveness on the three dimensions namely, attractiveness, trustworthiness, and

Bhutada & Rollins / Research in Social and Administrative Pharmacy 11 (2015) 891900

expertise. On the contrary, Boyd and Shank20

found that consumers (both men and women)
perceived endorsers of the same gender as their
own compared to the opposite gender to be
more trustworthy. Peetz, Parks, and Spencer21
found that male athlete endorsers compared to female athlete endorsers were signicantly more
likely to inuence consumers purchase intentions
and were four times more likely to be correctly
identied by the consumers; and male consumers
were more favorable even towards the less
familiar male athlete endorsers.
Despite the growing number of promotional
dollars being spent on celebrity endorsements in
a host of pharmaceutical marketing campaigns,
there is still a lot to be learned about the
consumers perceptions of the advertising eectiveness of the various endorsers used in DTCA.
While several studies in the past have been
grounded in assessing the impact of celebrity
endorsers in product-specic DTCA,22,23 only
one study assessed the impact of celebrity endorsements
in
disease-specic
DTCA.24
Further, inuence of expert (e.g., a physician)
endorsements on consumers remains unclear,
and, furthermore, the impact of the gender of
the endorser and the gender of the consumer
on consumers responses to such endorsements
has not been examined in the context of
DTCA. Historically in traditional media,
gender has been used as a key segmentation variable for product positioning and marketing; as
such, it is important to evaluate the impact of
gender eects on consumers responses to
advertising endorsements. Therefore, the aim
of this study was to assess how the gender of
the consumer inuences advertising eectiveness when exposed to an advertisement
endorsed by a male versus a female endorser
(whether celebrity, expert, or non-celebrity).
The following testable hypotheses were developed to achieve this purpose:
H01: Consumers will not dier in terms of their
levels of attitude toward the ad (Aad), attitude
toward the company (Ac), attention paid to
the ad (Atnad), endorser credibility (EC),
intent to seek more information (I-Seek),
intent to discuss the condition with their
physician (I-Discuss), and intent to ask for
a prescription for the condition (I-Ask) between the dierent type of endorsers (celebrity vs. expert vs. non-celebrity) featured in
the ad.

893

H02: Consumers will not dier in terms of their

levels of Aad, Ac, Atnad, EC, I-Seek,
I-Discuss, and I-Ask between male endorser
and female endorser.
H03: Male consumer will have the same levels
of Aad, Ac, Atnad, EC, I-Seek, I-Discuss,
and I-Ask compared to the female
consumer.
H04: Consumers will have the same levels of
Aad, Ac, Atnad, EC, I-Seek, I-Discuss, and
I-Ask when the product is either promoted
by a male endorser or a female endorser
regardless of the type of endorser (celebrity
vs. expert vs. non-celebrity) featured in the
ad.
H05: Male and female consumers will not dier
in terms of their levels of Aad, Ac, Atnad, EC,
I-Seek, I-Discuss, and I-Ask regardless of
the type of endorser (celebrity vs. expert vs.
non-celebrity) featured in the ad.
H06: Male and female consumers will not dier
in terms of their levels of Aad, Ac, Atnad, EC,
I-Seek, I-Discuss, and I-Ask regardless of
the gender of the endorser (male vs. female).
H07: Male and female consumers will have the
same levels of Aad, Ac, Atnad, EC, I-Seek, IDiscuss, and I-Ask for the dierent types of
endorsers (celebrity vs. expert vs. noncelebrity) regardless of the endorser gender.

Methods
Advertising stimuli
Fictitious disease-specic advertising stimuli were
used to maximize internal validity and reduce the
confounding eects emanating from previously used
disease-specic DTC ads. Previous research
in disease-specic DTCA25,26 focused on oral
contraceptives and allergies, both able to be selfdiagnosed and treated via over-the-counter medicines. To overcome this issue, depression was selected
as the disease condition for this study due to the level
of the disease complexity and consumers inability
to diagnose and/or treat the condition using
over-the-counter medications. The ad stimuli
(example Appendix I Male Expert Ad) featured
a slogan/website tag line (www.morethansadness.
com) and a ctitious pharmaceutical company (C
& C Pharmaceuticals, Inc.). Six ads were created
for this study taking into account the three levels of
endorser type and two levels for endorser gender.
The content of all 6 ads were exactly the same with

