Clin Oral Invest DOI 10.1007/s00784-016-1883-3
ORIGINAL ARTICLE
Association between salivary flow rate and depressive symptoms with adjustment for genetic and family environmental factors in Japanese twin study Haruka Tanaka 1,2 & Soshiro Ogata 1,3 & Kazunori Ikebe 4 & Yuko Kurushima 4 & Kenichi Matsuda 4 & Kaori Enoki 4 & Kayoko Omura 2 & Chika Honda 5 & Osaka Twin Research Group & Kazuo Hayakawa 2
Received: 13 November 2015 / Accepted: 13 June 2016 # Springer-Verlag Berlin Heidelberg 2016
Abstract Objectives The association between salivary flow rate (SFR) and depressive symptoms have been inconclusive. The present study aimed to investigate the association between SFR and depressive symptoms with and without adjustment for genetic and family environmental factors. Materials and methods We conducted a cross-sectional study using twins and measured SFR and depressive symptoms as the outcome and explanatory variables, respectively. We also performed three-step regression analyses by first analysing the association between SFR and depressive symptoms without adjustment for genetic and family environmental factors (individual-level analyses). We then performed between–within analyses using monozygotic (MZ) and same-sex dizygotic (DZ) twin pairs, and finally using only MZ twin pairs. These between–within analyses estimated the coefficients adjusted * Haruka Tanaka
[email protected]
1
2
for genetic and family environmental factors. Furthermore, differences in the associations between individual-level and between–within suggest confounding by genetic and family environmental factors. Results We conducted 448 twins aged ≥20 years. In individuallevel analyses in males and between–within analyses using MZ and same-sex DZ male twin pairs, SFR associated with depressive symptoms. In between–within analyses using only MZ male twin pairs, SFR did not associate with depressive symptoms. In females, SFR did not associate with depressive symptoms in both individual-level and between–within analyses. Conclusions The present study revealed that the association between SFR and depressive symptoms was affected by common genetic factors in males. Clinical relevance Understanding this association between SFR and depressive symptoms with adjustment for genetic and family environmental factors could lead to an important consideration for the prevention and treatment of hyposalivation.
Osaka Twin Research Group
Keywords Depressive symptoms . Salivary flow rate . Hyposalivation . Twin study
Department of Health Promotion Science, Osaka University Graduate School of Medicine, 1-7 Yamadaoka, Suita, Osaka 565-0871, Japan
Introduction
Mie Prefectural College of Nursing, 1-1-1 Yumegaoka, Tsu, Mie 514-0116, Japan
3
Research Fellow of Japan Society for the Promotion of Science, Tokyo, Japan
4
Department of Prosthodontics, Gerodontology and Oral Rehabilitation, Osaka University Graduate School of Dentistry, 1-8 Yamadaoka, Suita, Osaka 565-0871, Japan
5
Center for Twin Research, Osaka University Graduate School of Medicine, 1-7 Yamadaoka, Suita, Osaka 565-0871, Japan
Saliva is mainly produced by the submandibular, parotid and sublingual glands. Saliva moistens the oral cavity and maintains oral health. We previously reported that hyposalivation is associated with a pathologically high candidal activity [1], and a low oral health-related quality of life [2]. Hyposalivation has also been associated with dental caries [3]. Therefore, it is important to maintain sufficient salivary flow rate to prevent oral diseases.
Clin Oral Invest
Depressed patients are prone to hyposalivation. However, the association between salivary flow rate and depressive symptoms remains unclear. Salivary flow rate in psychiatric patients has been studied since Strongin and Hinsie’s investigation in 1938 [4]. The association between salivary flow rate and depressive symptoms remains controversial; some studies have shown a significant association [5–7], while others have not [8–10]. We have shown that salivary flow rate is influenced by genetic factors [11]. Depressive symptoms have also been reported to be influenced by genetic factors [12]. Therefore, we hypothesised that genetic factors influence the association between salivary flow rate and depressive symptoms. Although the previous studies suggested that noradrenaline [13, 14] and stress [15, 16] influences the association between salivary flow rate and depressive symptoms, the mechanism of the association is still unclear. In particular, it is not unknown whether genetic factors are involved in the mechanism. A twin study can be adjusted for genetic and family environmental factors because monozygotic (MZ) twin pairs share almost all genetic and family environmental factors [17, 18]. The investigation of the association between salivary flow rate and depressive symptoms with adjustment for genetic and family environmental factors could lead to important consideration for prevention of hyposalivation. The present study aimed to investigate the associations between salivary flow rate and depressive symptoms with and without adjustment for genetic and family environmental factors using genetically informative twin samples.
