514 RISK INDICATORS FOR DEPRESSIVE SYMPTOMSsider tradeoffs between treatments that prolong life sis was part of a larger study on risk factors for oraland the associated quality of life that remains. How- epithelial dysplasia (OED), a precancerous condition,ever, in a 1999 review, Rogers, Fisher, and Woolgar3 and oral cancer. As a component of that study, weuncovered only 65 papers published between 1980 assessed psychological distress and depressive symp-and 1997 on quality of life in relation to head and toms among the study participants. We also obtainedneck cancer, and fewer than 10 papers were speciﬁ- information on factors potentially associated with ancally concerned with oral cancer. Since that review, increased or a decreased risk of depressive symptoms.head and neck cancer in general and, to a lesser We hypothesize that a primary risk indicator associ-extent, oral cancer have continued to receive atten- ated with depression is the diagnosis of a malignant,tion in relation to quality of life issues.4-11 rather than a premalignant, lesion. Additionally, based on the results of Mathieson et al20 in their study ofPsychosocial Impacts of Oral Cancer oral cancer patients and the general literature on social support,43 we hypothesize that social support Studies assessing the psychological status of oral moderates the impact of oral cancer and dysplasia oncancer and of head and neck cancer patients indicate depression such that those with high social supportthat the psychological impacts of the disease and its would have fewer depressive symptoms. Finally, wetreatment are substantial and that the prevalence of assess the effects of sociodemographic factors anddepressive symptoms and/or anxiety ranges from 19% socioeconomic status on risk for psychological dis-to 50% of those studied.4,7,12-26 Moreover, the impact tress.of oral cancer and its treatment persists over time,with Espie et al25 reporting that 22% to 32% of theirstudy participants were anxious or depressed 10 yearsafter diagnosis and treatment. Factors previously iden- Materials and Methodstiﬁed as contributing to an increased risk of psycho- SAMPLElogical distress among oral cancer patients3 include Incident cases of oral cancer and OED were iden-persistent pain, age (generally, younger patients more tiﬁed by reviewing pathology reports generated bythan older patients), gender (females more than oral pathology laboratories at the University of Floridamales), stage of cancer, type of treatment, and social College of Dentistry, the New Jersey Dental School,support.3 and the University of Connecticut School of Dental Studies of general quality of life and functional Medicine. These pathology laboratories serve primar-status since 1999 reinforce earlier ﬁndings of the ily community-based oral and maxillofacial surgeonsserious impact of oral cancer on these domains im- as well as general dentists and other dental specialists.mediately following surgery as well as over the long Patients were eligible for inclusion in the study if theyterm.13,27-40 Oral functions, such as chewing andspeech, were impaired. Social functioning also was were 30 to 79 years of age at the time of diagnosis,signiﬁcantly affected by surgery and pain in the reha- had been diagnosed with an invasive cancer or epi-bilitation period. Worse quality of life was associated thelial dysplasia of the oral cavity (excluding the lipwith larger tumor size, type of surgical treatment, and and major salivary glands, ie, ICD-0-3 C01.9 throughmore advanced stage of disease.28,31,33,34,40 C06.9), could provide reliable information, and could A limited number of studies41,42 have also assessed speak and read English.the psychological status of people undergoing biopsy Surgeons who submitted the biopsy specimen thatfor suspicious lesions and evaluated differences in gave rise to the identiﬁed pathology report were ap-psychological status after histopathologic diagnosis. proached to obtain permission to contact their pa-These studies show that participants have high levels tient regarding the study. After obtaining surgeonof depression and anxiety before biopsy and that after permission, study personnel contacted potential