894

Bhutada & Rollins / Research in Social and Administrative Pharmacy 11 (2015) 891900

the only change being the endorser type (celebrity vs.

expert vs. non-celebrity) featured in the ad. Selection
of celebrity endorsers (male actor Harrison Ford
and female actress Ashley Judd) was based on their
documented history of depression, consensus among
a panel of experts, and availability of high resolution
digital pictures. To avoid confounding eects due to
the dierences between the dierent ad versions, persons with similar age, physical appearance, and
demeanor were used as expert (physician) and noncelebrity endorsers. All ads were created in accordance with the guidelines proposed by the FDA for
disease-specic DTC ads.13 The study protocol was
approved by the Institutional Review Board of the
Philadelphia College of Osteopathic Medicine.
Study measures
Independent variables
The type of endorser featured in the diseasespecic DTC ad (celebrity vs. expert vs. noncelebrity), gender of the endorser (male vs.
female), and gender of the consumer (male vs.
female) were the three independent variables used
in the study. The two variables, type of the
endorser and the gender of the endorser, were
manipulated by means of the advertising stimuli.
Dependent variables
The primary interest of this study was to assess
advertising eectiveness. Therefore, commensurate
with the literature and the interests of pharmaceutical advertisers, the dependent variables of interest
were attitude toward the ad (Aad), attitude toward
the company (Ac), attention paid to the ad (Atnad),
endorser credibility (EC), intent to seek more information (I-Seek), intent to discuss the condition
with their physician (I-Discuss), and intent to ask
for a prescription for the condition (I-Ask). Attention paid to the ad (Atnad) by consumers was
measured using a ve-item Likert-type scale (with
scale points 1 very little and 5 very much, coecient a .84).27,28 A three-item, seven-point, semantic dierential scale was used to measure Aad
(bad-good, unpleasant-pleasant, unfavorablefavorable, coecient a .93).29 Eectiveness of the
various types of endorsers used in the study was
assessed by measuring the endorser credibility
(EC) construct. Consumers attitude towards EC
was measured using a seven-item, seven-point, semantic dierential scale (unattractive-attractive,
unclassy-classy, insincere-sincere, untrustworthytrustworthy, not an expertexpert, inexperiencedexperienced, and very unfamiliar-very familiar, coecient a .86).30,31

A brand-name is not featured in diseasespecic DTC advertisements. As such, consumers

attitudes towards the company (Ac) promoting
the disease-specic DTC advertisement was
measured instead of the traditionally measured
attitude toward the brand. Another reason for
measuring Ac is that the logo and name of the
pharmaceutical company appears in both
product-specic and disease-specic DTC advertisements. Thus, Ac was used to replace attitude
toward the brand since the pharmaceutical company serves as the single identier in the case of
disease-specic DTC advertisements and was
measured using a three-item, seven-point, semantic dierential scale (badgood, negativepositive,
unfavorablefavorable, coecient a .98).25,32
Three, 11-point probability based and previously validated scales33,34 (with options ranging
from 0 chances in 100 to 99 chances in 100) were
used to measure the impact of disease-specic
DTCA on consumers behavioral intentions.
These three scales included consumers intention
to seek more information on depression (I-Seek),
make an appointment to discuss depression with
their physician (I-Discuss), and ask for a prescription drug for depression (I-Ask).
Procedure
A three (Endorser Type: Celebrity/Expert/
Non-Celebrity)
two (Endorser Gender: Male/
Female)
two (Consumer Gender: Male/
Female) full factorial between subjects experimental design was used to analyze the impact on
the dependent variables Aad, Ac, Atnad, EC, ISeek, I-Discuss, and I-Ask (i.e., the ad eectiveness variables). In order to achieve a power of
.8, at a priori signicance level of .05, and medium eect size (f .25), a sample of 27 participants was needed in each of the 12 cells created
by the factorial design leading to a total sample
size of at least 324 participants for the study.35
A Qualtrics consumer panel was used to obtain
the data for US adults (18 years and older). A total of 514 responses were obtained via Qualtrics
online survey tool. In a randomized control study
design, participants were randomly assigned to
one of the six ad stimuli. Quota sampling was
used to ensure that about equal number of
male and female participants were assigned to
each of the six ad stimuli in order to achieve
the minimum required sample size for each of
the cells created by the factorial design. Participants were prescreened prior to their inclusion

