Sibling risk of anxiety disorders based on hospitalizations in Sweden

Authors


  • Declaration of interest: There are no conflicts of interest.

Xinjun Li, MD, MPH, PhD, Center for Primary Health Care Research, Lund University, CRC, hus 28, plan 11, ing 72, UMAS, 205 02 Malmö, Sweden. Email: xinjun.li@med.lu.se

Abstract

Aims:  This study used nationwide hospital records to examine sibling risk of any type of anxiety disorder in Sweden over a 40-year period.

Methods:  This study, carried out between 1 January 1968 and 31 December 2007, of the entire population of Sweden, linked information on family relationships from the nationwide Multi-Generation Register with information from the nationwide Swedish Hospital Discharge Register on first diagnosis of anxiety disorder. A total of 42 602 persons hospitalized for anxiety disorders and 2093 affected siblings were identified. Standardized incidence ratios (SIR) were calculated by comparing risk in siblings of persons hospitalized for anxiety disorders with risk in persons whose siblings had no hospital diagnosis of anxiety disorders.

Results:  The sibling risk was 2.26, which was independent of sex and age differences between siblings. The SIR was highest in siblings <20 years of age (2.83). Analysis of risk by subtype showed that having a sibling diagnosed with any anxiety disorder resulted in increased risks of a number of disorders; the highest increased risk was of social phobia (SIR 3.68, 95% confidence interval, 1.68–7.69). Risk of panic disorder, generalized anxiety disorder, mixed anxiety and depressive disorder, and obsessive–compulsive disorder was raised in female but not male siblings.

Conclusions:  Heritable effects likely play an important role in the cause of anxiety disorders, but the extent of their role remains to be established. Important contributions could be made by studies of gene–environment interactions that have sufficient sample sizes to produce reliable results.

ANXIETY DISORDERS ARE common psychological problems, with a prevalence of approximately 25%. Despite their high prevalence and socioeconomic impact,1 little is known about their cause. Genetic susceptibility has been proven to play a role in the development of anxiety disorders. A population-based sample of twins from Virginia showed a modest genetic influence of 36% for agoraphobia, 28% for panic disorder, and 23% for generalized anxiety disorder.2 Another study on twins demonstrated genetic influences as high as 44% on the development of panic disorder.3 A recently published large-scale population-based study from our group found that, after accounting for socioeconomic status, age, geographic region, and period of diagnosis, age-specific familial risks of anxiety disorders had a twofold increased risk among individuals having a parent with anxiety disorder.4

In the present study we included hospital data on all individuals in Sweden and their siblings who were born in 1932 and onward, that is, a total of 7.6 million individuals. The aim of this study was to define familial risk of hospitalization for all anxiety disorders and their subtypes in the Swedish population. The Swedish family dataset, that is, the Multi-Generation Register, is a validated source that has been proved to be reliable in the study of many familial diseases.5–7

METHODS

MigMed research database

Data used in this study were retrieved from the MigMed database, located at the Center for Primary Health Care Research at Lund University. We used the main diagnoses for anxiety disorders recorded in the register. Additional linkages were carried out to national census data to obtain individual socioeconomic status, occupation, geographical region of residence, national Registry of Causes of Death (to identify date of death), and the Emigration Registry (to identify date of emigration). All linkages were performed by the use of an individual national identification number that is assigned to each person in Sweden for their lifetime. This number was replaced by a serial number for each person in order to provide anonymity.

Outcome variable

Information on first hospital diagnoses of anxiety disorders was retrieved from the Swedish Hospital Discharge Register, which is part of the MigMed Database. For the years 1968–1986, this register contains diagnostic codes from the 8th edition of the International Statistical Classification of Diseases and Related Health Problems (ICD-8). For the years 1987–1996, it contains diagnostic codes from the 9th edition (ICD-9), and for the years 1997–2007, it contains diagnostic codes from the 10th edition (ICD-10). The subtypes of anxiety disorders for the years 1968–1996, when the ICD-8 and 9 were used, included anxiety disorder (300.0), phobic disorder (300.2), and obsessive–compulsive disorder (300.3). The subtypes of anxiety disorders for the years 1997–2007, when the ICD-10 was used, included agoraphobia (F40.0), social phobia (F40.1), generalized anxiety disorder (F41.1), mixed anxiety and depressive disorder (F41.2), other anxiety disorders (F41.3, F41.8, and F41.9), and obsessive–compulsive disorder (F42).