sub-the diagnosis is established, those with malignant le- jects, ﬁrst via mail and subsequently by telephone.sions are more depressed and anxious than those with Consenting individuals were interviewed over thenegative ﬁndings. However, these studies consist of telephone by trained interviewers who were blindedsmall samples and use various measures of psycholog- to both study hypotheses and each respondent’s di-ical status, and few, if any, analyze risk factors that agnostic status. Surrogate interviews were not con-might contribute to the development of symptoms of ducted, and subjects were compensated $20 for thepsychological distress. time required to participate. The study used a stan- We report ﬁndings from a study that measured dardized, structured questionnaire that, in addition toassociations between sociodemographic risk indica- obtaining demographic and environmental risk factortors and depression among individuals diagnosed information, included questions regarding depressionwith oral cancer or a premalignant lesion. The analy- and social support.
REISINE ET AL 515 VARIABLES BSNI questions were included in the study question- Depression was assessed using the Center for naire. Of these subjects, 15 were too ill to participateEpidemiologic Studies–Depression (CES-D) Scale,44 a and 50 had died before being contacted regarding thewidely used measure of depressive symptoms and study. An additional 50 subjects were not contactedpsychological distress that is utilized in community based on the wishes of the clinician who performedstudies and for samples of patients with chronic dis- the biopsy, and 110 subjects could not be reached byeases. The CES-D is a 20-item scale that assesses symp- letter or telephone. Among the individuals who couldtom frequency over the 2-week period preceding the be contacted by telephone, 79% (404 of 510) agreedinterview. Scores can range from 0 to 60; the average to participate in the study. Although all 404 subjectsscore for community samples is 9, and the average completed the environmental risk factor portion ofscore among patients institutionalized for depression the questionnaire, 3 subjects refused to answer allis 22. A score of 16 is considered indicative of questions related to the CES-D and were excludedclinical depression, although the CES-D is not a diag- from the following analysis. All interviews were com-nostic interview and cannot be used to diagnose clin- pleted during the period April 1996 throughical depression.44 In an early community-based study, December 2000.18% of whites and 26% of blacks had scores of 16 45; DESCRIPTIVE CHARACTERISTICS OF THE SAMPLEhowever, the reported prevalence is often higheramong people with chronic diseases.46 Table 1 presents the descriptive characteristics of The Berkman Social Network Inventory (BSNI)47 the study subjects. The majority had a diagnosis ofmeasures the level of social support. This social sup- dysplasia (58%). As expected, most of the subjectsport measure assesses the size of the social network were over the age of 50 and nearly one third wereand the degree of the individual’s integration within over the age of 70. There were more men thanthe network. The BSNI is based on family size, marital women in the sample, which was primarily white,status, number of memberships in organizations, and non-Hispanic. Study subjects also tended to be mar-number of social contacts in a week. It is categorized ried, not employed outside the home, have more thaninto 4 levels (BSNI score): low (0 or 1), medium (2 to a high school education, and report incomes of5), medium-high (6 or 7), and high (8 to 12). $30,000 or over. Subjects appeared to have good social networks as measured by the BSNI; most scor- ANALYSIS ing in the medium to high level, with only 24% re- Bivariate and multivariate analyses of the data were porting low social support. The majority of subjectsconducted using standard statistical methods. Odds were diagnosed at the University of Florida, and mostratios (ORs) and their 95% conﬁdence intervals (CIs) were interviewed within 15 months of diagnosis. Thewere obtained using unconditional logistic regression overall CES-D score was 8.2 (SE 