Metabolic syndrome and insulin resistance in schizophrenia patients receiving antipsychotics genotyped for the methylenetetrahydrofolate reductase (MTHFR) 677C/T and 1298A/C variants

1. Introduction 

The metabolic syndrome has been associated with significant cardiovascular mortality (Maggi et al., 2006, Onat et al., 2006), and the growing association between atypical antipsychotic (AAPs) use and this syndrome represents a particular problem for psychiatric populations treated with these medications. Although several definitions for the metabolic syndrome have been cited, the National Cholesterol Education Program Adult Treatment Protocol (NCEP ATP III) is most often used and requires at least three of the following: abdominal obesity, elevated triglycerides, low HDL, elevated blood pressure, or elevated fasting glucose (Lee et al., 2001). AAPs, primarily used for the treatment of schizophrenia, have been associated with significant metabolic complications, including hyperlipidemia (Gaulin et al., 1999, Osser et al., 1999, Spivak et al., 1999, Henderson et al., 2000, Huang and Chen, 2005), insulin resistance and diabetes mellitus (Henderson et al., 2000, Gianfrancesco et al., 2003, Ollendorf et al., 2004, Carlson et al., 2006, Henderson et al., 2005, Lambert et al., 2005, Sernyak et al., 2005, Guo et al., 2006), obesity (Henderson et al., 2000, Volavka et al., 2002, Simpson, 2005, Zipursky et al., 2005), and hyperhomocysteinemia (Levine et al., 2002, Applebaum et al., 2004, Muntjewerff and Blom, 2005, Muntjewerff et al., 2005). These metabolic complications result in a two-to-four-fold increase in the rate of metabolic syndrome in patients with schizophrenia (Kato et al., 2004, McEvoy et al., 2005, Nasrallah, 2006). The presence of the metabolic syndrome itself is associated with an increased risk for coronary heart disease, CVD, and diabetes mellitus (Wilson et al., 2005a, Wilson et al., 2005b), and patients with schizophrenia treated with AAPs are at greater risk for vascular disease and have a three fold increase in sudden cardiac death compared to the general population (Saari et al., 2005).

Because APPs such as clozapine and olanzapine may provide superior clinical benefits (Lieberman et al., 2005, McEvoy et al., 2006) there is a pressing need to identify patients at risk for the metabolic syndrome and the associated complications. Among the risk factors associated with the development of CVD, insulin resistance, and diabetes mellitus in the general population are the 677C/T and 1298A/C genetic variants of methylenetetrahydrofolate reductase (MTHFR) which are involved in folate and homocysteine metabolism (Klerk et al., 2002, Lewis et al., 2005). These polymorphisms have been associated with up to a 70% reduction in folate acid metabolism, hyperhomocysteinemia, and a greater risk for CVD (Klerk et al., 2002, Matthews, 2002, Gueant-Rodriguez et al., 2005). The relationship between these variants and risk for the metabolic syndrome or insulin resistance in the schizophrenia population receiving AAPs has not been previously investigated, as to our knowledge; our research group is the first to report an association.

2. Methods and materials 

2.3. Determination of MTHFR 677C/T and 1298A/C variants 

Genomic DNA was isolated from whole blood with the salt precipitation method. Genotyping was done with Pyrosequencing™ Technology. Polymerase Chain Reaction (PCR) and pyrosequencing primers were designed using Pyrosequencing single nucleotide polymorphism (SNP) Primer Design Version 1.01 software (http://www.biotage.com). Pyrosequencing was performed and analyzed as previously described using a PSQ MD instrument and software (Pyrosequencing AB, Uppsala, Sweden).

The MTHFR 677C/T and 1298A/C variants previously described (Frosst et al., 1995, van der Put et al., 1998) were analyzed for their relationship to metabolic syndrome and HOMA-IR value in subjects with schizophrenia receiving antipsychotic treatment for at least 12 months. The sequence was accessed from GENBANK (NM 005957). Forty five PCR cycles were done for reactions in a 30 μl volume with 1.5 mM Mg2+ and 10 pmol of each primer according to the following specifications: 94 °C for 30 s, 56 °C for 30 s and, 72 °C×30 s. A 168 bp (677C/T) and 85 bp (1298A/C) PCR products were visualized by electrophoresis on 1.5% agarose gels stained with ethidium bromide prior to Pyrosequencing. Table 1, lists the primers used for genotyping of the 677C/T and 1298A/C variants.