analyses, which measured the concordance of DNA markers. Twin pairs with one discordant marker were considered DZ, and twin pairs with seven concordant markers were considered MZ. Measurements Salivary flow rate We used stimulated salivary flow rate as salivary flow rate. Stimulated whole saliva was collected by the mastication method [1, 2, 21]. Participants chewed a piece of paraffin wax (1 g, melting point 42 °C) (Orion Diagnostica, Espoo, Finland) for 2 min after swallowing all the saliva in their mouths. Thereafter, total saliva was collected in paper cups. The volume of stimulated whole saliva was calculated by converting the weight to volume, assuming a specific gravity of 1.0. The stimulated salivary flow rate (ml/min) was calculated by dividing the total saliva by two. Depressive symptoms
Materials and methods
Depressive symptoms were measured using the depression– dejection scale, which was adapted from the Japanese version of the profile of mood-states brief (POMS) [22]. The POMS measured mood states over a 7-day period. The depression– dejection scale was comprised of five self-report items. This Likert-type scale ranged from 0 (not at all) to 4 (extremely) points and the total score ranged from 0 to 20 points. High scores indicated high depressive symptoms. The depression– dejection scale has been reported to be associated with major depressive disorder [23].
Study design and participants
Possible confounders
This study was completed at the Osaka University Center for Twin Research founded 2009 in Japan [19]. We recruited twins using several methods, such as those using newspaper advertisements and posters in hospital and universities. The inclusion criteria for the present study were as follows: (a) twin pairs who participated in the cross-sectional, comprehensive medical examination between January 2009 and July 2014, (b) aged ≥20 years, (c) being either MZ twins, samesex dizygotic (DZ) twins or different-sex DZ twins and (d) twin pairs who participated in dental examination and questionnaire for depressive symptoms. We excluded twin pairs whose zygosity was unknown because they refused to participate in the genetic tests. Zygosity was determined using 15 loci from short tandem repeat (STR) markers [20]. A completely concordant twin pair with these STRs was considered as MZ. To assess whether DNA was identical, twin pairs underwent DNA microsatellite
To identify possible confounders, we collected the following information: (a) age, (b) number of teeth, (c) current smoking status and (d) current medication use. The reason we used the number of teeth and current smoking status as possible confounders was that they were associated with both with salivary flow rate [24, 25] and depressive symptoms [26, 27]. The teeth were counted by a dentist during a dental examination. Current smoking status was classified by responses to the following statements: (a) I smoke every day, (b) I do not smoke or (c) I do not smoke now, but I have smoked previously. Participants who responded positively to (a) were classified as current smokers. We classified current medication use regardless of the type of medication, as salivary flow rate is affected by numerous medications [28, 29]. In addition, current medication use has been associated with decreased salivary flow rate [8, 30]. For statistical analyses, age and number of teeth were considered numerical variables and
Clin Oral Invest
smoking status and medication use were considered categorical variables.