Bhutada & Rollins / Research in Social and Administrative Pharmacy 11 (2015) 891900

in the study using the Patient Health Questionnaire (PHQ), a validated nine-item scale used
by primary care physicians for diagnosis and
monitoring of depression.36,37 In order to qualify
to be included in the study, participants should at
least demonstrate minimal symptoms of depression and need for monitoring (i.e., PHQ
score O 4). Further, physicians, pharmacists,
and employees of an advertising/market research
agency were excluded from the study.
Results
Demographic characteristics
Given the quota sampling used in this study,
there was an even distribution of gender in the
sample. Over half (58.5%) of the sample indicated
they had completed at least some college education. The major group represented in the sample
was Caucasians (86.8%) followed by AfricanAmericans (7.4%). Almost half of the sample
(47.3%) was comprised of middle-aged individuals

895

(3655 years of age). Table 1 provides the specic

sample demographics.
Hypotheses testing
The overall inuence of the type of the
endorser, gender of the endorser, and gender of
the consumer on the dependent variables (Aad, Ac,
Atnad, EC, I-Seek, I-Discuss, and I-Ask) was
examined using a 3
2
2 between subjects
factorial multivariate analysis of variance (MANOVA). The inequality of the variance-covariance
matrix between the groups (Boxs M 447.93,
P ! .001) singled violation of MANOVA
assumption. Therefore, MANOVA results were
interpreted using Pillais Trace, a more robust
test statistic to MANOVA violations.
Endorser type was the only independent variable that had a signicant main eect (F 4.22,
P ! .001) on the dependent variables. Gender
of the endorser (F 1.368, P .217) and gender
of the consumer (F 1.032, P .408) did not
have a signicant main eect on the dependent

Table 1
Demographic characteristics of survey participants (n 514)
Variable

Categories

Frequency

Percentage (%)

Gender

Male
Female
Less than high school
High school graduate
Associates degree
Completed some college
College graduate
Graduate school or higher
American Indian or Alaska native
Asian
Black or African-American
Caucasian
Hispanic or Latino
Native Hawaiian or other Pacic Islander
1825 yrs
2635 yrs
3645 yrs
4655 yrs
5665 yrs
Above 65 yrs
Less than $15,000
$15,000$24,999
$25,000$34,999
$35,000$49,999
$50,000$74,999
$75,000$99,999
$100,000 or more

255
259
11
113
89
148
104
49
3
6
38
446
19
2
44
94
107
136
113
20
92
90
70
86
95
43
38

49.6
50.4
2.1
22.0
17.3
28.8
20.2
9.5
.6
1.2
7.4
86.8
3.7
.4
8.6
18.3
20.8
26.5
22.0
3.9
17.9
17.5
13.6
16.7
18.5
8.4
7.4

Level of education

Race

Age

Annual household
income

896

Bhutada & Rollins / Research in Social and Administrative Pharmacy 11 (2015) 891900

Table 2
MANOVA depicting
manipulations

signicance

of

various

Eect

Pillais
F
Trace Test
statistic

Endorser type
Endorser gender
Consumer gender
Endorser type
endorser
gender
Endorser type
consumer
gender
Endorser gender
consumer
gender
Endorser type
endorser
gender
consumer gender

.112
.019
.014
.038

ad

P
value

4.220 !.001
1.368
.217
1.032
.408
1.380
.156

.014

.508

.929

.005

.383

.912

.027

.964

.489

variables; thus, we failed to reject null hypothesis

H02 and H03, respectively. As a result, the 2-way
and 3-way interactions among the independent
variables were not signicant. Accordingly, we
failed to reject null hypotheses H04 thru H07. Results of MANOVA with all the main and interaction eects are provided in Table 2.
Since endorser type had a signicant main
eect, univariate ANOVAs were performed to
examine the eects of the various endorsers on
the dependent variables. Results of the ANOVAs
(see Table 3) indicate that H01 is partially supported, as consumers only exhibited signicantly
dierent levels of Aad (F 5.38, P .005), Atnad
(F 4.64, P .01), and EC (F 20.54,
P ! .001) among the three endorser types (celebrity vs. expert vs. non-celebrity). Post hoc tests
were performed to further analyze these dierences. Results indicate that consumers exposed