Statistical analysis

Person-years were calculated from the start of follow up on 1 January 1968, until first diagnosis of anxiety disorder, death, emigration, or the end of the study on 31 December 2007. Age-standardized incidence ratios were calculated for the whole follow-up period, divided into eight 5-year periods (1968–1972, 1973–1977, 1978–1982, 1983–1987, 1988–1992, 1993–1997, 1998–2002, and 2003–2007). Standardized incidence ratios (SIR) were calculated as the ratio of the observed and the expected number of cases.8 The expected numbers were calculated by the incidence rates for all individuals without a sibling history of anxiety disorder, and the rates were standardized by age (5-year groups), sex, time period (5-year intervals), socioeconomic status (manual worker, blue collar, professional, private, farmer and other) and residential area (big cities, Southern region, and Northern region). Sibling risks were calculated for individuals with siblings affected by anxiety disorder, compared with individuals whose siblings were not affected by this condition, using previously described cohort methods.9 Confidence intervals (95%CI) were calculated assuming a Poisson distribution, and they were adjusted for dependence between the sibling pairs.9

Ethical considerations

This study was approved by the ethics committee of Lund University, Sweden.

RESULTS

From 1968 to 2007 there were 16 942 first hospital diagnoses of anxiety disorders in male subjects and 25 660 in female subjects between the ages 0 and 75 years (Table 1). For the years in which the ICD-10 diagnostic codes were used (1997–2007), other anxiety disorders accounted for 41.6% of the first hospital diagnoses of anxiety disorders, followed by mixed anxiety and depressive disorder (21.9%), panic disorder (15.9%), and generalized anxiety disorder (8.8%). Age-specific first-hospitalization rates per 100 000 person years for anxiety disorders in men and women are shown in Figure 1. The overall hospitalization rates of anxiety disorders were 15.0/100 000 and 23.8/100 000 for men and women, respectively. The rates increased with increasing age up to the age of 40, and then increased in the older age groups. Female rates exceeded male rates after 10 years of age.

Table 1.  Number of first hospitalizations of anxiety disorders during the study period in persons aged 0–75 years in Sweden
SubtypeMenWomenAll
Cases%Cases%Cases%
ICD-8 (1968–1986), ICD-9 (1987–1996)      
 Anxiety state (300.0)697079.411 08883.418 05881.8
 Phobic disorder (300.2)7318.31 0507.91 7818.1
 Obsessive–compulsive disorder (300.3)108112.31 1598.72 24010.1
All8782100.013 297100.022 079100.0
ICD-10 (1997–2007)      
 Agoraphobia (F40.0)1231.51721.42951.4
 Social phobia (F40.1)2853.52211.85062.5
 Other phobic anxiety disorder (F40.2, F40.8, F40.9)811.0830.71640.8
 Panic disorder (F41.0)136316.71 90115.43 26415.9
 Generalized anxiety disorder (F41.1)6848.41 1259.11 8098.8
 Mixed anxiety and depressive disorder (F41.2)156319.22 92223.64 48521.9
 Other anxiety (F41.3, F41.8, F41.9)332240.75 20842.18 53041.6
 Obsessive–compulsive disorder (F42)7399.17315.91 4707.2
All8160100.012 363100.020 523100.0
Figure 1.

Age-specific incidence rates of anxiety disorders in men and women aged 0–75 years.

The age-specific sibling risks of any anxiety disorder among siblings are shown in Table 2. Altogether, 2093 affected siblings were scored, giving an overall SIR of 2.26. A gradient was found for both male and female subjects so that, with increasing age, the SIR decreased but remained significant in all age groups. For male siblings over age 50, the risk of any anxiety disorder was not significant.

Table 2.  Risk of anxiety disorders in siblings by age at diagnosis
Age at diagnosis (years)Male siblingFemale siblingAll siblings
OSIR95%CIOSIR95%CIOSIR95%CI
  1. Bold type: 95%CI does not include 1.00.

  2. CI, confidence interval; O, observed number of cases; SIR, standardized incidence ratio.

<20593.011.625.50872.711.544.731462.831.694.70
20–292082.741.694.443142.751.744.355222.751.784.24
30–392332.341.453.763842.501.593.916172.441.593.73
40–491611.861.123.062842.241.413.574452.091.343.24
≥501221.410.832.392411.821.132.923631.661.062.60
All7832.131.403.2313102.341.573.5020932.261.533.33

Table 3 shows SIR for specific types of anxiety disorders when a sibling was diagnosed with any form of anxiety disorder. For diagnoses made during the years the ICD-10 codes were used (1997–2007), the highest risk was of social phobia (3.68). The SIR for other anxiety disorders was 2.05. Risks of certain disorders were increased only in either male or female siblings. These included agoraphobia (3.64 for men), social phobia (4.49 for men), panic disorder (2.05 for women), generalized anxiety disorder (2.16 for women), mixed anxiety and depressive disorder (2.61 for women), and obsessive–compulsive disorder (3.01 for women).

Table 3.  Risk for specific anxiety disorders in siblings when a co-sibling has been diagnosed with any anxiety disorder
SubtypeMale siblingFemale siblingAll siblings
OSIR95%CIOSIR95%CIOSIR95%CI
  1. Bold type: 95%CI does not include 1.00.