Statistical analyses To investigate the associations between stimulated salivary flow rate and depressive symptoms with and without adjustment for genetic and family environmental factors, we performed three-step regression analyses with individual-level analyses using twins, and between–within analyses using MZ and same-sex DZ twin pairs, and using only MZ twin pairs. We used stimulated salivary flow rate and depressive symptoms as the outcome and explanatory variables, respectively. First, we analysed the association between stimulated salivary flow rate and depressive symptoms using generalized linear mixed models (GLMMs) because GLMMs can adjust for the clustering of twin pairs [17]. Therefore, GLMMs analyses considered twins as individuals (individual-level analyses). Individual-level analyses did not adjust for genetic and family environmental factors; in these analyses, we analysed the associations with adjustment for age, number of teeth, current smoking status and current medication use. Second, we performed between–within analyses using MZ and same-sex DZ twin pairs for investigating the associations with adjustment for some genetic factors and almost all family environmental factors using the mean score of depressive symptoms within a twin pair [βB: between twin pairs coefficient], and difference in score between individual depressive symptoms and mean within a twin pair [βw: within twin pair coefficient] [18]. Finally, we performed between–within analyses using the only MZ pairs for controlling almost all genetic factors. Between–within analyses were used to adjust for genetic and family environmental factors. MZ twin pairs shared almost all genetic and family environmental factors, and samesex DZ twin pairs shared half of the genetic and almost all family environmental factors [17, 18]. βB represents the total effects of factors shared by a twin pair, such as genetic and family environmental factors [17, 18]. βw represents the total effects of factors independent of confounding genetic and family environmental factors [17, 18]. Differences in the association between depression coefficients among individuals and βw coefficients from between–within analyses of MZ and same-sex DZ pairs suggests that some genetic factors and almost all family environmental factors were confounding [17, 18]. Differences in the association between βw coefficients from between–within analyses of MZ and same-sex DZ twin pairs, and of MZ twin pairs only suggest that almost all genetic factors were confounding [17, 18]. Similar associations would indicate no confounding genetic and family environmental factors [17, 18].
We stratified all analyses by sex. Standard coefficients and 95 % confidence intervals (CIs) were obtained using GLMMs. The lme4 package version 1.1–5 [31] was used for GLMMs in the R statistics software version 3.0.3 [32]. Missing data was handled using list-wise case deletion; case deletion with a missing rate of 10 % or less was considered suitable [33].
Results Participants We evaluated 448 twins aged ≥20 years, which included 200 MZ twin pairs (292 females and 108 males), 19 same-sex DZ twin pairs (22 females and 16 males) and 5 different-sex DZ twin pairs. The overall rate for the missing data was 4.91 %, which included 3.13 % (n = 14) for stimulated salivary flow rate, 0.89 % (n = 4) for depressive symptoms and 0.89 % (n = 4) for current medication use. There were no missing data for age, the number of teeth, current smoking status and current medication use. The descriptive statistics are summarised in Table 1. In our study, there was only one female participant who took medication for Sjogren’s syndrome. Another female participant was diagnosed with xerostomia, but she was not taking medications for this. Table 2 shows current medication use for each disease. In our sample, only females had medication use for depression, osteoporosis and cancer. Medication use for hypertension was significantly different between male and female; however, medication use for other diseases did not significantly differ between sex. Individual-level analyses In individual-level analyses, we examined the association between stimulated salivary flow rate and depressive symptoms without adjustment for genetic and family environmental factors using the 448 participants. In females, depressive symptoms were not significantly associated with stimulated salivary flow rate adjusted for age, number of teeth, current smoking status and current medication use (Table 3). In males, depressive symptoms had a significant negative association with stimulated salivary flow rate adjusted for age, number of teeth, current smoking status and current medication use (Table 4). Furthermore, we investigated the association between salivary flow rate and depressive symptoms with adjustment for current medication use for each disease in individual-level analyses. The standardized coefficients of depressive symptoms on salivary flow rate ranged from −0.062 to −0.058 in female and all coefficients were not statistically significant (p > 0.24). In male, the standardized coefficients ranged from −0.27 to −0.24 and all coefficients were statistically significant (p < 0.003). In addition, we investigated the crude association between salivary flow rate
Clin Oral Invest Table 1 Characteristics of the participants
Variables
Female (n = 319)
Male (n = 129)
Age, year, mean (SD), range
49.21 (16.8), 21–88
63.46 (16.8), 21–87
1.69 (0.9), 0–5.5
1.75 (1.2), 0.05–5.7
2.45 (2.8), 0–12 25.51 (6.1), 0–32
2.64 (3.2), 0–20 21.57 (9.3), 0–32
Stimulated salivary flow rates: mg/min, mean (SD), range Depressive symptoms, mean (SD), range Total teeth, mean (SD), range Current smoking status, n (%) Current medication status, n (%)
12 (3.8)
24 (18.6)
132 (41.4)
56 (43.4)
SD standard deviation
and depressive symptoms. The association was significant in males [standardized coefficients: −0.25 (95 % CI −0.41 to −0.09)], but not significant in females [standardized coefficients: −0.03 (95 % CI −0.14 to 0.07)]. To investigate the significance of the sex-interaction effect, the main effect of depressive symptoms was not significant [standardized coefficients: −0.05 (95 % CI −0.16 to 0.06)]. However, sexinteraction effect (males) [standardized coefficients: −0.22 (95 % CI −0.40 to −0.05)] was significant. This result indicated that depressive symptoms were associated with salivary flow rate for males, but not for females. Between–within analyses In between–within analyses, we examined the associations with adjustment for genetic and family environmental factors. In between–within analyses using both MZ and same-sex DZ female twin pairs, and using only MZ female twin pairs, βw was not significantly associated with stimulated salivary flow rate (Table 3). In between–within analyses using MZ and same-sex DZ male twin pairs, βw was significantly associated with stimulated salivary flow rate adjusted for the number of Table 2 Characteristic of current medication use for each disease
teeth, current smoking status and current medication use (Table 4). In between–within analyses using only MZ male twin pairs, βw was not significantly associated with stimulated salivary flow rate adjusted for possible confounders (Table 4).