0.48), slightlymodels in which CES-D score (dichotomized at 16, lower than that reported in community samples16 ) was the dependent variable of interest.48 To (CES-D 9).assist in model building, we used both forward and BIVARIATE ANALYSISbackward stepwise regression with levels for entryand exclusion set at P .15 and based on the likeli- The mean CES-D score was 9.9 (SD 11.2) for thehood ratio statistic. These program-driven stepwise cancer series and 7.1 (SD 8.1) for the OED series.regressions were conﬁrmed manually. As seen in Figure 1, the distribution of CES-D scores The study protocol was reviewed and approved by for the cancer group was shifted somewhat to thethe applicable institutional review boards. right relative to that of the OED cases (P .04 via Mann-Whitney). When we dichotomized CES-D scores at 9.0 (the mean score for community samples)Results or less and 10 or greater, cancer patients were more likely to score 10 or higher (35%) relative to persons RECRUITMENT OF SUBJECTS with a diagnosis of OED (25%; OR 1.61, 95% CI, A total of 301 surgeons were approached for per- 1.02 to 2.55; P .03). The mean CES-D score wasmission to contact their patients to recruit the sample similar for males (8.2, SD 10.0) and females (8.2, SDof subjects who completed the questionnaire that 9.2) (P .98).included the CES-D and BSNI questions. Of these In this analysis, CES-D scores of 16 were consid-surgeons, 282 (94%) could be contacted regarding the ered indicative of clinical depression. Of the 401study, and of these, 94% (265 of 282) were willing to respondents, 76 (19%) had scores of 16 and 37 (9%)assist with the project. had scores of 22 . Table 1 presents the characteris- A total of 735 potentially eligible subjects were tics of the study participants by depression status.identiﬁed during the time in which the CES-D and Although similar in terms of gender, race, employ-
516 RISK INDICATORS FOR DEPRESSIVE SYMPTOMS Table 1. CHARACTERISTICS OF STUDY SUBJECTS IN TOTAL AND BY CENTER FOR EPIDEMIOLOGIC STUDIES DEPRESSION SCALE (CES-D) SCORE DICHOTOMIZED AT <15 AND 16 (n 401) Total Sample, CES-D Score Variable n (%) 0 to 15 16 P ValueOral diagnosis Oral epithelial dysplasia 234 (58.4) 197 (60.6) 37 (48.7) Cancer 167 (41.6) 128 (39.4) 39 (51.3) .06*Age at interview (yr) 50 69 (17.2) 46 (14.2) 23 (30.3) 51 to 60 81 (20.2) 65 (20.0) 16 (21.1) 61 to 70 119 (29.7) 106 (32.6) 13 (17.1) 71 132 (32.9) 108 (33.2) 24 (31.6) .002* Mean (SD) 63.3 (11.9) 64.1 (11.4) 60.2 (13.3) .04†Gender Female 192 (47.9) 159 (48.9) 33 (43.4) Male 209 (52.1) 166 (51.1) 43 (56.6) .39*Race White, non-Hispanic 384 (95.8) 313 (96.3) 71 (93.4) Other 17 (4.2) 12 (3.4) 5 (6.6) .26*Currently Married No 135 (33.7) 100 (30.8) 35 (46.1) Yes 266 (66.3) 225 (69.2) 41 (53.9) .01*Employed outside the home No 256 (63.8) 203 (62.5) 53 (69.7) Yes 145 (36.2) 122 (37.5) 23 (30.3) .24*Education 12 yr 43 (10.7) 31 (9.5) 12 (15.8) 12 yr 111 (27.7) 88 (27.1) 23 (30.3) Some college 123 (30.7) 99 (30.5) 24 (31.6) At least college 124 (30.9) 107 (32.9) 17 (22.4) .19*Annual income $20,000 77 (21.2) 55 (18.8) 22 (31.4) $20,000 to $29,999 67 (18.5) 53 (18.1) 14 (20.0) $30,000 to $49,999 92 (25.3) 74 (25.3) 18 (25.7) $50,000 127 (35.0) 111 (37.9) 16 (22.9) .045*Berkman Social Network Inventory Low (0 or 1) 95 (23.9) 65 (20.2) 30 (39.5) Medium (2 to 5) 160 (40.2) 131 (40.7) 29 (38.2) Medium-High (6 or 7) 55 (13.8) 49 (15.2) 6 (7.9) High (8 ) 88 (22.1) 77 (23.9) 11 (14.5) .002*Participating institution University of Florida 204 (50.9) 164 (50.5) 40 (52.6) New Jersey Dental School 52 (13.0) 40 (12.3) 12 (15.8) University of Connecticut 145 (36.2) 121 (37.2) 24 (31.6) .56*Months between diagnosis and interview 12 121 (30.2) 96 (29.5) 25 (32.9) 13 to 15 189 (47.1) 153 (47.1) 36 (47.4) 16 91 (22.7) 76 (23.4) 15 (19.7) .74* Mean (SD) 13.4 (4.2) 13.5 (4.2) 13.0 (4.2) .31†CES-D score Mean (SD) 8.2 (9.6)NOTE. Percentages may not total to 100% due to rounding. Thirty-eight subjects refused to answer the question regarding income, and 3subjects did not answer questions related to the Berkman Social Network Inventory. *Based on Pearson 2 test. †Based on Mann-Whitney test.Reisine et al. Risk Indicators for Depressive Symptoms. J Oral Maxillofac Surg 2005.ment status, education, the laboratory making the score. Those who were younger, were not married,diagnosis, and months between the date of diagnosis had lower family incomes, and had lower BSNI scoresand interview, the distribution of clinical depression were more likely to score 16 on the CES-D com-status was statistically different in terms of age at pared with those who were older, were married, hadinterview, marital status, family income, and BSNI higher family incomes, and had higher scores on the