  2. CI, confidence interval; O, observed number of cases; SIR, standardized incidence ratio.

ICD-8 (1968–1986), ICD-9 (1987–1996)            
 Anxiety state (300.0)3912.231.433.497052.471.623.7610962.381.583.57
 Phobic disorder (300.2)563.251.735.97612.291.244.151172.661.564.51
 Obsessive–compulsive disorder (300.3)391.670.843.22562.111.133.87951.901.093.29
All4862.251.453.488222.431.603.6813082.361.583.52
 ICD-10 (1997–2007)            
 Agoraphobia (F40.0)73.641.0210.6641.410.265.14112.310.815.86
 Social phobia (F40.1)184.491.8810.0772.510.707.35253.681.687.69
 Other phobic anxiety disorder (F40.2, F40.8, F40.9)32.670.3611.1921.440.107.4951.990.446.62
 Panic disorder (F41.0)361.420.702.78712.051.133.661071.781.033.05
 Generalized anxiety disorder (F41.1)282.020.954.14482.161.134.05762.111.173.73
 Mixed anxiety and depressive disorder (F41.2)491.640.863.061332.611.544.371822.251.373.68
 Other anxiety (F41.3, F41.8, F41.9)1362.081.233.481932.021.243.293292.051.293.22
 Obsessive–compulsive disorder (F42)201.880.814.12303.011.446.09502.431.274.53
All2971.951.233.094882.211.433.427852.111.393.20

Gender-specific risks of anxiety disorders were considered. The overall SIR was 1.83 (95%CI 1.17–2.84) and 2.28 (95%CI 1.51–3.44) for brother pairs and sister pairs, separately (data not shown).

In order to test for the extent of environmental sharing in the observed risks, the risk of anxiety disorders was calculated by age differences between siblings. Overall, few differences were found by age difference. Siblings with a difference in age of < 5 years had a SIR of 3.03 (95%CI 2.00–4.58), whereas those with ≥ 5 years difference in age had a SIR of 1.94 (95%CI 1.30–2.90) (data not shown).

DISCUSSION

The present study provided evidence of sibling risks of anxiety disorders in the 0–75-year-old population. Agoraphobia and social phobia in male siblings, and panic disorder, generalized anxiety disorders, mixed anxiety disorder and depressive disorder, and obsessive–compulsive disorder in female siblings all had substantial familial aggregation. The overall sibling risk of hospitalization for anxiety disorders if a sibling has had a hospital diagnosis of anxiety disorders was 2.26, a figure that is in line with familial risks found in earlier studies on anxiety disorders.10

Familial aggregation of anxiety disorders may be due to environmental factors shared by family members or due to shared genes. In this study, the analysis of sibling risks for anxiety disorders according to the difference in age assumed that a small age difference is a proxy for a higher degree of a shared family environment. However, the analysis gave no indication of an association between risk and a higher degree of environmental sharing. Accordingly, the twin studies on anxiety disorders are in line with our findings.10

Consistent with prior evidence in twins,11 we found no significant differences in familial risks of anxiety disorders between the sexes. However, for specific types of anxiety disorders, gender effects were noted, such as an excess of panic disorder, generalized anxiety disorder, mixed anxiety and depressive disorder, and obsessive–compulsive disorder in female siblings. In male siblings, an excess of agoraphobia and social phobia was found. When broken down by sex, the number of persons in each subtype group was small. In the largest subtype group (i.e. those with an ICD-8 or ICD-9 diagnosis of anxiety disorders), the risk was higher in female siblings (2.47) than in male siblings (2.23); however, the 95%CI overlapped between the sexes.

There were some limitations in the present study. First, we were not able to test for the validity of anxiety disorders diagnoses because our data were based on the entire Swedish population. However, we only used main diagnoses for anxiety disorders recorded in the hospital registers; that is, all patients were hospitalized mainly for anxiety disorders, which increases the possibility that the diagnoses are valid. Hospitalizations for anxiety disorders normally require a doctor's referral from the primary care service. Thus, each hospitalized patient is likely seen by at least two medical doctors, of which the hospital doctor is almost certainly likely to be a specialist. A further disadvantage is that we were bound by the ICD codes as used. Some of them are relatively broad and some codes for anxiety disorders are unspecified in the Swedish Hospital Discharge Register. During the years the ICD-10 codes were in use, other anxiety disorders were a common diagnosis. Furthermore, the ICD-8, ICD-9 and ICD-10 diagnostic codes were quite different and a completely unambiguous merging was not possible. However, we could partly account for this potential bias by ‘adjusting’ the analysis for time period. The study population was large enough to detect high familial risks for a number of specific diagnostic groups.

Heritable effects likely play an important role in the cause of anxiety disorders, but the extent of the roles of genetics and environment remains to be established.

ACKNOWLEDGMENTS

This work was supported by grants to Drs Kristina and Jan Sundquist from the Swedish Research Council (K2005-27X-15428-01A), the Swedish Council for Working Life and Social Research (2006-0386 and 2007-1754), and the Swedish Research Council Formas (2006-4255-6596-99 and 2007-1352). The project was also supported by an ALF project grant from the Region of Skåne in Sweden, and Lars Hiertas minne (FO2010-0317).

Ancillary