Discussion Our study aimed to investigate the associations between salivary flow rate and depressive symptoms with and without adjustment for genetic and family environmental factors using genetically informative twin samples. In male, the present study suggests confounding by genetic factors between depressive symptoms and salivary flow rate because the association between the coefficients of βB in between–within analyses using MZ and same-sex DZ twin pairs and using only MZ twin pairs were different. In females, depressive symptoms did not significantly associate with salivary flow rate both with and without adjustment for genetic and family environmental factors. Previous studies have shown that a low salivary flow rate is significantly associated with depressive symptoms [5–7]. However, other previous studies have reported that salivary
Current medication use
Females (n = 319)
Males (n = 129)
P valuesb Females/Males
Depression, n (%) Sleep disturbance, n (%) Diabetes, n (%) Hypertension, n (%) Allergiesa, n (%) Hyperlipidemia, n (%) Gastritis or Reflux esophagitis, n (%) Arrhythmias, n (%) Osteoporosis, n (%) Cancer, n (%) Joint disorder, n (%)
6 (1.88) 8 (2.51) 11 (3.45) 33 (10.35) 33 (10.35) 15 (4.70) 11 (3.45) 3 (0.94) 8 (2.51) 6 (1.88) 9 (2.82)
– 2 (1.55) 7 (5.43) 25 (19.38) 7 (5.43) 4 (3.10) 6 (4.65) 4 (3.10) – – 5 (3.88)
– 0.55 0.41 0.01 0.35 0.45 0.39 0.14 – – 0.57
We investigated medication use for each disease which had more than six current medication users a b
Allergies include Pollinosis and Urticaria
We examined whether or not current medication use for each disease differed between sex using generalized estimating equations (GEE)
Clin Oral Invest Table 3 Standardised coefficients for the association between stimulated salivary flow rate and depressive symptoms by individual-level and between– within analyses in females
Individual-level analyses Standardized coefficients
Between–within analyses (MZ + same-sex DZ) Standardized coefficients
(95 % CI) n = 319 Depressive symptoms βB βW Age Number of teeth Current smoking statusa Current medication useb
−0.06 (−0.17 to 0.04)
(95 % CI)
Between–within analyses (MZ) Standardized coefficients
n = 314
(95 % CI) n = 292
–
–
–
0.11 (−0.04 to 0.25)
0.07 (−0.08 to 0.22)
–
−0.07 (−0.15 to 0.005)
−0.07 (−0.15 to 0.01)
−0.28 (−0.45 to −0.12) 0.03 (−0.11 to 0.17) −0.12 (−0.77 to 0.54)
– 0.13 (0.01 to 0.26) 0.07 (−0.63 to 0.78)
– 0.14 (0.01 to 0.28) −0.08 (−0.85 to 0.70)
0.04 (−0.18 to 0.26)
−0.05 (−0.27 to 0.17)
−0.03 (−0.26 to 0.20)
MZ monozygotic, DZ dizygotic, CI confidence interval, βB mean score of depressive symptoms within a twin pair, βW difference in score between individual depressive symptoms and mean within a twin pair Age and sex were not entered in the between–within analyses because one of the twin pair shared sex and age with the co-twin of the twin pair a
Current smoker = 1, not current smoker = 0
b
Current medication user = 1, not current medication user = 0
flow rate is not significantly associated with depressive symptoms [8–10]; therefore, the association between salivary flow rate and depressive symptoms remains inconclusive. We hypothesised that genetic factors influence the association between salivary flow rate and depressive symptoms. Stimulated salivary flow rate [11] and depressive symptoms [12] are both Table 4 Standardized coefficients for the association between stimulated salivary flow rate and depressive symptoms by individual-level and between– within analyses in males
influenced by genetic factors. In the present study, we demonstrated that common genetic factors influenced the association between salivary flow rate and depressive symptoms in male twins. However, the complete association remains unclear because of the limited amount of research. Therefore, the present study warrants further studies.
