Psychoneuroendocrinology (2015) 56, 179—189
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/psyneuen
REVIEW
Plasma adiponectin levels in schizophrenia and role of second-generation antipsychotics: A meta-analysis Francesco Bartoli a,∗, Annamaria Lax a, Cristina Crocamo a, Massimo Clerici a, Giuseppe Carrà b a
Department of Surgery and Translational Medicine, University of Milano Bicocca, Via Cadore 48, 20900 Monza, MB, Italy b Division of Psychiatry, Faculty of Brain Sciences, University College London, Charles Bell House, 67—73 Riding House Street, London W1W7EJ, UK Received 4 February 2015; received in revised form 6 March 2015; accepted 6 March 2015
KEYWORDS Adiponectin; Schizophrenia; Second-generation antipsychotics; Metabolic abnormalities; Weight gain; Cardiovascular diseases
∗
Abstract Background: People with schizophrenia are more likely than general population to suffer from metabolic abnormalities, with second-generation antipsychotics (SGAs) increasing the risk. Low plasma adiponectin levels may lead to metabolic dysregulations but evidence in people with schizophrenia, especially for the role of SGAs, is still inconclusive. Objective: To compare plasma adiponectin levels between people with schizophrenia and healthy controls, and to estimate the relative effect of schizophrenia and SGAs on adiponectin. Methods: We performed a systematic review and meta-analysis of observational studies published up to 13 June 2014 in main electronic databases. Pooled standardized mean differences (SMDs) between index and control groups were generated. Appropriate subanalyses and additional subgroup analyses were carried out. Results: Data from 2735 individuals, 1013 with and 1722 without schizophrenia, respectively, were analysed. Schizophrenia was not associated with lower adiponectin levels (SMD of −0.28, 95%CI: −0.59, 0.04; p = 0.09). However, individuals with schizophrenia taking SGAs had plasma levels significantly lower than controls (p = 0.002), which was not the case of drug free/drug naïve subjects (p = 0.52). As regards single antipsychotic drugs clozapine (p < 0.001) and olanzapine (p = 0.04) — but not risperidone (p = 0.88) — were associated with adiponectin levels lower than controls. Conclusions: People with schizophrenia per se may not have levels of adiponectin lower than controls, though treatment with SGAs is associated with this metabolic abnormality. This bears
Corresponding author. Tel.: +39 0392333982; fax: +39 0392332277. E-mail address: [email protected] (F. Bartoli).
http://dx.doi.org/10.1016/j.psyneuen.2015.03.012 0306-4530/© 2015 Elsevier Ltd. All rights reserved.
180
F. Bartoli et al. clinical significance because of hypoadiponectinemia involvement in cardiovascular diseases, even if mechanisms whereby SGAs affect adiponectin remain unexplained. Longitudinal studies evaluating long-term effects of SGAs on adiponectin are needed. © 2015 Elsevier Ltd. All rights reserved.
Contents 1. 2.
3.
4.
Introduction............................................................................................................ Methods................................................................................................................ 2.1. Search strategy ................................................................................................. 2.2. Eligibility criteria ............................................................................................... 2.3. Data collection .................................................................................................. 2.4. Quality assessment .............................................................................................. 2.5. Data analyses ................................................................................................... Results ................................................................................................................. 3.1. Study selection.................................................................................................. 3.2. Study characteristics ............................................................................................ 3.3. Adiponectin levels among people suffering from schizophrenia .................................................. Discussion .............................................................................................................. 4.1. Summary of findings............................................................................................. 4.2. Interpretations of findings....................................................................................... 4.3. Limitations ...................................................................................................... 4.4. Conclusions ..................................................................................................... Role of the funding source ............................................................................................. Conflict of interest statement .......................................................................................... Acknowledgments ...................................................................................................... Appendix A. Supplementary data..................................................................................... References .............................................................................................................