REISINE ET AL 517 30 The odds of CES-D scores indicative of depression 25 (16 ) were 79% higher among oral cancer patients Oral Cancer Oral Epithelial Dysplasia relative to those of subjects with a diagnosis of OED. 20 On the other hand, the odds of depression were% 15 reduced in persons over the age of 50 compared with those aged 50 years and under; in persons with 10 higher, relative to low, levels of social support; and in 5 persons who were employed outside the home com- pared with those who were not. Each of these ORs 0 was statistically signiﬁcant. BSNI was also ﬁtted as a 0-1 2-3 4-5 6-7 8-9 10-15 16-21 22+ dummy variable; however, the similarity in estimated CES-D Score coefﬁcients for the medium, high-medium, and highFIGURE 1. Center for Epidemiologic Studies–Depression Scale BSNI categories suggested that the variable be coded(CES-D) score by diagnostic status (n 401). dichotomously, with the lowest quartile remainingReisine et al. Risk Indicators for Depressive Symptoms. J Oral the referent category (as presented in Table 2). Al-Maxillofac Surg 2005. though not statistically signiﬁcant (P .07), CES-D scores indicative of clinical depression were moreBSNI. Although not statistically signiﬁcant at the .05 frequently observed among males than amonglevel, a higher proportion of cancer patients (23%) females (adjusted OR 1.73, 95% CI, 0.95 to 3.25).relative to OED patients (16%) were classiﬁed as de- No interactions between terms presented in Table 2pressed (OR 1.62, 95% CI, 0.95 to 2.76, P .06). were statistically signiﬁcant at the P .15 level. How- ever, adding an interaction term for oral diagnosis status MULTIVARIATE ANALYSIS (ie, cancer/OED) * age stratum to the main effects model Logistic regression was used to assess relationships (P .16) suggested that young adults with cancer hadbetween elevated CES-D scores and the variables in- the highest odds of depression (adjusted OR 3.7, 95%cluded in Table 1 while simultaneously controlling for CI, 1.2 to 11.6, relative to OED patients aged 50 years).each of the other independent variables in the modelas well as drinking (7 versus 7 drinks per week)and current smoking. Because a number of subjects Table 2. ADJUSTED ODDS RATIOS FOR SYMPTOMrefused to answer the question on annual income, we SEVERITY INDICATIVE OF CLINICAL DEPRESSIONcarried out separate analyses that either included or (CENTER FOR EPIDEMIOLOGIC STUDIES DEPRESSIONexcluded that variable. Forward and backward step- SCALE SCORE >16) ASSOCIATED WITH SOCIODEMOGRAPHIC VARIABLES AT THE FINALwise selection of variables resulted in the same ﬁnal STEP OF THE ANALYSISmodel regardless of whether annual income was in-cluded in the list of available independent variables. 95% Conﬁdence Variable ORadj* Interval Table 2 presents the independent variables associ-ated with depression that were included in the ﬁnal Oral diagnosismodel, along with their adjusted ORs and 95% CIs. Oral epithelial dysplasia 1.00We also ﬁtted a model that included all of the vari- Cancer 1.79 1.03 to 3.12 Age (yr)ables listed in Table 1 (oral diagnosis, age, gender, 50 1.00race, marital status, employment, education, income, 51-60 0.38 0.17 to 0.87BSNI, pathology laboratory, and months between di- 61-70 0.13 0.05 to 0.34agnosis and interview) as well as terms for drinking 71 0.24 0.10 to 0.59and current smoking. The further adjusted ORs were Berkman Social Network Inventory scorenot meaningfully different than those presented in Low 1.00Table 2. In addition, to further evaluate whether the Moderate 0.44 0.24 to 0.79time span between the dates of diagnosis and inter- Employed outside the homeview may have inﬂuenced our ﬁndings, we also con- No 1.00ducted separate analyses for persons interviewed Yes 0.28 0.13 to 0.59 Genderwithin 12-months of diagnosis and those interviewed Female 1.00more than 12 months after diagnosis. The general Male 1.73 0.95 to 3.15trends observed in Table 2 were also identiﬁed in *Odds ratios adjusted for each of the other variables in the tableeach of the stratiﬁed analyses; however, there was no as well as drinking (7 , 7 drinks per week) and current smokingassociation between gender and a CES-D score of 16 (yes, no).