Table 1. Primers used for genotyping MTHFR 677C/T and 1298A/C variants

	MTHFR 677C/T variant
	Forward primer	5′-Biotin/ACTGTCATCCCTATTGGCAGGTTA-3′
	Reverse primer	5′-TCGGTGCATGCCTTCACAA-3′
	Pyrosequencing primer	5′-AGGAGCTGACCAGTGA-3′
	MTHFR 1298A/C variant
	Forward primer	5′-GGAGCTGCTGAAGATGTGG-3′
	Reverse primer	5′-Biotin/TGGTTCTCCCGAGAGGTAAAG-3′
	Pyrosequencing primer	5′-AGGAGCTGACCAGTGA-3′

3. Results 

A total of fifty-nine subjects were recruited for this study, with one opting to not provide a DNA sample which resulted in a final sample size of fifty-eight (males=38 (66%), females=20 (34%)). There were no differences in age, gender, race, smoking, or psychopathology between the genotype groups. Table 2 provides details regarding the baseline characteristics of this population by 677 genotype group.

Table 2. Genotype groups characteristics

		MTHFR genotype		p-value
		CC (n=26)	T allele carriers (n=32)
	Age (±S.D.) in years [age range in years]	36.7±10.5 [23–57]	36.2±9.6 [20–52]	0.85
	Gender F/M (% male)	11 female/15 male (58%)	9 female/ 23 male (72%)	0.26
	Race (% Caucasian)	81%	96%	0.10
	BMI (±S.D.)	30.83±9.47	30.16±5.51	0.73
	% receiving AAPs	92%	84%	0.20
	Cigarette pack year history (±S.D.)	13.07±17.68	11.69±17.07	0.77
	Minutes of physical activity per week (±S.D.)	82.8±114.97	217.44±324.81	0.05
	% meeting metabolic syndrome criteria	23%	53%	0.02
	Fasting blood glucose in mg/dl (±S.D.)	113.42±14.7	113.65±29.89	0.97
	Fasting total cholesterol in mg/dl (±S.D.)	171.08±42.41	186.55±40.87	0.17
	Fasting triglycerides in mg/dl	119.17±62.13	164.45±98.77	0.06
	Fasting HDL in mg/dl	52.46±21.11	48.03±12.12	0.33
	Blood pressure in mmHg (±S.D.)	123/72±14/9	119/73 ±15/12	0.39
	Waist circumference in cm (±S.D.)	103.9±20.7	105.88±13.15	0.66
	HOMA-IR (±S.D.)	3.79±3.37	5.44±6.41	0.24
	Brief Psychiatric Rating Scale (BPRS) Score (±S.D.)	39.35±8.95	38.91±11.82	0.88
	Scale for Assessment of Negative Symptoms (SANS) Score (±S.D.)	13.31±7.45	11.31±7.78	0.33

At the time of screening, subjects were receiving a variety of antipsychotic medications with 42 on monotherapy (n=11 clozapine, n=8 olanzapine, n=8 risperidone, n=6 quetiapine, n=4 haloperidol, n=3 aripiprazole, and n=2 ziprasidone) and 16 on polytherapy (n=6 two AAPs, n=6 one AAP and one typical antipsychotic, n=2 ziprasidone and one typical antipsychotic, and n=2 aripiprazole and one AAP). Overall, 81% (47/58) were currently receiving an AAP associated with DM and weight gain, although all subjects had significant past AAP exposure.

The distribution of 677 and 1298 genotypes was as follows CC (44%), CT (47%), TT (9%), AA (47%), AC (46%), CC (7%), both of which were in Hardy Weinberg equilibrium (677: χ2=0.296, p=0.58, 1298: χ2=0.04, p=0.46). We found high degree of linkage between these two SNPS, which has been reported previously (D'=1) (Martin et al., 2006).

Twenty-three subjects met criteria for metabolic syndrome, giving a prevalence rate of 40%, which is similar to national estimates of 30–50% in schizophrenia (McEvoy et al., 2005, Nasrallah, 2006). For the 677 T allele carriers (male=23, female=9) the incidence of metabolic syndrome was 53%, compared to the CC genotype allele carriers (male=15, female=11), where 23% of these subjects met metabolic syndrome criteria (OR=3.7, 95% CI=1.24–12.66, p=0.02). Thus, for the T allele subjects the risk of developing metabolic syndrome while receiving AAPs was 3.6 times greater than those with the CC genotype, despite similar AAPs exposure. Addition of the 1298 genotypes to the analysis did not significantly add to the model and so this data is not presented.

In looking at this data further we also examined the relationship between the homeostasis model assessment insulin resistance (HOMA-IR) value, waist circumference, and the MTHFR genotype (Fig. 1) with an interaction between these two terms which was statistically significant (F=8.6, df=2, p=0.0006) suggesting that subjects with a 677TT genotype are at the greatest risk for increased insulin resistance with increasing waist circumference compared to TC and CC genotype subjects. This was independent of age, gender, BMI, or metabolic syndrome diagnosis. Overall the power of this model to detect differences based on genotype was 0.72. The effect size for the TT genotype/waist interaction variable is 19. The addition of the 1298 genotypes to this analysis did not contribute significantly to the model. We also examined plasma homocysteine in a subset of subjects (n=46) and found an interesting trend. For the 677TT group the mean homocysteine concentration was 10.2±4.9, where the CT and CC groups were lower (CT 9.1±2.8, CC 9.9±3.5). Although these are not statistically different (p=0.67), the TT group's values are higher, suggesting a greater risk for CVD, which accompanies the increased risk of metabolic syndrome and insulin resistance in this group.