to expert endorsers had signicantly more favorable Aad compared to the consumers exposed to
non-celebrity endorsers (M 5.40 vs. 4.93,
P .004). However, consumers exposed to
non-celebrity endorsers did not dier signicantly from celebrity endorsers in terms of their
Aad (M 4.93 vs. 5.24, P .09). Similarly, consumers exposed to celebrity endorsers did not
dier signicantly from consumers exposed to
expert endorsers in terms of their Aad
(M 5.24 vs. 5.40, P .51).
Consumers paid more attention to the ad
endorsed by the celebrity endorser compared to
the ad endorsed by the non-celebrity endorser
(M 3.82 vs. 3.56, P .009). However, there was
no signicant dierence among non-celebrity and
expert endorsers in their potential to attract consumers attention to the ad (M 3.56 vs. 3.63,
P .71). Similarly, expert endorsers and celebrity
endorsers did not signicantly dier from each
other in terms of attracting consumers attention
to the ad (M 3.63 vs. 3.82, P .07).
Expert (M 5.29, P ! .001) and celebrity
(M 5.04, P ! .001) endorsers were perceived
as signicantly more credible than non-celebrity
endorsers (M 4.56). However, there was no signicant dierence in perceived endorser credibility
between expert and celebrity endorsers (M 5.29
vs. 5.04, P .09).
Discussion
Legally permissible only in two countries, the
US and New Zealand, the highly popular promotional strategy of advertising prescription
medications directly to consumers has faced
considerable criticism and scrutiny over the
past two decades. Given its potential to create

Table 3
Analyses of variance by endorser type across dependent variables
Source

Dependent variable

Type III sum

of squares

Mean square

P value

Endorser type

Attitude toward the ad

Attitude toward the company
Endorser credibility
Attention paid to the ad
Intent to seek more information
Intent to discuss the condition with
their physician
Intent to ask for a prescription for
the condition

20.31
6.06
46.94
6.46
18.51
36.78

10.16
3.03
23.47
3.23
9.26
18.39

5.38
1.65
20.54
4.64
1.04
1.73

.005
.193
!.001
.010
.356
.178

68.70

34.35

2.86

.059

Bhutada & Rollins / Research in Social and Administrative Pharmacy 11 (2015) 891900

unrealistic expectations among consumers about

drug benets,38,39 DTCA can possibly lead to
unnecessary demand for prescription medications39,40 and can potentially have a negative
impact on public health. Critics are particularly
concerned when pharmaceutical companies use
certain strategies to market their products, such
as using a celebrity to endorse a product.41
Critics argue that because of their popularity, celebrities might have a persuasive impact on consumers to request a prescription drug they may
not actually need.
Pharmaceutical marketers are increasingly employing celebrity endorsers in DTCA of prescription medications for the same reason they are
heavily used in traditional consumer good advertising, their public recognition. Given the current
issues surrounding DTCA912 and the considerable fees associated with celebrity endorsements,
the question of particular interest to marketers is
whether or not the use of celebrity endorsers is
benecial for their campaigns and generates
more ad eectiveness in terms of positive attitudinal and behavior change and eventual dialog
with physicians.
In line with previous ndings, celebrity endorsers were able to attract more of consumers
attention towards the advertisement. This is
important from a pharmaceutical marketers
perspective, as this strategy could help them
distinguish their message/campaign in an increasingly cluttered marketplace particularly in the
case of the ad type used in this evaluation,
disease-specic DTCA. Additionally, consumers
showed signicantly more favorable attitude towards the ad when it was endorsed by an expert
compared to a non-celebrity endorser, and
perceived the expert endorser to be signicantly
more credible than the non-celebrity endorser.
Health information and treatment can be associated with signicant risk, and, thus, consumers
might exhibit a greater condence level in the
message when it is coming from an expert or
trusted source. By virtue of their training and
prescribing power, physicians are seen as health
information and treatment experts.
Post hoc comparisons conrmed expert endorsers are not dierent from celebrity endorsers
in terms of their impact on ad eectiveness.
Therefore, our data support that for the kind of
fees paid to the celebrity endorser to feature in the
campaign; pharmaceutical manufacturers are not
necessarily getting a comparable return on their
investment. In fact, pharmaceutical marketers