(OR 1.01) among those interviewed within 12 Reisine et al. Risk Indicators for Depressive Symptoms. J Oralmonths of diagnosis. Maxillofac Surg 2005.
518 RISK INDICATORS FOR DEPRESSIVE SYMPTOMSDiscussion cancer group relative to oral cancer patients as a whole. Had we included oral cancer cases identiﬁed via hospi- This study measured the association between socio- tal-based pathology laboratories, those patients may welldemographic risk indicators and depression in a sam- have reported higher levels of depression, thereby rais-ple of persons diagnosed with either OED or oral ing the mean CES-D score, the overall occurrence ofcancer. depression within the cancer series, and the OR for On average, the subjects in this study had lower depression among subjects with cancer relative to thosethan expected levels of depressive symptoms. Rela- with OED.tive to mean CES-D scores previously reported for Other ﬁndings were generally consistent with re-community samples ( 9), mean scores for subjects sults from previous studies investigating risk factorswith a diagnosis of oral cancer (9.9) were only slightly associated with elevated depression scores. Personselevated, whereas scores for persons with a diagnosis who were older, were employed outside the home,of OED (7.1) were actually lower. The proportion of and reported good social support were less likely tosubjects with CES-D scores of 16 (19%) was similar have elevated scores than were persons who wereto the prevalence observed in community studies, younger, were unemployed, and had poor social sup-while in the current study, the percentage of oral port. Surprisingly, however, we also found thatcancer patients with CES-D scores of 16 (23%) was among these individuals diagnosed with either oralsomewhat lower than the 26% recently reported for a cancer or OED, men were more likely to have CES-Dseries of 82 head and neck cancer cases treated in scores indicative of clinical depression than wereToronto.7 women after we adjusted for potential confounding. We hypothesized that persons diagnosed with oral Persons aged 50 and under were more likely thancancer were more likely to show signs of depression older individuals to have CES-D scores of 16 . In fact,than were persons diagnosed with a precancerous one third (23 of 69) of all patients aged 50 andlesion. Consistent with this hypothesis, our analyses younger were classiﬁed as depressed, a proportionrevealed that the distribution of CES-D scores for the notably higher than seen in the other age strata. Thecancer cases was shifted toward higher scores com- elevated CES-D scores in individuals aged 50 years andpared with those of the OED cases. Further, the odds younger persisted even after controlling for oral diag-of having a CES-D score of 16 were nearly 80% nostic status, employment, social support, gender,higher among the cancer cases compared with those drinking, and current smoking. These ﬁndings suggestcases diagnosed with OED after controlling for other that younger individuals carry a particularly high bur-covariates. The differences, however, were smaller den of depression among persons diagnosed with oralthan anticipated. precancerous or cancerous lesions. Moreover, our A number of factors may explain why we observed analysis of interaction terms further suggests that oralonly minimal differences in the occurrence of depres- cancer patients aged 50 and younger are at a notablysion between the OED and oral cancer cases. Our elevated risk of depression. Practitioners should becomparison group was composed of individuals with particularly cognizant of depressive symptoms ina precancerous lesion, and such persons