View Large Image Download to PowerPoint

Fig. 1. Relationship between MTHFR 677C/T genotype, HOMA-IR and waist circumference. Graphical representation of relationship between HOMA-IR (Y-axis) waist circumference in centimeters (X-axis) and, MTHFR 677C/T genotype (TT= •, TC=×, CC=+).

4. Discussion 

This investigation is the first to examine the relationship between the MTHFR 677C/T and 1298A/C variants and metabolic syndrome and insulin resistance risk in a schizophrenia population receiving atypical antipsychotics. Recent research has highlighted the importance of genes as well as environmental/nutritional factors in the body's ability to regulate genome machinery (Ames, 2001). This has lead to the field of epigenetics, whose focus is on how diet and nutrition affect DNA functioning or DNA binding proteins without altering the nucleotide sequence of DNA (Oommen et al., 2005, Reik, 2007).

Mechanistically, MTHFR plays an important role in folate metabolism and DNA maintenance. Folic acid is a water soluble B-vitamin involved in the synthesis, repair, and methylation of DNA (Friso and Choi, 2005), whose effective utilization is dependant on adequate daily intake as well as genetic variants altering metabolism (Friso and Choi, 2005). Methylenetetrahydrofolate reductase (MTHFR) is the enzyme responsible for the formation of methyltetrahydrofolate (5-methyl THF) from dietary folic acid, which allows conversion of homocysteine to methionine and adenosyl methionine which is the universal methyl donor for DNA and protein synthesis. Lack of this universal donor may result in DNA hypomethylation, leading to alterations in gene silencing. Functionally, the MTHFR variant is very important, particularly in the field of obstetrics as presence of the variant allele is a risk factor for the development of neural tube defects in an unborn child (Gueant et al., 2003). Previous reports have linked MTHFR variants to cardiovascular risk as well as to schizophrenia risk, but have not been examined in the context of metabolic complications seen with antipsychotic use (Klerk et al., 2002, Gueant et al., 2003, Lewis et al., 2005, Muntjewerff et al., 2005, Kempisty et al., 2006, Lee et al., 2006, Zintzaras, 2006). Additionally, a recent report showed the MTHFR T variant to be associated with the occurrence of negative symptoms in schizophrenia (Roffman et al., in press).

In general, the metabolism of both folate and homocysteine is undeniably linked and dependent on MTHFR. Thus, reduction in folate due to either low dietary intake, genetic predisposition, or both, may result in hyperhomocysteinemia, which has been linked to cardiovascular disease in the general population (Danesh and Lewington, 1998). Folate administration prevents homocysteine oxidative stress and consequential endothelial damage (Setola et al., 2004) as hyperhomocysteinemia (Moat et al., 2004) directly damage endothelial cells (Wall et al., 1980) and impair the release of nitrous oxide (Stamler et al., 1993), reducing endothelial dependent vasodilatation (Tawakol et al., 1997). Prolonged increases in plasma homocysteine of ≥1 μmol/L result in a 5% increased CVD risk.

From our data it appears that the MTHFR 677T allele is associated with a 3.6 fold greater risk for developing AAP associated metabolic syndrome, and the TT genotype may place individuals at greater risk for insulin resistance with greater central adiposity. Higher levels of insulin resistance pose an elevated danger for developing diabetes mellitus, metabolic syndrome, and endothelial dysfunction, which are significant risk factors for the development of CVD. As stated previously, the relationship between MTHFR and cardiovascular disease is not unique to patients with schizophrenia, but within this investigation the risk associated with the T allele was much higher than that expected in the general population. In normal controls, previous meta-analyses have found that patients with the MTHFR 677TT genotype have approximately a 16% increase in CVD compared to the CC genotype which is potentially related to hyperhomocysteinemia, and poor dietary folate intake (Klerk et al., 2002, Lewis et al., 2005). Previous reports have shown that on average, homocysteine concentrations in individuals with a MTHFR 677TT genotype are 2.5 μmol/L higher than CC patients (Brattstrom et al., 1998, Danesh and Lewington, 1998, Ueland et al., 2000). Thus, it may be that addition of medications, such as the atypical antipsychotics, associated with weight gain, insulin resistance, diabetes mellitus, and hyperlipidemia may be the “trigger” that results in a greater risk for metabolic syndrome in the schizophrenia population. Possibly most importantly, in looking at this group of subjects, this was a fairly young group with an age range of 20–57 years, and an overall mean age of 36 years. The occurrence of these potentially fatal conditions at such an early age, place patients at risk for developing even more cardiovascular complications, leading to greater physical and health related impairments. Not surprisingly, when the individual components of the metabolic syndrome are compared between the genotype groups very little differences were seen (Table 2). Given that presence of a metabolic syndrome diagnosis places individuals at greater risk for CVD than each of its individual components (Wilson et al., 2005a, Wilson et al., 2005b), better characterization of the unexplored relationship between folate pharmacogenetics, metabolic syndrome, insulin resistance, endothelial function, dietary folate intake, and AAPs in schizophrenia may help to identify risk factors for these serious complications, and aid in developing preventative measures.