1. Introduction A large body of evidence shows that people who suffer from schizophrenia are more likely than the general population to have metabolic abnormalities, such as diabetes, dyslipidemia, abdominal obesity, and metabolic syndrome (e.g., Bartoli et al., 2013a; De Hert et al., 2011a; Mitchell et al., 2013; Osborn et al., 2008). A number of potential explanations of this association have been proposed. In part, the development of metabolic abnormalities is attributable to the individual-level detrimental health behaviours common in people with schizophrenia, including unhealthy dietary habits, excessive alcohol intake, smoking, and lack of exercise (Clerici et al., 2014; De Hert et al., 2011a). However, based on unmedicated clinical populations, it has also been argued that people with schizophrenia may have a specific susceptibility to metabolic abnormalities (e.g., Cohn et al., 2006; Ryan et al., 2003; Verma et al., 2009). Furthermore, antipsychotic medications, particularly second generation antipsychotics (SGAs) such as olanzapine and clozapine, seem to raise the risk of metabolic side effects such as hyperglycaemia, dyslipidemia, and weight gain (Correll et al., 2011; Das et al., 2012; Newcomer, 2005). Mechanisms underlying these metabolic abnormalities in people with schizophrenia are not fully understood so far, but in recent years it has been suggested that adipokines — biologically active cytokines secreted by adipose tissue — may play a major role, influencing energy and vascular homeostasis, and immune response (e.g., Klemettilä et al.,
180 181 181 181 181 181 182 182 182 182 183 183 183 185 186 187 187 187 187 187 187
2014; Kubota et al., 2006; Maury and Brichard, 2010; Song et al., 2013). In particular, adiponectin is an adipokine which has recently attracted considerable attention because of its interactions with both genetic and environmental (e.g., food intake and exercise) factors (Imbeault, 2007; Kadowaki et al., 2008; Okamoto et al., 2006). It is synthesized by white adipose tissue (Maeda et al., 1996) and circulates at relatively high concentrations, accounting for up to 0.05% of total serum protein (Lihn et al., 2005). Although its physiological function is not entirely clear, adiponectin is certainly involved in the modulation of glucose and lipid metabolism (Liu et al., 2012), with low plasma levels associated with obesity (Kadowaki et al., 2006), insulin resistance, and, in a dose—response manner, with type 2 diabetes as shown by a rather conclusive meta-analysis (Li et al., 2009). Furthermore, plasma adiponectin levels are negatively correlated with triglycerides concentration and waist circumference, possibly explaining the associations between hypoadiponectinemia, dyslipidemia and metabolic syndrome, respectively (Gardener et al., 2013; Matsuzawa et al., 2004; Ryo et al., 2004). Preliminary findings on adiponectin in people with schizophrenia as compared with healthy controls are contradictory (Jin et al., 2008), with research in drug free subjects showing both higher (Song et al., 2013) and lower (Cohn et al., 2006) plasma levels. In addition, atypical antipsychotics might influence adiponectin regulation independently of components of metabolic syndrome and gender (Hanssens et al., 2008), but also findings on their effect are
Plasma adiponectin levels in schizophrenia and role of second-generation antipsychotics: A meta-analysis inconclusive, with reports on both increased (Togo et al., 2004) and diminished (Richards et al., 2006) plasma levels of adiponectin in people with schizophrenia taking SGAs. Clarifying the inconsistencies in adiponectin alterations in people with schizophrenia is important for several reasons. First, in terms of pathophysiology, schizophrenia and metabolic abnormalities such as obesity may share an imbalance in adiponectin, perhaps involving also proinflammatory cytokines which play a role in the chronic, low-grade, inflammatory state detectable in both the conditions (Harwood, 2012; Leonard et al., 2012). On the other hand, the potentially crucial role of SGAs, along with other candidate factors which are likely to be correlated with adiponectin plasma levels such as age, gender and weight (Cohen et al., 2011; Obata et al., 2013; Okauchi et al., 2009; Saltevo et al., 2009; Santaniemi et al., 2006), needs to be elucidated. Personalized, preventive and treatment strategies aimed at avoiding severe metabolic disorders in people with schizophrenia need valid early markers. In addition, a more thorough knowledge of adiponectin plasma levels might pave the way also for a role of other adipokines. Although a body of evidence of acceptable size has accumulated in the last decade, previous narrative reviews (Jin et al., 2008; Scigliano and Ronchetti, 2013) reported not conclusive results. To the best of our knowledge, meta-analytic data comparing plasma adiponectin between people with schizophrenia and healthy subjects, in particular exploring the role of SGAs, are lacking. We conducted a systematic review and meta-analysis of relevant research to clarify adiponectin plasma levels in people with schizophrenia as compared with healthy controls. We also aimed at making clearer the role of SGAs, taking into account other known sociodemographic and clinical confounding factors. It is designed to include the most recent data and to overcome some of the limitations of studies to date. This is the first quantitative synthesis of the available evidence and, as such, it will allow for the assessment of strength and consistency of the relationship and for the identification of moderators of effect and publication bias.