would be these patients.expected to have a greater likelihood of depression Although social support and psychological well-than would persons without such lesions. This, in being have been studied extensively among cancerturn, would tend to attenuate group differences in patients in general, it has been far less frequentlyterms of depression. While this may have occurred, it investigated among people with oral cancer or thoseis notable that the OED cases in our sample had an with precancerous oral lesions. Hassanein and col-average CES-D score that was lower than previously leagues14 did ﬁnd a weak relationship between func-reported for community samples, and this would have tional status, social support, and satisfaction with so-enhanced any group difference in depression. cial support among oral cancer patients. They also Another possible explanation for the small difference reported that better functional scores were associatedin depressive symptoms between the 2 groups is that with higher levels of support and greater satisfaction.the OED and cancer cases included in this analysis were Various measures of social support have also beenall identiﬁed via oral pathology laboratories that serve inversely related to depression in head and neck pa-primarily community-based oral surgeons and general tients.5,20,49 In our study, social support, as measureddentists. Although we do not have information on stage by the BSNI, was inversely associated with a depres-at diagnosis, it is probable that the identiﬁed oral cancer sion score of 16 . Persons with little or no socialcases were more likely than those diagnosed in a hospi- support were most likely to have elevated CES-Dtal setting to be of an earlier stage, with more favorable scores, but even moderate social support was associ-prognoses and less traumatic treatment needs. This ated with a notable reduction in the odds of a CES-Dcould have minimized the level of depression in the score of 16 . Given the cross-sectional nature of the
REISINE ET AL 519data, however, it is not possible to ascribe a temporal 3. Rogers SN, Fisher SE, Woolgar JA: A review of quality of life assessment in oral cancer. Int J Oral Maxillofac Surg 28: 99,relationship to the observed association or to make 1999statements that social support leads to a reduced 4. Petruson KM, Silander EM, Hammerlid EB: Effects of psycho-likelihood of depression. It is possible, for example, social intervention on quality of life in patients with head andthat depression occurs ﬁrst and limits one’s ability to neck cancer. Head Neck 25:576, 2003 5. Kugaya A, Akechi T, Okamura H, et al: Correlates of depressedengage in social interaction. A similar limitation exists mood in ambulatory head and neck cancer patients. Psy-in terms of the observed link between employment chooncology 8:494, 1999and the risk of depression. It remains to be seen if 6. Kugaya A, Akechi T, Okuyama T, et al: Prevalence, predictive factors, and screening for psychologic distress in patients withclinical interventions to assist in the maintenance of newly diagnosed head and neck cancer. Cancer 88:2817, 2000social support structures, or in establishing new ones, 7. Katz MR, Irish JC, Devins GM, et al: Psychosocial adjustment incould improve the quality of life for oral cancer pa- head and neck cancer: The impact of disﬁgurement, gender and social support. Head Neck 25:103, 2003tients. In a recent report from Sweden, a psychosocial 8. Duffy SA, Terrell JE, Valenstein M, et al: Effect of smoking,support intervention program for head and neck can- alcohol, and depression on the quality of life of head and neckcer patients did not reduce the proportion of persons cancer patients. Gen Hosp Psychiatry 24:140, 2002 9. de Graeff A, de Leeuw JR, Ros WJ, et al: A prospective study onwith possible or probable depression as measured by quality of life of patients with cancer of the oral cavity orthe Hospital Anxiety and Depression Scale.4 oropharynx treated with surgery with or without radiotherapy. Gender differences in CES-D scores and the propor- Oral Oncol 35:27, 1999tion of persons with a score indicative of clinical 10. de Graeff A, de Leeuw JR, Ros WJ, et al: Pretreatment