This study assessed quality of life (QoL), depression, and anxiety before and after orthognathic surgery and identified risk factors for poorer postoperative outcome. This multicentre prospective study included 140 patients from five French medical centres. We assessed patients before surgery (T1), 3 months after surgery (T2), and 12 months after surgery (T3). We assessed the severity of the orofacial deformity, physical, psychological, social, and environmental QoL (WHOQOL-BREF), and depression and anxiety (GHQ-28). Risk factors for poorer outcome were identified using linear mixed models. Between baseline and 12 months, there was significant improvement in psychological and social QoL and in depression (although below the norms reported in the general population), but not in anxiety. Physical QoL was poorer in patients who were younger, who had a mild orofacial deformity, and who were depressed. Psychological QoL was poorer in younger patients and in depressed patients. Social QoL was poorer in patients who were single, who had a mild orofacial deformity, and who were depressed. Although orthognathic surgery provides a moderate improvement in psychological and social QoL, the systematic screening and treatment of depression could further improve QoL after surgery because it is a major predictor of poor QoL in this population.
Orofacial deformities are conditions associated with poor aesthetic and functional outcomes, as well as poor physical, psychological, and social quality of life (QoL). They can be treated effectively by orthognathic surgery, which can improve aesthetics, functional status, and QoL. There is a growing body of literature that focuses on the effects of orthognathic surgery on QoL ; pre- and post-surgery assessment of QoL can provide data that complement objective evaluations provided by clinicians. The concept of QoL, which is now used widely in health care and medicine, refers to the patient’s subjective evaluation of his/her health, and it enables many health dimensions to be assessed, contributing to a better understanding of the patient’s subjective perception of his/her health (e.g., physical, psychological, social, and environmental QoL). In patients with orofacial deformities, although it has been demonstrated that orthognathic surgery generally provides improvements in physical, psychological, and social QoL, postoperative outcomes in each of these dimensions may vary between patients. It is still unclear why some patients experience improved QoL while others do not.
To improve our ability to understand how QoL changes after surgery, some studies have aimed to identify the factors associated with changes in each QoL dimension (physical, psychological, social, and environmental QoL). This approach, which has been conducted successfully in patients undergoing other types of surgery, including bariatric surgery, surgery for cancer, and transplantation, might enable the early identification of patients at higher risk of poor postoperative outcomes. In a recent literature review about the QoL of patients undergoing orthognathic surgery, Soh and Narayanan identified a number of deficiencies in the literature, including many studies with small sample sizes, few prospective studies (only seven studies included more than 100 patients, and only two of these used a prospective design), and little use of standardized and validated questionnaires. In addition, only a few studies have investigated whether certain psychiatric comorbidities such as anxiety or depression could be associated with poor QoL in this population, and few studies have simultaneously assessed multiple variables (e.g., the severity of the orofacial deformity, type of orofacial deformity, type of treatment, individual characteristics, environmental factors, and depression). Since these variables may be confounding factors that could account for changes in QoL, an integrative approach is needed to determine whether they should be incorporated in the statistical analyses of future studies.
To improve our knowledge of the factors associated with QoL in patients undergoing orthognathic surgery, an integrative model that aims to determine the factors associated with each QoL dimension in this population is proposed. It is based on Wilson and Cleary’s QoL model, which is one of the models that is most widely used and referred to in the health-related QoL literature. In the working model presented here, focus is placed on three of the five levels of health outcome measures proposed by Wilson and Cleary, namely biological and physiological variables, symptom status, and functional status (the latter corresponds to health-related QoL). In this model, described in detail in Fig. 1 , it is hypothesized that each QoL dimension (physical, psychological, social, and environmental) would be associated with different biological and physiological variables (type of surgery, severity of orofacial deformity), variables related to symptom status (depression), individual characteristics (gender, age), and characteristics of the environment (marital status) ( Fig. 1 ). A similar model has already been tested in bariatric surgery patients. In the present study, this model was tested in a large sample of orthognathic patients ( N = 140), taking a longitudinal and prospective approach, with an integrative assessment of all the above-mentioned factors in order to control for potential confounding variables. As recommended by Soh and Narayanan, validated questionnaires were used: QoL was assessed with a widely used generic QoL questionnaire (the World Health Organization Quality of Life Questionnaire BREF version; WHOQOL-BREF), and anxiety and depression were assessed with a questionnaire that is used widely to screen for general psychiatric disorders in patients with medical disorders (the General Health Questionnaire, GHQ).