2. Methods We conducted a systematic review and meta-analysis according to the meta-analysis of observational studies in epidemiology (MOOSE) guidelines (Stroup et al., 2000). The protocol was registered on PROSPERO (International prospective register of systematic reviews; registration and revision date: 18 July 2014; registration code: CRD42014010726).
2.1. Search strategy We systematically searched for papers published up to 13 June 2014 the following Electronic databases: PubMed; Embase; and PsycInfo (via ProQuest). No language restrictions were applied. Search phrases combined thesaurus terms related to adiponectin, schizophrenia, and SGAs. In addition, synonyms of thesaurus terms were used for free search in titles and abstracts. The search phrases used were as follows:
181
a. PubMed: (‘‘Adiponectin’’[Mesh] or adiponectin [title/abstract]) and (‘‘schizophrenia’’[Mesh] or schizo* [title/abstract] or ‘‘Antipsychotic Agents’’[Mesh] or antipsychotic* [title/abstract]); b. Embase: (’adiponectin’/exp OR adiponectin) AND (’schizophrenia’/exp OR schizophrenia); c. PsycInfo (via ProQuest): adiponectin and (SU.EXACT(‘‘Schizophrenia’’) OR schizo* OR SU.EXACT(‘‘Neuroleptic Drugs’’)). Additionally, we searched via web for non-indexed articles and grey literature on schizophrenia and adiponectin. References were managed using EndNote Web Software.
2.2. Eligibility criteria We included observational studies providing information on adiponectin plasma levels in both (a) an index group of adults suffering from schizophrenia or related disorders (i.e., schizoaffective, schizophreniform, delusional, and psychotic NOS disorders), and (b) a control group of healthy subjects. We excluded studies with participants’ mean age
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/psyneuen
REVIEW
Plasma adiponectin levels in schizophrenia and role of second-generation antipsychotics: A meta-analysis Francesco Bartoli a,∗, Annamaria Lax a, Cristina Crocamo a, Massimo Clerici a, Giuseppe Carrà b a
Department of Surgery and Translational Medicine, University of Milano Bicocca, Via Cadore 48, 20900 Monza, MB, Italy b Division of Psychiatry, Faculty of Brain Sciences, University College London, Charles Bell House, 67—73 Riding House Street, London W1W7EJ, UK Received 4 February 2015; received in revised form 6 March 2015; accepted 6 March 2015
KEYWORDS Adiponectin; Schizophrenia; Second-generation antipsychotics; Metabolic abnormalities; Weight gain; Cardiovascular diseases
∗
Abstract Background: People with schizophrenia are more likely than general population to suffer from metabolic abnormalities, with second-generation antipsychotics (SGAs) increasing the risk. Low plasma adiponectin levels may lead to metabolic dysregulations but evidence in people with schizophrenia, especially for the role of SGAs, is still inconclusive. Objective: To compare plasma adiponectin levels between people with schizophrenia and healthy controls, and to estimate the relative effect of schizophrenia and SGAs on adiponectin. Methods: We performed a systematic review and meta-analysis of observational studies published up to 13 June 2014 in main electronic databases. Pooled standardized mean differences (SMDs) between index and control groups were generated. Appropriate subanalyses and additional subgroup analyses were carried out. Results: Data from 2735 individuals, 1013 with and 1722 without schizophrenia, respectively, were analysed. Schizophrenia was not associated with lower adiponectin levels (SMD of −0.28, 95%CI: −0.59, 0.04; p = 0.09). However, individuals with schizophrenia taking SGAs had plasma levels significantly lower than controls (p = 0.002), which was not the case of drug free/drug naïve subjects (p = 0.52). As regards single antipsychotic drugs clozapine (p < 0.001) and olanzapine (p = 0.04) — but not risperidone (p = 0.88) — were associated with adiponectin levels lower than controls. Conclusions: People with schizophrenia per se may not have levels of adiponectin lower than controls, though treatment with SGAs is associated with this metabolic abnormality. This bears
Corresponding author. Tel.: +39 0392333982; fax: +39 0392332277. E-mail address: [email protected] (F. Bartoli).
http://dx.doi.org/10.1016/j.psyneuen.2015.03.012 0306-4530/© 2015 Elsevier Ltd. All rights reserved.