5. Study limitations 

Due to the small number of subjects included in this analysis and the cross-sectional nature of this study, our data has several limitations. First, in looking at our current analysis, it appears that the majority of the relationships found with metabolic syndrome and insulin resistance is being driven by the small number of subjects with the 677TT genotype. Thus, future investigations need to expand the number of subjects included. Additionally, for many of these subjects this was the first time they had undergone screening for the metabolic syndrome, which is similar to other reports in the literature concerning the monitoring for these metabolic consequences in this population (McEvoy et al., 2005). Thus we cannot determine if these subjects developed the metabolic syndrome before or after starting AAP treatment. Secondly, we did not have dietary information available concerning folate intake, which may also limit our data analysis, as the effect of the MTFHR T allele has been shown to be more exaggerated in the face of poor dietary folate intake. Lastly, the subjects included in this study were not excluded due to substance abuse and were receiving a variety of antipsychotics. The decision to include all subjects was done in an effort to gain an understanding of metabolic syndrome risk in the “real world”, but unfortunately does introduce variability in the data.

Regardless, this investigation represents an important first step in delineating the relationship between MTHFR and cardiovascular risk in schizophrenia subjects treated with AAPs. Thus, it is imperative that we continue our investigation, expanding our sample size and including variants associated with other enzymes involved in folate and homocysteine metabolism resulting in reduced metabolic activity to further elucidate their role in the development, prevention, and treatment of AAP associated metabolic complications. Lastly, understanding the role of supplemental folate in the attenuation of these metabolic complications may help to improve the treatment of our patients and needs to be further investigated as we look for ways to reduced overall cardiovascular risk.

Role of funding source 

This project was supported by a grant from the National Institute of Mental Health (K08 MH64158) and the National Center for Research Resources, General Clinical Research Centers Program, National Institutes of Health (M01-RR-59). Both the NIMH and the NCRR had no further role in the study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.

Contributors 

Dr. Ellingrod designed the study, wrote the protocol, managed the literature searches and data analyses, and wrote the first draft of the manuscript. Dr. Miller oversaw the clinical care of the subjects, and aided in the interpretation of the results, and preparation of the manuscript. Dr. Taylor aided in the interpretation of the results and the manuscript preparation. Ms. Moline aided in the genotype analysis, and the management of the literature searches and manuscript writing. Mr. Holman and Ms. Kerr were responsible for subject recruitment, data management, and aided in the literature searches and manuscript writing.

Conflict of interest 

The authors have no conflicts of interest to disclose in relation to this manuscript.

Disclosure statement for Vicki L. Ellingrod, Pharm.D., BCPP

June 25, 2007 — for the last 3 years

Dr. Ellingrod is a paid author for LexiComp Inc., and has no other direct financial relationships with any pharmaceutical company, including consulting arrangements, speakers' bureau, or board memberships.

List of entities from whom compensation for professional services was received:

National Institutes of Health

Disclosure statement for Del D. Miller, Pharm.D., M.D.

June 25, 2007 — for the last 3 years

Dr. Miller has received research funds from Bristol Myers Squibb, Eli Lilly, and Pfizer, manufacturers of antipsychotics. He receives compensation as a consultant and for speaking for Astra Zeneca, Bristol Myers Squibb, Janssen Pharmaceutica, Organon USA, and Pfizer, manufacturers of antipsychotics.

List of entities from whom compensation for professional services was received:

Disclosure statement for Stephan F. Taylor, M.D.

June 25, 2007 — for the last 3 years

Dr. Taylor has performed paid consulting work for Pfizer and Organon. He has performed unpaid consulting work for Advanced Neuromodulation Systems, and has been in negotiations for clinical trial support from Advanced Neuromodulation Systems and Medtronic. He has participated in a clinical trial sponsored by Neuronetics.

Dr. Taylor has no other direct financial relationships with any pharmaceutical company, including consulting arrangements, speaker's bureau, or board memberships. He has a significant stock holding in Proctor and Gamble (>$10,000).

List of entities from whom compensation for professional services was received:

Ms. Moline, Mr. Holman and Ms. Kerr have nothing to disclose.