897

might achieve similar, or even better, returns by

using the more economical expert endorsers (or an
actor portraying an expert). This is a strategy
pharmaceutical marketers might consider adopting given the criticisms surrounding DTCA for its
undue inuence on consumers. Further, given the
uncertainty in the health care sector with the new
health care reform act and current economic
climate, from a pharmaceutical marketers
perspective, use of expert endorsers appears to
be the most cost-eective choice. The results of
our study should also ease some concerns of
critics and government regulators as consumers
exposure to the celebrity ads did not signicantly
increase their intention to make an appointment
with their physician or request for a prescription
product for depression.
Compared to the credibility dimension of the
source/endorser of the message, another important source characteristic gender has not
received much attention in the DTCA literature.
Moreover, mixed evidence exists about the impact
of endorsers gender on the ad eectiveness variables.1921 Our results indicate that endorser
gender does not have an impact on the ad eectiveness variables. Given the ubiquitous nature
of DTCA over the past 15 years, gender response
dierences appear to no longer exist. In particularly for non-gender specic diseases or product
classes, such as depression, pharmaceutical marketers need not alter their messaging hoping to
gain a response from a non-responding gender.
Generally, women are more caring in nature
than men,42 and, for the most part, are the caretakers of family health care matters.43,44 Not surprisingly, ndings from past research suggest that
awareness of DTCA is high among women4547
and they are more likely to be inuenced by these
types of promotions.4649 The non-signicant
interaction eects are an indication that even
though women generally are more receptive to
DTCA and tend to approach a doctor about
advertised drugs,4648 they will not necessarily
get swayed away because a product is endorsed
by a celebrity.
Limitations and directions for future research
The results of this study should be considered
in light of its limitations. There are limits to the
generalizability of the results of this study, given
the overrepresentation of Caucasians in the study
sample compared the US population distribution.50 Another limitation is the forced exposure

898

Bhutada & Rollins / Research in Social and Administrative Pharmacy 11 (2015) 891900

to the advertising stimuli. It is likely that consumers respond dierently to a forced exposure
compared to a natural exposure. Further, this
research focused only on one disease state
depression. Future studies should examine if these
results hold up for advertisements within other
disease states. The cross-sectional design was
used in the study captures consumers attitudes
and behavioral intentions at a particular point in
time. Since consumers attitudes and behavioral
intentions evolve over time, future studies should
attempt to assess consumers attitudes and behavioral intentions in a longitudinal study design and
check if the results are comparable. Celebrities
used in the study were from the entertainment industry. Future eorts should assess if similar results are observed with celebrities from other
arenas, such as sports, politics, etc.

Conclusion
Results of this study indicate that endorsers
used in the disease-specic DTC ad had some
inuences on consumers. In particular, the expert
endorser (i.e., the physician portrayed in the ad)
generated more favorable consumer attitudes
towards the disease-specic DTC ad compared
to the non-celebrity endorser. The celebrity
endorser was able to signicantly attract consumers attention to the disease-specic DTC ad
and compared to the non-celebrity endorser. Both
the expert endorser and the celebrity endorser
were perceived as more credible compared to the
non-celebrity endorser. However, celebrity and
expert endorsers did not signicantly dier from
each other on the abovementioned aspects.
Lastly, endorser gender and consumer gender
did not emerge as signicant predictors of consumer attitudes and behavioral intentions, and
thus did not have a signicant inuence on ad
eectiveness.
References
1. Ace Metrix. Celebrity Advertisements: Exposing a
Myth of Advertising Eectiveness, http://mktg.
acemetrix.com/acton/fs/blocks/showLandingPage/
a/563/p/p-001d/t/page/fm/0; 2010.
2. Macias W, Lewis LS, Baek TH. The changing face
of direct-to-consumer print advertising: policy and
content issues. Pharmaceut Med 2010;24:165177.
3. Main KJ, Argo JJ, Huhmann BA. Pharmaceutical
advertising in the USA: information or inuence?
Int J Advert 2004;23:119142.