factors predicting quality of life after treatment for head and neckdepression were minimal before controlling for other cancer. Head Neck 22:398, 2002covariates. After adjusting for potential confounding, 11. Hammerlid E, Bjordal K, Ahlner-Elmqvist M, et al: A prospectivehowever, the odds of having a CES-D score indicative study of quality of life in head and neck cancer patients. Part I: At diagnosis. Laryngoscope 111:669, 2001of clinical depression was 70% higher among males 12. Freidlander P, Caruana S, Singh B, et al: Functional status afterrelative to females, although this difference was seen primary surgical therapy for squamous cell carcinoma of theonly among individuals interviewed more than base of the tongue. Head Neck 24:111, 2002 13. Rogers SN, Hannah L, Lowe D, et al: Quality of life 5-10 years12 months after diagnosis. By comparison, Katz et al7 after primary surgery for oral and oropharyngeal cancer. Jreported that, among head and neck cancer patients, Craniomaxillofac Surg 27:187, 1999women (40%) were more likely than were men (23%) 14. Hassanein KA, Musgrove BT, Bradbury E: Functional status ofto have CES-D scores of 16 . patients with oral cancer and its relation to style of coping, social support and psychological status. Br J Oral Maxillofac In our study, clinical depression was deﬁned as a Surg 39:340, 2001CES-D score of 16 . Persons with a diagnosis of oral 15. Bjordal K, Ahlner-Elmqvist M, Hammerlid E, et al: A prospectivecancer were more likely to have depression than were study of quality of life in head and neck cancer patients. Part II: Longitudinal data. Laryngoscope 111:1440, 2001OED patients. On the other hand, a protective asso- 16. Epstein JB, Emerton S, Kolbinson D, et al: Quality of life andciation against depression was identiﬁed for persons oral function following radiotherapy for head and neck cancer.over the age of 50 compared to those aged 50 years Head Neck 21:1, 1999 17. Allison PJ, Locker D, Wood-Dauphinee S, et al: Correlates ofand younger, as well as for persons having higher, health-related quality of life in upper aerodigestive tract cancerrelative to low, levels of social support and for per- patients. Qual Life Res 7:713, 1998sons employed outside the home compared with 18. Hammerlid E, Mercke C, Sullivan M, et al: A prospective quality of life study of patients with oral or pharyngeal carcinomathose who were not. Although not statistically signif- treated with external beam irradiation with or without brachy-icant, men were more likely than women to have therapy. Eur J Cancer 33:189, 1997CES-D scores indicative of clinical depression. 19. Chatuvedi SK, Shenoy A, Prasad KM, et al: Concerns, coping Oral and maxillofacial surgeons who diagnose and and quality of life in head and neck cancer patients. Support Care Cancer 4:186, 1996treat patients with oral cancer and oral dysplasia are 20. Mathieson CM, Logan-Smith LI, Phillips J, et al: Caring for headin a unique position to identify persons with an in- and neck oncology patients: Does social support lead to bettercreased risk of depression. The characteristics pre- quality of life? Can Fam Physician 42:1712, 1996 21. McDonough EM, Boyd JH, Varvares MA, et al: Relationshipsented here may assist the clinician in identifying between psychological status and compliance in a sample ofthose individuals with an elevated risk of concomitant patients treated for cancer of the head and neck. Head Neckdepressive symptoms and may have use for research- 18:487, 1996 22. Rapoport U, Kreitler S, Chaitchik S, et al: Psychosocial prob-ers seeking to better understand depression in these lems in head and neck cancer patients and their change withpatients. time since diagnosis. Ann Oncol 4:69, 1993 23. Telfer MR, Shepher JP: Psychological distress in patients attend- ing an oncology clinic after deﬁnitive treatment for maxillofa- cial malignant neoplasia. Int J Oral Maxillofac Surg 22:347,References 19931. de Haes JC, van Knippenberg FC: The quality of life of cancer 24. Gotay CC, Moore TD: Assessing quality of life in head and neck patients: A review of the literature. Soc Sci Med 20:809, 1985 cancer. Qual Life Res 1:5, 19922. Michael YL, Berkman LF, Colditz GA, et al: Social networks and 25. Espie CA, Freelander E, Campsie LM, et al: Psychological dis- health-related quality of life in breast cancer survivors: A pro- tress at follow-up after major surgery for intraoral cancer. spective study. J Psychosom Res 52:285, 2002 J Psychosom Res 33:441, 1989