180
F. Bartoli et al. clinical significance because of hypoadiponectinemia involvement in cardiovascular diseases, even if mechanisms whereby SGAs affect adiponectin remain unexplained. Longitudinal studies evaluating long-term effects of SGAs on adiponectin are needed. © 2015 Elsevier Ltd. All rights reserved.
Contents 1. 2.
3.
4.
Introduction............................................................................................................ Methods................................................................................................................ 2.1. Search strategy ................................................................................................. 2.2. Eligibility criteria ............................................................................................... 2.3. Data collection .................................................................................................. 2.4. Quality assessment .............................................................................................. 2.5. Data analyses ................................................................................................... Results ................................................................................................................. 3.1. Study selection.................................................................................................. 3.2. Study characteristics ............................................................................................ 3.3. Adiponectin levels among people suffering from schizophrenia .................................................. Discussion .............................................................................................................. 4.1. Summary of findings............................................................................................. 4.2. Interpretations of findings....................................................................................... 4.3. Limitations ...................................................................................................... 4.4. Conclusions ..................................................................................................... Role of the funding source ............................................................................................. Conflict of interest statement .......................................................................................... Acknowledgments ...................................................................................................... Appendix A. Supplementary data..................................................................................... References .............................................................................................................
1. Introduction A large body of evidence shows that people who suffer from schizophrenia are more likely than the general population to have metabolic abnormalities, such as diabetes, dyslipidemia, abdominal obesity, and metabolic syndrome (e.g., Bartoli et al., 2013a; De Hert et al., 2011a; Mitchell et al., 2013; Osborn et al., 2008). A number of potential explanations of this association have been proposed. In part, the development of metabolic abnormalities is attributable to the individual-level detrimental health behaviours common in people with schizophrenia, including unhealthy dietary habits, excessive alcohol intake, smoking, and lack of exercise (Clerici et al., 2014; De Hert et al., 2011a). However, based on unmedicated clinical populations, it has also been argued that people with schizophrenia may have a specific susceptibility to metabolic abnormalities (e.g., Cohn et al., 2006; Ryan et al., 2003; Verma et al., 2009). Furthermore, antipsychotic medications, particularly second generation antipsychotics (SGAs) such as olanzapine and clozapine, seem to raise the risk of metabolic side effects such as hyperglycaemia, dyslipidemia, and weight gain (Correll et al., 2011; Das et al., 2012; Newcomer, 2005). Mechanisms underlying these metabolic abnormalities in people with schizophrenia are not fully understood so far, but in recent years it has been suggested that adipokines — biologically active cytokines secreted by adipose tissue — may play a major role, influencing energy and vascular homeostasis, and immune response (e.g., Klemettilä et al.,
180 181 181 181 181 181 182 182 182 182 183 183 183 185 186 187 187 187 187 187 187
