Estrogen receptor α gene polymorphisms and anxiety disorder in an elderly population

SIR—Estrogens act on various monoamine systems in the brain, affect synaptogenesis and have neuroprotective properties. Previously, estrogen receptor (ER) polymorphisms have been related to personality traits. We found that in women ERα haplotypes created by the PvuII (c.454–397T>C) and the XbaI (c.454–351A>G) polymorphisms increased the likelihood of having an anxiety but not of having a depressive disorder; no relation between mood or anxiety disorders and ERα haplotypes was observed in men.

This study is based on data from the Rotterdam Study, a population-based cohort of persons over 55 years. A random selection of 2468 participants of a Caucasian origin filled in the Hospital Anxiety and Depression Scale (HADS) during the second examination and was genotyped. Anxiety was defined as a score of 11 or more on the HADS-anxiety subscale, which traditionally denotes probable anxiety disorder.1 We found a probable anxiety disorder in 148 (6.0%) persons. Additionally, we defined a more homogenous group with anxiety and no comorbid depression. This group (n=84) was restricted to those with a HADS-depression score of less than 11. Depressive disorders were assessed with the Center for Epidemiological Studies Depression scale (CES-D) in 4098 persons participating in the third examination (1997–1999). Screen-positive subjects (cut-off score was 16) had a psychiatric interview with the present state examination to classify depression according to DSM-IV criteria. In total, 321 (7.8%) screened positive for depressive symptoms and of these, 132 had a DSM-IV depressive disorder.

We studied two single nucleotide polymorphisms (SNP) identified in the ERα gene, the c.454–397T>C (also known as PvuII) and the c.454–351A>G (also known as XbaI), to define haplotypes as described previously.2 All SNPs were genotyped with Taqman Assays-by-Design SNP Genotyping Service and ABI PRISM® 7900HT Sequence Detection System (both: Applied Biosystems). The association of the ERα haplotype with depressive symptoms, and depressive disorders, and anxiety was assessed separately in men and women using logistic regression and testing an allele–dose model. Only the results of the ERα PvuII–XbaI haplotype 1 (frequency 52.8%) are reported as the haplotype alleles are closely related. Genotype distribution was in Hardy–Weinberg equilibrium.

The ERα haplotype 1 was associated with the likelihood of having anxiety in women, but not in men (Table 1). Furthermore, we found a significant allele–dose effect per copy of haplotype 1 (odds ratio: 1.42; 95% CI: 1.05–1.91, P=0.02). The effect of haplotype 1 on anxiety without comorbid depression was even stronger. No relationship was observed with depressive symptoms (Table 1) or DSM-IV depressive disorder (P=0.73 for women and P=0.72 for men).

Table 1 Relationship of anxiety and depressive symptoms with estrogen receptor α genotype in men and women from the Rotterdam study

Studies of the putative association of ERα haplotypes and variations in behavior have been sparse. ERα polymorphisms have been associated with parent's report of anxiety at the age of 13–14 years.3 Other studies concentrated on personality traits rather than specific mood states.4, 5 Furthermore, opposite findings were reported in men and women. A gender difference like in our study may be explained by the cessation of gonadal function in postmenopausal women, who have lower estrogen levels than men.6 Interestingly, our observations are supported by very detailed studies of behavior in ERα knockout mice.7 Female ERα knockout mice were more aggressive, whereas the respective male mice showed hardly any offensive attacks as compared to wild-type mice. At the same time, anxiety and fear levels were increased in female mice but not in male knockout mice. Further support for a role of the ERα in anxiety comes from studies of the receptor distribution in the human brain. ERα mRNA is expressed at relatively higher levels in the amygdala and the hypothalamus, both involved in emotional regulation.8

Our results suggest that ERα haplotypes may increase the risk of having an anxiety, but not a depressive, disorder. This is remarkable as anxiety is very frequently accompanied by symptoms of depression. Our observation could be a false-positive finding. However, the population-based design, the large sample, and the allele–dose effect support a true association. Furthermore, ERα haplotypes increased the likelihood of anxiety without significant comorbid depression. This is in line with studies showing that the perimenopausal period is characterized by mild affective symptoms, whereas the incidence of mood disorders does not increase.9 Moreover, the core features of the perimenopausal period and the perimenstrual syndrome are irritability, nervousness and tension rather than a depressed effect.10 Taken together, this suggests that ERα polymorphisms have a causal role in the anxious-irritable phenotype.


  1. 1

    Zigmond AS, Snaith RP . Acta Psychiatr Scand 1983; 67: 361–370.

  2. 2

    Schuit SC et al. JAMA 2004; 291: 2969–2977.

  3. 3

    Prichard Z et al. Am J Med Genet 2002; 114: 169–176.

  4. 4

    Westberg L et al. Mol Psychiatry 2003; 8: 118–122.

  5. 5

    Comings DE et al. Mol Psychiatry 1999; 4: 374–377.

  6. 6

    Burger HG et al. Novartis Found Symp 2002; 242: 161–167; discussion 167–71.

  7. 7

    Ogawa S et al. Endocrinology 1998; 139: 5070–5081.

  8. 8

    Ostlund H et al. Ann NY Acad Sci 2003; 1007: 54–63.

  9. 9

    Robinson GE . CNS Drugs 2001; 15: 175–184.

  10. 10

    Landen M, Eriksson E . Depress Anxiety 2003; 17: 122–129.

Download references

Author information



Corresponding author

Correspondence to H Tiemeier.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Tiemeier, H., Schuit, S., den Heijer, T. et al. Estrogen receptor α gene polymorphisms and anxiety disorder in an elderly population. Mol Psychiatry 10, 806–807 (2005).

Download citation

Further reading


Quick links