This study aimed to assess how QoL, depression, and anxiety change after orthognathic surgery and to test an integrative model in order to identify the risk factors for a poorer postoperative outcome (i.e., lower physical, psychological, social, and environmental QoL, higher depression, and higher anxiety). The relative impacts of age, gender, type of orofacial deformity, time since surgery, marital status, and depression on each QoL dimension were studied. Finally, we assessed whether changes in QoL, anxiety, or depression could be predicted by some of these variables.
Materials and methods
Participants and procedure
This multicentre prospective study was conducted in five public and private French medical centres. Patients who attended one of these centres between March 2010 and March 2011 and for whom orthognathic surgery was prescribed were included. Inclusion criteria were age ≥14 years and an orofacial deformity that had developed during childhood. Exclusion criteria were previous orthognathic surgery and orofacial deformity due to trauma, cancer, or malformation (e.g., cleft lip/palate). After inclusion, patients were assessed at baseline (1 month before surgery, T1), 3 months after surgery (T2), and 12 months after surgery (T3). At T1, patients answered the questionnaires shortly after their preoperative consultation with the surgeon; at T2 and at T3, they answered the questionnaires by mail. At each time point, patients completed self-administered questionnaires. The final population comprised 140 patients.
Marital status, age, QoL, depression, and anxiety were assessed at each time point (T1, T2, and T3). Baseline data also included gender and the severity of the orofacial deformity (mild, moderate, severe), assessed by the surgeon. Patients were considered to have a mild orofacial deformity if they had either a maxillary or a mandibular osteotomy, a moderate orofacial deformity if they had a maxillomandibular osteotomy without genioplasty, and a severe orofacial deformity if they had a maxillomandibular osteotomy with genioplasty.
The main outcome variable was QoL, assessed with the French version of the WHOQOL-BREF. This is a 26-item questionnaire exploring physical health (seven items, e.g. ‘To what extent do you feel physical pain prevents you from doing what you need to do?’), psychological health (six items, e.g. ‘How often do you have negative feelings such as despair, anxiety, depression?’), social health (three items, e.g. ‘How satisfied are you with the support you get from your friends?’), the environment (eight items, e.g. ‘How satisfied are you with your access to health services?’) and overall QoL and general health (two items). Participants had to answer each item on a 5-point Likert scale, which was then summed and transformed to a scale ranging from 0 to 100 for each score, a higher score indicating better QoL. The WHOQOL-BREF has good psychometric properties and provides a reliable assessment of generic health-related QoL. The WHOQOL-BREF is a shorter version of the original WHOQOL, which was created by the WHOQOL project initiated in 1991 in 15 international field centres and which aimed to develop an international cross-culturally comparable QoL assessment instrument. The WHOQOL is one of the two most widely used and validated questionnaires assessing generic health-related QoL (along with the 36-item Short Form Health Survey, SF-36).
Depression was assessed using the severe depression subscale of the 28-item version of the General Health Questionnaire (GHQ-28) and anxiety using the anxiety and insomnia subscale of the GHQ-28. The GHQ-28 is a self-administered questionnaire designed by Goldberg and Hillier, which has been used widely and is validated to screen for general psychiatric disorders in primary care and in patients with somatic disorders. It has already been used in patients undergoing oral and maxillofacial surgery, demonstrating good construct validity in this population, with good internal consistency and good convergent validity with measures of depression. In this study, the French version of the GHQ-28 (Likert scoring), validated by Pariente et al., was used. In line with the recommended thresholds, patients were considered to have significant depression when their depression sub-score at baseline was ≥6. The GHQ anxiety subscale was considered as a continuous variable.
Institutional review board approval was obtained for this study. It was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. Informed consent was also obtained from each patient to participate in this follow-up study.
Analyses were conducted using the R statistical package version 2.15.2 with the nlme package, which enables the use of linear and non-linear mixed effects models. Initial analyses involved descriptive statistics (means and standard deviations, numbers and percentages).