Individual-level analyses Standardized coefficients (95 % CI) n = 129 Depressive symptoms βB βW Age Number of tooth Current smoking statusa Current medication useb
−0.26 (−0.42 to −0.10) – – −0.35 (−0.57 to −0.13) −0.01 (−0.21 to 0.19) −0.42 (−0.86 to 0.02) 0.23 (−0.08 to 0.55)
Between–within analyses (MZ + same-sex DZ) Standardized coefficients (95 % CI)
Between–within analyses (MZ) Standardized coefficients
n = 124
(95 % CI) n = 108
– −0.20 (−0.40 to 0.01) −0.13 (−0.26 to −0.0001) – 0.14 (−0.04 to 0.33) −0.33 (−0.79 to 0.13) 0.12 (−0.20 to 0.44)
– −0.23 (−0.46 to 0.004) −0.05 (−0.18 to 0.08) – 0.13 (−0.07 to 0.32) −0.33 (−0.86 to 0.21) 0.05 (−0.28 to 0.39)
MZ monozygotic, DZ dizygotic, CI confidence interval, βB mean score of depressive symptoms within a twin pair, βW difference in score between individual depressive symptoms and mean within a twin pair Age and sex were not entered in the between–within analyses because one of the twin pair shared sex and age with the co-twin of the twin pair a
Current smoker = 1, not current smoker = 0
b
Current medication user = 1, not current medication user = 0
Clin Oral Invest
Salivary flow rate was associated with depressive symptoms in males, but not in females before adjustment for genetic and family environmental factors. This sexrelated difference may be controlled by sex hormones; salivary flow rate and depressive symptoms have previously been associated with menopause and the menstrual cycle [34–36]. The influence of sex hormones was not evaluated in the present study. Several mechanisms could explain the association between salivary flow rate and depressive symptoms. First, a low salivary flow rate has been associated with a reduction in noradrenergic transmission [13] and deficit in brain noradrenaline has been associated with depression [14]. Here, we have shown that genetic factors affect the association between salivary flow rate and depressive symptoms. Therefore, common genetic factors may be associated with reduced noradrenergic transmission. Second, salivary flow rate decreases after stressful tasks [15], and depression has been associated with stressful life events [16]. Perceived stress is also affected by genetic factors [37], therefore the association between salivary flow rate and depressive symptoms may be induced by genetic factors associated with perceived stress. These mechanisms could not be examined in this cross-sectional epidemiological study. A strength of our study was the use of twins, which allowed adjustment for potentially confounding genetic and family environmental factors. However, there are also limitations to the present study. First, it was a crosssectional design, therefore the direction of causal associations remains unclear. Second, the sample size was small, particularly for analyses of male twin pairs. Although the statistical power was low because of the small sample size, the effects of depressive symptoms on salivary flow rate were statistically significant in individual-level analyses and between–within analyses using MZ and same-sex DZ twin pairs. Third, analyses were not adjusted for gene–environment interactions. However, almost all other genetic effects were successfully adjusted by analysing MZ twin pairs. Fourth, in our study, the mean of age was different between males and females, which led us to consider that further studies are necessary to clarify the sex differences for the association between depressive symptoms and salivary flow rate. Fifth, in this investigation, we were unable to classify medication use by drug name because of insufficient information. Further studies adjusting specific drug types as confounders are required. Finally, we did not measure unstimulated salivary flow rate, although this has been investigated previously [5–10]. In the future, the association between unstimulated salivary flow and depressive symptoms with adjustment for genetic and family environmental factors should be investigated.