520 RISK INDICATORS FOR DEPRESSIVE SYMPTOMS26. Strauss RP: Psychosocial responses to oral and maxillofacial 38. Epstein JB, Robertson M, Emerton S, et al: Quality of life and surgery for head and neck cancer. J Maxillofac Surg 1:343, oral function in patients treated with radiation therapy for head 1989 and neck cancer. Head Neck. 23:389, 200127. Morton RP, Izzard ME: Quality-of-life outcomes in head and 39. Gritz ER, Carmack CL, de Moor C, et al: First year after head and neck cancer patients. World J Surg 27:884, 2003 neck cancer: Quality of life. J Clin Oncol 17:352, 199928. Rogers SN, Devine J, Lowe D, et al: Longitudinal health-related 40. Konstantinovic VS: Quality of life after surgical excision fol- quality of life after mandibular resection for oral cancer: A lowed by radiotherapy for cancer of the tongue and ﬂoor of the comparison between rim and segment. Head Neck 26:54, 2004 mouth: Evaluation of 78 patients. J Craniomaxillofac Surg 27:29. Kwok HC, Morton RP, Chaplin JM, et al: Quality of life after 192, 1999 parotid and temporal bone surgery for cancer. Laryngoscope 41. Baile WF, Gibertini M, Scott L, et al: Pre-biopsy assessment of 112:820, 2002 patients with suspected head and neck cancer. J Psychosoc30. Perlmutter MA, Johnson JT, Snyderman CH, et al: Functional outcomes after treatment of squamous cell carcinoma of the Oncol 10:79-91, 1993 base of the tongue. Arch Otolaryngol Head Neck Surg 116:785, 42. Davies ADM, Davies C, Delpo MC: Depression and anxiety in 2002 patients undergoing diagnostic investigations for head and31. Rogers SN, Lowe D, Fisher SE, et al: Health-related quality of life neck cancers. Br J Psychiatry 149:491, 1986 and clinical function after primary surgery for oral cancer. Br J 43. Cohen S, Underwoood, LG, Gottleib BH: Social Support Mea- Oral Maxillofac Surg 40:11, 2002 surement and Intervention. A Guide for Health and Social32. Gellrich NC, Schimming R, Schramm A, et al: Pain, function, Scientists. Oxford, UK, Oxford University Press, 2000 and psychologic outcome before, during, and after intraoral 44. Radloff L: The CES-D scale: A self-report depression scale for tumor resection. J Oral Maxillofac Surg 60: 772, 2002 research in the general population. Appl Psychol Meas 1:385,33. Schliephake H, Jamil MU: Prospective evaluation of quality of 1977 life after oncologic surgery for oral cancer. Int J Oral Maxillofac 45. Comstock GW, Helsing KJ: Symptoms of depression in two Surg 31:427, 2002 communities. Psychol Med 6:551, 197634. Schliephake H, Jamil MU: Impact of intraoral soft-tissue recon- 46. McQuillan J, Fiﬁeld J, Sheehan TJ, et al: A comparison of struction on the development of quality of life after ablative self-reports of distress and affective disorder diagnoses in rheu- surgery in patients with oral cancer. Plast Reconstr Surg 109: matoid arthritis: A receiver operator characteristic analysis. 421, 2002 Arthritis Rheum 49:368, 200335. Robertson ML, Gleich LL, Barrett WL, et al: Base-of-tongue 47. Berkman LF, Syme SL: Social networks, host resistance, and cancer: survival, function, and quality of life after external- mortality: A nine year follow-up study of Alameda county beam irradiation and brachytherapy. Laryngoscope 111:1362, 2001 residents. Am J Public Health 109:186, 197936. Knezevic P, Uglesic V, Virag M, et al: Comparison of three 48. Hosmer DW, Lemeshow S: Applied Logistic Regression. New methods of oral reconstruction for quality-of-life assessment. York, NY, John Wiley & Sons, 2000 Ann Plast Surg 48:334, 2002 49. De Leeuw JR, De Graeff A, Ros WJ, et al: Negative and positive37. Klug C, Neuburg J, Glaser C, et al: Quality of life 2-10 years after inﬂuences of social support on depression in patients with combined treatment for advanced oral and oropharyngeal can- head and neck cancer: A prospective study. Psychooncology cer. Int J Oral Maxillofac Surg 31:664, 2002 9:20-28, 2002