First, QoL, depression, and anxiety scores at baseline, 3 months, and 12 months were compared using linear mixed models adjusted only for time; in these initial analyses, time was considered as the only fixed effect; as random effects, there were intercepts for subjects, as well as by-time random slopes for the effect of the dependent variable. Physical, psychological, and social QoL scores at 12 months were compared with the French norms reported by Baumann et al. using the Student t -test. Next, to determine which factors were associated with each QoL dimension (physical, psychological, social, and environmental) and with depression and anxiety, linear mixed models were used to adjust the analyses on other variables (age, gender, time, type of orofacial deformity, marital status, and depression). As fixed effects, age, gender, time, type of orofacial deformity, marital status, and significant depression were entered; the only exception was that when depression was predicted, depression was not entered as a fixed effect because it was already the dependent variable. In this model, the type of orofacial deformity was considered as a dummy variable (moderate vs. mild deformity; severe vs. mild deformity). As random effects, there were intercepts for subjects, as well as by-time random slopes for the effect of the dependent variable. Interactions between time and each of the other variables entered as fixed effects was also tested to investigate whether some variables could be associated with changes in QoL over time. For a conceptual introduction to mixed effects modelling, the reader can refer to Winter. Linear mixed models have some advantages over traditional analysis of variance (ANOVA) models, including increased statistical power compared to univariate models, and the possibility of including both categorical and continuous predictors. Moreover, it is not necessary to make any assumptions about the structure of the residual variances and covariances. However, it was checked that the basic assumption for linear mixed models was met (for each analysis, visual inspection of residual plots did not reveal any obvious deviations from homoscedasticity or normality).
Of the 288 patients who were eligible for this study, 286 were included at baseline. Of these 286 patients, 279 produced fully exploitable questionnaires at T1, 210 were followed up at 3 months after surgery (with 206 fully exploitable questionnaires), and 152 were followed up at 12 months after surgery (with 140 fully exploitable questionnaires). The final sample was thus based on 140 patients who were successfully assessed at T1, T2, and T3 and who had fully exploitable questionnaires for all QoL dimensions.
Descriptive data and changes in QoL, depression, and anxiety after surgery
Table 1 presents the main characteristics of the study sample at the preoperative, 3-month, and 12-month visits. Between baseline and 12 months, there was significant improvement in psychological and social QoL, but a significant decrease in physical QoL ( Table 1 ). Patients reported a significant decrease in depression (0.9 points on a 28-point scale), but there was no change in anxiety. Patients reported a significant increase in environmental QoL between baseline and 3 months, but no change was observed between baseline and 12 months.
|Preoperative visit||3-Month visit||P -value||12-Month visit||P -value|
|Age, years||25.3 ± 10.7||25.7 ± 10.7||26.4 ± 10.7|
|Single||80 (57.1%)||81 (57.9%)||80 (57.1%)|
|Married or in a relationship||60 (42.9%)||59 (42.1%)||60 (42.9%)|
|Gender, female||100 (71.4%)||–||–|
|Type of orofacial deformity|
|Quality of life (WHOQOL-BREF)|
|Physical health||61.4 ± 12.8||58.3 ± 12.1||<0.01||59.2 ± 11.4||<0.05|
|Psychological health||59.5 ± 14.2||60.9 ± 13.7||0.19||62.1 ± 13.3||<0.05|
|Social health||60.8 ± 24.1||63.2 ± 23.8||0.22||66.7 ± 23.1||<0.01|
|Environmental health||74.4 ± 15.3||76.7 ± 14.5||<0.05||74.9 ± 14.3||0.64|
|Depression (GHQ-28)||4.2 ± 4.6||3.2 ± 4.4||<0.05||3.3 ± 4.4||<0.05|
|Significant depression||46 (32.9%)||32 (22.9%)||<0.05||32 (22.9%)||<0.05|
|Anxiety (GHQ-28)||6.1 ± 5.8||6.1 ± 5.8||0.99||6.0 ± 5.5||0.99|
a Descriptive data are presented as the mean ± standard deviation, or number (percentage). Quality of life, depression, and anxiety scores were compared between the preoperative visit and 3-month and 12-month visits using linear mixed models.
At 12 months, patients still had lower QoL scores than the norms reported in the general population ( P < 0.001) for physical QoL (59.5 vs. 81.4 for males; 59.1 vs. 79.4 for females), psychological QoL (62.6 vs. 69.3 for males; 61.8 vs. 66.6 for females), and social QoL (66.1 vs. 75.6 for males; 66.9 vs. 78 for females).
Factors associated with physical QoL and with changes in physical QoL following surgery ( Table 2 )
After adjustment, the factors associated with lower physical QoL were younger age (a mean decrease of 0.26 points for every year younger), mild orofacial deformity (when compared to moderate deformity), time (i.e., a decrease in physical QoL between baseline and 12 months), and significant depression (associated with a mean decrease of 4.75 points in the physical health score) ( Table 2 ). Patients who had a mild orofacial deformity tended to have poorer physical QoL than patients with a severe deformity ( P = 0.08), but there was no interaction between time and severity of the deformity (i.e., there was no difference in the change in physical QoL between different types of orofacial deformity). There was no interaction between time and gender, time and marital status, or time and depression (i.e., there was no difference in the change in physical QoL between males and females, between patients who were in a relationship and patients who were single, or between patients who were and were not depressed).