Conclusions In conclusion, the present study revealed that common genetic factors can influence the association between salivary flow and depressive symptoms in males. Understanding this association with adjustment for genetic and family environmental factors could lead to important considerations for preventing and treating the hyposalivation. Acknowledgments We would like to thank all the participants of the study. We would like to acknowledge Kei Matsumaru’s contribution to the English revision. We also would like to thank Yoshinori Iwatani, Jun Hatazawa, Shiro Yorifuji and Mikio Watanabe who are members of Center for Twin Research, Osaka University Graduate School of Medicine. Compliance with ethical standards Conflict of interest The authors declare that they have no conflict of interest. Funding This study was supported by University Grants from Japanese Ministry of Education, Culture, Sports, Science and Technology (MEXT) and MEXT KAKENHI Grant Number 26293151. Ethical approval Participants provide written informed consent in the present study. The consent procedure and the present study were approved by the Institutional Review Board for Clinical Research at Osaka University Hospital and Research Ethics Committee of Osaka University. Informed consent We obtained informed consent from all participants.
References 1.
2.
3.
4. 5. 6.
7.
8.
Ikebe K, Morii K, Matsuda K et al (2006) Association of candidal activity with denture use and salivary flow in symptom-free adults over 60 years. J Oral Rehabil 33:36–42 Ikebe K, Matsuda K, Morii K et al (2007) Impact of dry mouth and hyposalivation on oral health-related quality of life of elderly Japanese. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 103:216–222. doi:10.1016/j.tripleo.2005.12.001 Cunha-Cruz J, Scott J, Rothen M, et al (2013) Salivary characteristics and dental caries: evidence from general dental practices. J Am Dent Assoc 144(5):e31–e40 Strongin EI, Hinsie LE (1938) Parotid gland secretions in manicdepressive patients. Am J Psychiatry 94:1459–1466 Bolwig TG, Rafaelsen OJ (1972) Salivation in affective disorders. Psychol Med 2:232–238 Takiguchi T, Yoshihara A, Takano N, Miyazaki H (2015) Oral health and depression in older Japanese people. Gerodontology. doi:10.1111/ger.12177 Bergdahl M, Bergdahl J (2000) Low unstimulated salivary flow and subjective oral dryness: association with medication, anxiety, depression, and stress. J Dent Res 79:1652–1658 Hugo FN, Hilgert JB, Corso S et al (2008) Association of chronic stress, depression symptoms and cortisol with low saliva flow in a sample of south-Brazilians aged 50 years and older. Gerodontology 25:18–25. doi:10.1111/j.1741-2358.2007.00188.x
Clin Oral Invest 9.
10.
11.
12.
13.
14.
15.
16. 17.
18.
19.
20. 21.
22.
Ohara Y, Hirano H, Yoshida H et al (2013) Prevalence and factors associated with xerostomia and hyposalivation among community-dwelling older people in Japan. Gerodontology. doi:10.1111/ger.12101 Bergdahl J, Bergdahl M (2001) Environmental illness: evaluation of salivary flow, symptoms, diseases, medications, and psychological factors. Acta Odontol Scand 59:104–110 Kurushima Y, Ikebe K, Matsuda K et al (2015) Influence of genetic and environmental factors on oral diseases and function in aged twins. J Oral Rehabil 42:49–56. doi:10.1111/joor.12228 Sullivan PF, Neale MC, Kendler KS (2000) Genetic epidemiology of major depression: review and meta-analysis. Am J Psychiatr 157: 1552–1562. doi:10.1176/appi.ajp.157.10.1552 Vatta MS, Hope SI, Prendes GM et al (2002) Salivary glands and noradrenergic transmission in diabetic rats. Auton Autocoid Pharmacol 22:65–71 Lambert G, Johansson M, Agren H, Friberg P (2000) Reduced brain norepinephrine and dopamine release in treatment-refractory depressive illness: evidence in support of the catecholamine hypothesis of mood disorders. Arch Gen Psychiatry 57:787–793 Toda M, Ichikawa H (2012) Effect of laughter on salivary flow rates and levels of chromogranin a in young adults and elderly people. Environ Health Prev Med 17:494–499. doi:10.1007/s12199-012-0279-5 Tennant C (2002) Life events, stress and depression: a review of recent findings. A N Z J Psychiat 36:173–182 Carlin JB, Gurrin LC, Sterne JA et al (2005) Regression models for twin studies: a critical review. Int J Epidemiol 34:1089–1099. doi:10.1093/ije/dyi153 Dwyer T, Blizzard L (2005) A discussion of some statistical methods for separating within-pair associations from associations among all twins in research on fetal origins of disease. Paediatr Perinat Epidemiol 19(Suppl 1):48–53. doi:10.1111/j.13653016.2005.00615.x Hayakawa K, Iwatani Y (2013) An overview of multidisciplinary research resources at the Osaka University Center for twin research. Twin Res Hum Genet : Off J Int Soc Twin Stud 16:217–220. doi:10.1017/thg.2012.141 Krenke BE, Tereba A, Anderson SJ et al (2002) Validation of a 16locus fluorescent multiplex system. J Forensic Sci 47:773–785 Ikebe K, Sajima H, Kobayashi S et al (2002) Association of salivary flow rate with oral function in a sample of community-dwelling older adults in Japan. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 94:184–190 Bougoeis A, Unes Le A, Michael M (2010) Full-sacle and shortform of profile of mood states: a factor analytic comparison. Journal of Sport Behavior 33(4):355–375
23.