2014; Kubota et al., 2006; Maury and Brichard, 2010; Song et al., 2013). In particular, adiponectin is an adipokine which has recently attracted considerable attention because of its interactions with both genetic and environmental (e.g., food intake and exercise) factors (Imbeault, 2007; Kadowaki et al., 2008; Okamoto et al., 2006). It is synthesized by white adipose tissue (Maeda et al., 1996) and circulates at relatively high concentrations, accounting for up to 0.05% of total serum protein (Lihn et al., 2005). Although its physiological function is not entirely clear, adiponectin is certainly involved in the modulation of glucose and lipid metabolism (Liu et al., 2012), with low plasma levels associated with obesity (Kadowaki et al., 2006), insulin resistance, and, in a dose—response manner, with type 2 diabetes as shown by a rather conclusive meta-analysis (Li et al., 2009). Furthermore, plasma adiponectin levels are negatively correlated with triglycerides concentration and waist circumference, possibly explaining the associations between hypoadiponectinemia, dyslipidemia and metabolic syndrome, respectively (Gardener et al., 2013; Matsuzawa et al., 2004; Ryo et al., 2004). Preliminary findings on adiponectin in people with schizophrenia as compared with healthy controls are contradictory (Jin et al., 2008), with research in drug free subjects showing both higher (Song et al., 2013) and lower (Cohn et al., 2006) plasma levels. In addition, atypical antipsychotics might influence adiponectin regulation independently of components of metabolic syndrome and gender (Hanssens et al., 2008), but also findings on their effect are
Plasma adiponectin levels in schizophrenia and role of second-generation antipsychotics: A meta-analysis inconclusive, with reports on both increased (Togo et al., 2004) and diminished (Richards et al., 2006) plasma levels of adiponectin in people with schizophrenia taking SGAs. Clarifying the inconsistencies in adiponectin alterations in people with schizophrenia is important for several reasons. First, in terms of pathophysiology, schizophrenia and metabolic abnormalities such as obesity may share an imbalance in adiponectin, perhaps involving also proinflammatory cytokines which play a role in the chronic, low-grade, inflammatory state detectable in both the conditions (Harwood, 2012; Leonard et al., 2012). On the other hand, the potentially crucial role of SGAs, along with other candidate factors which are likely to be correlated with adiponectin plasma levels such as age, gender and weight (Cohen et al., 2011; Obata et al., 2013; Okauchi et al., 2009; Saltevo et al., 2009; Santaniemi et al., 2006), needs to be elucidated. Personalized, preventive and treatment strategies aimed at avoiding severe metabolic disorders in people with schizophrenia need valid early markers. In addition, a more thorough knowledge of adiponectin plasma levels might pave the way also for a role of other adipokines. Although a body of evidence of acceptable size has accumulated in the last decade, previous narrative reviews (Jin et al., 2008; Scigliano and Ronchetti, 2013) reported not conclusive results. To the best of our knowledge, meta-analytic data comparing plasma adiponectin between people with schizophrenia and healthy subjects, in particular exploring the role of SGAs, are lacking. We conducted a systematic review and meta-analysis of relevant research to clarify adiponectin plasma levels in people with schizophrenia as compared with healthy controls. We also aimed at making clearer the role of SGAs, taking into account other known sociodemographic and clinical confounding factors. It is designed to include the most recent data and to overcome some of the limitations of studies to date. This is the first quantitative synthesis of the available evidence and, as such, it will allow for the assessment of strength and consistency of the relationship and for the identification of moderators of effect and publication bias.
2. Methods We conducted a systematic review and meta-analysis according to the meta-analysis of observational studies in epidemiology (MOOSE) guidelines (Stroup et al., 2000). The protocol was registered on PROSPERO (International prospective register of systematic reviews; registration and revision date: 18 July 2014; registration code: CRD42014010726).
2.1. Search strategy We systematically searched for papers published up to 13 June 2014 the following Electronic databases: PubMed; Embase; and PsycInfo (via ProQuest). No language restrictions were applied. Search phrases combined thesaurus terms related to adiponectin, schizophrenia, and SGAs. In addition, synonyms of thesaurus terms were used for free search in titles and abstracts. The search phrases used were as follows:
181
a. PubMed: (‘‘Adiponectin’’[Mesh] or adiponectin [title/abstract]) and (‘‘schizophrenia’’[Mesh] or schizo* [title/abstract] or ‘‘Antipsychotic Agents’’[Mesh] or antipsychotic* [title/abstract]); b. Embase: (’adiponectin’/exp OR adiponectin) AND (’schizophrenia’/exp OR schizophrenia); c. PsycInfo (via ProQuest): adiponectin and (SU.EXACT(‘‘Schizophrenia’’) OR schizo* OR SU.EXACT(‘‘Neuroleptic Drugs’’)). Additionally, we searched via web for non-indexed articles and grey literature on schizophrenia and adiponectin. References were managed using EndNote Web Software.
2.2. Eligibility criteria We included observational studies providing information on adiponectin plasma levels in both (a) an index group of adults suffering from schizophrenia or related disorders (i.e., schizoaffective, schizophreniform, delusional, and psychotic NOS disorders), and (b) a control group of healthy subjects. We excluded studies with participants’ mean age