|Effect||Physical health||Psychological health||Social health|
|Estimate||P -value||Estimate||P -value||Estimate||P -value|
|Intercept||53.67 ± 2.95||<0.001||51.72 ± 3.55||<0.001||50.15 ± 5.83||<0.001|
|Age||0.26 ± 0.09||<0.01||0.32 ± 0.10||<0.01||0.16 ± 0.17||0.35|
|Gender, male||−0.17 ± 1.69||0.92||2.15 ± 2.10||0.31||−1.01 ± 3.42||0.77|
|T2 (versus T1)||−3.47 ± 1.05||<0.01||−0.09 ± 1.21||0.94||1.28 ± 1.98||0.52|
|T3 (versus T1)||−2.58 ± 1.17||<0.05||1.47 ± 1.14||0.20||3.85 ± 2.07||0.06|
|Type of orofacial deformity|
|Moderate||4.28 ± 2.15||<0.05||1.44 ± 2.67||0.59||6.52 ± 4.36||0.14|
|Severe||4.15 ± 2.34||0.08||2.55 ± 2.89||0.38||11.04 ± 4.72||<0.02|
|Marital status, married or in a relationship||−2.33 ± 1.61||0.15||−0.97 ± 1.69||0.57||10.71 ± 2.98||<0.001|
|Significant depression (GHQ-28)||−4.75 ± 1.35||<0.001||−5.30 ± 1.36||<0.001||−13.32 ± 2.40||<0.001|
a Results are based on linear mixed models adjusted for age, gender, time, severity of orofacial deformity, marital status, and depression. Data are the parameter estimates ± standard errors. T1, T2, and T3 are assessments at the preoperative visit, the 3-month postoperative visit, and the 12-month postoperative visit, respectively.
b Examples to clarify the results: (1) ‘Age’ shows a significant estimated effect of +0.26 for physical health, indicating that each increase in 1 year is associated with a significant mean increase of 0.26 points in the physical health score. (2) ‘T3 (versus T1)’ shows a significant estimated effect of −2.58 for physical health, indicating that patients at T3 reported a significant mean decrease of 2.58 points in the physical health score (compared to T0). (3) ‘Significant depression (GHQ-28)’ shows a significant estimated effect of −4.75 for physical health, indicating that significant depression is associated with a significant mean decrease of 4.75 points in the physical health score (compared to patients with no significant depression).
Factors associated with psychological QoL and with changes in psychological QoL following surgery ( Table 2 )
After adjustment, the factors associated with lower psychological QoL were younger age (a mean decrease of 0.32 points for each year younger) and significant depression (which was associated with a mean decrease of 5.30 points on the psychological health score) ( Table 2 ). Gender, type of orofacial deformity, time, and marital status were not associated with psychological QoL. There was no interaction between time and severity of the orofacial deformity, time and gender, time and marital status, or time and depression (i.e., there was no difference in the change in psychological QoL between different types of orofacial deformity, between males and females, between patients who were in a relationship and those who were single, or between patients who were and were not depressed).
Factors associated with social QoL and with changes in social QoL following surgery ( Table 2 )
After adjustment, the factors associated with lower social QoL were a mild orofacial deformity (i.e., patients with a mild orofacial deformity reported a mean decrease of 11.04 in social QoL compared to those with a severe deformity), marital status (single patients reported a mean decrease of 10.71 points in social QoL compared to patients who were in a relationship), and significant depression (associated with a mean decrease of 13.32 points on the social QoL score) ( Table 2 ). Patients reported a non-significant increase in social QoL between baseline and the 12-month visit ( P = 0.06). Age and gender were not associated with social QoL. There was no interaction between time and the type of orofacial deformity, time and gender, time and marital status, or time and depression (i.e., there was no difference in the change in social QoL between different types of orofacial deformity, between males and females, between patients who were in a relationship and patients who were single, or between patients who were and were not depressed).
Factors associated with environmental QoL and with changes in environmental QoL following surgery ( Table 3 )
After adjustment, the factors associated with lower environmental QoL were time (i.e., there was a transient increase in the environmental score between baseline and 3 months, but not between baseline and 12 months) and significant depression (associated with a mean decrease of 5.61 points in environmental QoL) ( Table 3 ). Age, gender, severity of orofacial deformity, and marital status were not associated with environmental QoL. There was no interaction between time and type of orofacial deformity, time and gender, time and marital status, or time and depression (i.e., there was no difference in the change in environmental QoL between different types of orofacial deformity, between males and females, between patients who were in a relationship and patients who were single, or between patients who were and were not depressed).