Patterson K, Young C, Woods SP et al (2006) Screening for major depression in persons with HIV infection: the concurrent predictive validity of the profile of mood states depression-dejection scale. Int J Methods Psychiatr Res 15:75–82. doi:10.1002/mpr.184 24. Bergdahl M (2000) Salivary flow and oral complaints in adult dental patients. Community Dent Oral Epidemiol 28:59–66 25. Rad M, Kakoie S, Niliye Brojeni F, Pourdamghan N (2010) Effect of long-term smoking on whole-mouth salivary flow rate and oral health. J Dent Res Dent Clin Dent Prosp 4:110–114. doi:10.5681/joddd.2010.028 26. Wiener RC, Wiener MA, McNeil DW (2015) Comorbid depression/anxiety and teeth removed: behavioral risk factor surveillance system 2010. Community Dent Oral Epidemiol 43:433– 443. doi:10.1111/cdoe.12168 27. Luger TM, Suls J, Vander Weg MW (2014) How robust is the association between smoking and depression in adults? A metaanalysis using linear mixed-effects models. Addict Behav 39: 1418–1429. doi:10.1016/j.addbeh.2014.05.011 28. Sreebny LM, Schwartz SS (1986) A reference guide to drugs and dry mouth. Gerodontology 5:75–99 29. Saleh J, Figueiredo MAZ, Cherubini K, Salum FG (2015) Salivary hypofunction: an update on aetiology, diagnosis and therapeutics. Arch Oral Biol 60:242–255. doi:10.1016/j.archoralbio.2014.10.004 30. Wu AJ, Ship JA (1993) A characterization of major salivary gland flow rates in the presence of medications and systemic diseases. Oral Surg Oral Med Oral Pathol 76:301–306 31. Bates D, Maechler M, Bolker B, Walker S (2014) lme4: Linear mixed-effects models using Eigen and S4 32. R Core Team (2014) R: a language and environment for statistical computing 33. Marshall A, Altman DG, Holder RL (2010) Comparison of imputation methods for handling missing covariate data when fitting a cox proportional hazards model: a resampling study. BMC Med Res Methodol 10:112. doi:10.1186/1471-2288-10-112 34. Calil CM, Lima PO, Bernardes CF et al (2008) Influence of gender and menstrual cycle on volatile sulphur compounds production. Arch Oral Biol 53:1107–1112. doi:10.1016/j.archoralbio.2008.06.008 35. Minicucci EM, Pires RBC, Vieira RA et al (2013) Assessing the impact of menopause on salivary flow and xerostomia. Aust Dent J 58:230–234. doi:10.1111/adj.12057 36. Altemus M, Sarvaiya N, Neill Epperson C (2014) Sex differences in a n x i e t y a n d d e p re s s i o n c l i n i c a l p er s p e c t i v e s . F r o n t Neuroendocrinol 35:320–330. doi:10.1016/j.yfrne.2014.05.004 37. Federenko IS, Schlotz W, Kirschbaum C et al (2006) The heritability of perceived stress. Psychol Med 36:375–385. doi:10.1017/S0033291705006616