4. Kahle LR, Homer PM. Physical attractiveness of

the celebrity endorser: a social adaptation perspective. J Consum Res 1985;11:954961.
5. Kamins MA. An investigation into the match-up
hypothesis in celebrity advertising: when beauty
may be only skin deep. J Advert 1990;19:413.
6. Keller KL. Conceptualizing, measuring, managing
customer-based brand equity. J Mark 1993;57:122.
7. Mowen JC. On product endorser eectiveness: a
balance model approach. Curr Issues Res Advert
1994;3:41.
8. Ohanian R. The impact of celebrity spokespersons
perceived image on consumers intention to Purchase. J Advert Res 1991;31:4654.
9. Berenson A. Analysts See Merck Victory in Vioxx
Deal. (Cover Story). New York Times (Front
Page, Morning Edition)157:. p. A1.
10. Kesselheim AS, Studdert DM, Mello MM. Whistleblowers experiences in fraud litigation against
pharmaceutical companies. New Engl J Med 2010;
362:18321839.
11. Friedman M, Gould J. Consumer attitudes and behaviors associated with direct-to-consumer prescription drug marketing. J Consum Mark 2007;
24:100109.
12. Spake DF, Joseph M. Consumer opinion and eectiveness of direct-to-consumer advertising. J Consum Mark 2007;24:283292.
13. FDA. Guidance for Industry: Help-seeking and
Other Disease Awareness Communications by or on
Behalf of Drug and Device Firms, http://www.fda.
gov/CDER/GUIDANCE/6019dft.pdf; 2004.
14. Maddux JE, Rogers RW. Eects of source expertness, physical attractiveness, and supporting arguments on persuasion: a case of brains over beauty.
J Pers Soc Psychol 1980;39:235244.
15. Meyers-Levy J, Maheswaran D. Exploring dierences in males and females processing strategies.
J Consum Res 1991;18:6370.
16. Meyers-Levy J, Sternthal B. Gender dierences in
the use of message cues and judgments. J Mark
Res 1991;28:8496.
17. Widgery R, McGaugh J. Vehicle message appeals
and the new generation woman. J Advert Res
1993;33:3642.
18. Berney-Reddish IA, Areni CS. Eects of probability
markers on advertising claim acceptance. J Mark
Commun 2005;11:4154.
19. Wolin LD. Gender issues in advertising an oversight synthesis of research: 19702002. J Advert
Res 2003;43:111129.
20. Boyd TC, Shank MD. Athletes as product endorsers: the eect of gender and product relatedness. Sport Mark Q 2004;13:8293.
21. Peetz TB, Parks JB, Spencer NE. Sport heroes as
sport product endorsers: the role of gender in
the transfer of meaning process for selected undergraduate students. Sport Mark Q 2004;13:
141150.

Bhutada & Rollins / Research in Social and Administrative Pharmacy 11 (2015) 891900
22. Bhutada NS, Menon AM, Deshpande AD,
Perri M III. Impact of celebrity pitch in direct-toconsumer advertising of prescription drugs. Health
Mark Quart 2012;29:3548.
23. Limbu YB, Huhmann BA, Peterson RT. An
examination of humor and endorser eects on
consumers responses to direct-to-consumer
advertising: the moderating role of product
involvement. Int J Pharmaceut Healthc Mark
2012;6:2338.
24. Rollins BL, Bhutada NS. Impact of celebrity endorsements in disease-specic direct-to-consumer
(DTC) advertisements: an elaboration likelihood
model approach. Int J Pharmaceut Healthc Mark
2014;8:164177.
25. Rollins BL, King K, Zinkhan G, Perri M III. Behavioral intentions and information-seeking behavior:
a comparison of nonbranded versus branded
direct-to-consumer prescription advertisements.
Drug Inf J 2010;44:673683.
26. Rollins BL, King K, Zinkhan G, Perri M III. Nonbranded or branded direct-to-consumer prescription drug advertising-which is more eective?
Health Mark Quart 2011;28:8698.
27. Laczniak RN, Muehling DD. Toward a better understanding of the role of advertising message
involvement in ad processing. Psychol Mark 1993;
10:301319.
28. Laczniak RN, Muehling DD, Grossbart S. Manipulating message involvement in advertising
research. J Advert 1989;18:2838.
29. MacKenzie SB, Lutz RJ. Am empirical examination
of the structural antecedents of attitude toward the
ad in an advertising pretesting context. J Mark
1989;53:4865.
30. Laerty BA, Goldsmith RE. Corporate credibilitys
role in consumers attitudes and purchase intentions
when a high versus a low credibility endorser is used
in the ad. J Bus Res 1999;44:109116.
31. Laerty BA, Goldsmith RE, Newell SJ. The dual
credibility model: the inuence of corporate and
endorser credibility on attitudes and Purchase intentions. J Mark Theor Pract 2002;10:112.
32. Muehling DD, Laczniak RN. Advertisings immediate and delayed inuence on brand attitudes: considerations across message-involvement levels. J
Advert 1988;17:2334.
33. Bemmaor AC. Predicting behavior from intentionto-buy measures: the parametric case. J Mark Res
1995;32:176191.
34. Juster FT. Consumer buying intentions and Purchase probability: an Experiment in survey design.
J Am Stat Assoc 1966;61:658696.
35. Cohen J. Statistical Power Analysis for the Behavioral Sciences. ed. 2nd ed. Hillsdale, NJ: L. Erlbaum
Associates; 1988.

899

36. Kroenke K, Spitzer RL. The PHQ-9: a new depression diagnostic and severity measure. Psychiatr Ann
2002;32:509515.
37. Spitzer RL, Kroenke K, Williams JW. Validation
and utility of a self-report version of PRIME-MD:
the PHQ primary care study. JAMA 1999;10:
17371744.
38. Paul DP, Handlin A, Stanton ADA. Primary
care physicians attitudes toward direct-toconsumer advertising of prescription drugs: still
crazy after all these years. J Consum Mark
2002;19:564574.
39. Weissman JS, Blumenthal D, Silk AJ, et al. Physicians report on patient encounters involving
direct-to-consumer advertising. Health A (Millwood) 2004;23:219233.
40. Lipsky MS, Taylor CA. The opinions and experiences
of family physicians regarding direct-to-consumer
advertising. J Fam Pract 1997;45:495499.
41. J W. Attack Mounts on DTC Ads. Pharmaceut
Exec 2008;28:18.
42. Gilligan C. In a Dierent Voice: Psychological Theory and Womens Development. Cambridge, MA:
Harvard University Press; 1993.
43. Kahn C. How the media inuences womans perceptions of health care. Mark Health Serv 2001;
21:1217.
44. U.S. Department of Labor. General Facts on
Women and Job Based Health, http://www.dol.
gov/ebsa/newsroom/fshlth5.html; 2013. Accessed
14.11.14.
45. Wilkes MS, Bell RA, Kravitz RL. Direct-to-consumer prescription drug advertising: trends, impact,
and implications. Health A (Millwood) 2000;19:
110128.
46. Peyrot M, Alperstein NM, Van Doren D, Poli LG.
Direct-to-consumer ads can inuence behavior.
Advertising increases consumer knowledge and prescription drug requests. Mark Health Serv 1998;18:
2632.
47. Bell RA, Kravitz RL, Wilkes MS. Direct-to-consumer prescription drug advertising and the public.
J Gen Intern Med 1999;14:651657.
48. Huh J, Becker LB. Direct-to-consumer prescription
drug advertising: understanding its consequences.
Int J Advert 2005;24:441466.
49. Murray E, Lo B, Pollack L, Donelan K, Lee K.
Direct-to-consumer advertising: public perceptions
of its eects on health behaviors, health care, and
the doctor-patient relationship. J Am Board Fam
Pract 2004;17:618.
50. U. S. Census Bureau. Prole of General Population and Housing Characteristics: 2010, http://
factnder2.census.gov/faces/tableservices/jsf/pag
es/productview.xhtml?srcbkmk; 2010. Accessed
14.11.14.

900

Appendix I

Bhutada & Rollins / Research in Social and Administrative Pharmacy 11 (2015) 891900