This cohort study examines the association between family major depression history and risk of major depression including association with age, sex, type of kinship, and age of the affected family member.
Key Points
Question
Does the association of familial depression with risk of major depression (MD) vary with age, sex, and type of affected family members?
Findings
In this cohort study of 2 903 430 individuals, maternal, paternal, full sibling, or half-sibling with MD were associated with 2-fold higher risks of MD in men and women. The risk increased with number of affected family members; individuals exposed when 30 years or older had markedly lower risk.
Meaning
Risk of MD was associated with increased number of affected family members, but it did not vary by sex or type of kinship; exposure to family MD during childhood and adolescence was associated with increased risk.
Abstract
Importance
Major depression (MD) aggregates within families, but how family history of MD confers risk of MD over the life course is unclear. Such knowledge is important to identify and prevent possible depressogenic effects of family environment.
Objective
To examine the association between family MD history and risk of MD including association with age, sex, type of kinship, and age of the affected family member.
Design, Setting, and Participants
This cohort study included all Danish citizens born from 1960 to 2003 with known parental identity followed up from their 15th birthday until time of MD, censoring, or December 31, 2018. Analysis took place between April 2022 and December 2022.
Exposures
Family members with first-time MD using International Classification of Diseases, Eighth Revision codes 296.09, 296.29, 298.09, and 300.49 or 10th Revision codes F32.0-F33.9, family members’ age at MD onset, and individuals’ age at exposure to family MD
Main Outcomes and Measures
Multivariable Poisson regression was applied to estimate the incidence rate ratio (IRR) with 95% CI of first-time MD.
Results
Of 2 903 430 individuals (1 486 574 [51.2%] men), 37 970 men (2.6%) and 70 223 women (5.0%) developed MD during follow-up. For men, exposure to maternal, paternal, or full sibling MD were associated with a 2-times higher risk of MD (IRR, 2.10 [95% CI, 2.02-2.19]; IRR, 2.04 [95% CI, 1.94-2.14]; IRR, 2.08 [95% CI, 1.97-2.19]) and the associated risk increased with number of affected family members. This pattern was similar for women. For men, family members’ age at MD onset was not associated with MD. For women, maternal MD onset at 69 years or younger was associated with higher IRRs of MD (age <40 years: IRR, 1.64 [95% CI, 1.28-2.10]; age 40-49 years: IRR, 1.62 [95% CI, 1.27-2.07]; age 50-59 years: IRR, 1.56 [95% CI, 1.22-2.00]; and age 60-69 years: IRR, 1.67 [95% CI, 1.28-2.16]) compared with women with maternal MD onset at 70 years or older. For men, exposure to maternal MD younger than 30 years (age <1 year: IRR, 1.95 [95% CI, 1.70-2.25]; age 1 to <12 years: IRR, 2.31 [95% CI, 2.16-2.47]; age 12 to <19 years: IRR, 2.18 [95% CI, 2.03-2.35]; age 19 to <30 years: IRR, 1.42 [95% CI, 1.32-1.53]) was associated with increased IRRs, while exposure to maternal MD at 30 years or older was associated with a lower IRR (0.77 [95% CI, 0.70-0.85]). The findings were similar across type of kinships and for women.
Conclusions and Relevance
In this study, risk of MD was associated with increased numbers of affected family members but did not vary by gender or type of kinship. Exposure to family MD during childhood and adolescence was associated with increased risk.
Introduction
Current evidence suggests that major depression (MD) aggregates within families. Studies have shown that parental history of MD confers risk of MD in offspring,1,2,3,4 and studies of twins and adoptees have provided evidence that the heritability of MD range between 30% and 50%.5,6 A 2- to 3-fold increased risk of MD has been identified among children and adolescents with at least 1 parent with MD compared with unexposed individuals.7,8,9,10 Similarly, an increased risk of MD has been found between grandparents and grandchildren.11,12,13 However, common to these studies is their relatively small sample size (range, 151-11 200 study individuals) and their sampling methods, often using probands to identify the study population, which may be limited by lack of representativeness to the general population. One study included a large population (N = 1 413 336) of all Swedish women giving birth between 1973 and 1992; however, this study only included daughters and granddaughters of later generations.12 In addition, only a few studies have examined the effect of MD among full siblings14 and none have examined the effect of half-siblings on the individual’s own risk of MD, to our knowledge. Finally, previous studies have not explored the independent effect of each family member’s MD.
It has been suggested that the heritability of MD varies with age of familial MD onset,15 with early-onset MD being more prominent among individuals with a family history.16,17,18 Additionally, the timing of MD in family members may be important as childhood and adolescence may be sensitive periods with shared environment with parents and siblings. Finally, women are twice as likely as men to experience MD,19,20,21 but these findings are inconsistent as some studies report no difference in MD risk between the sexes.14,15
Consequently, to understand the mechanisms behind aggregation of MD within families, more research is needed including information on which family relations experience the greatest aggregation and whether sensitive periods and sex differences exist.
The aim of this study was to provide a comprehensive investigation of how family history of MD is associated with risk of MD over the life course in men and women separately by examining the risk of MD based on (1) the type of kinship and number of family members affected, (2) the family members’ age at MD onset, and (3) the individual’s own age when exposed to family MD
Methods
This study was approved by the Danish Data Protection Agency. In Denmark, no informed consent is required for register-based studies relying exclusively on deidentified data. Results are reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Study Population
This cohort study used information on the entire Danish population from nationwide registers identified in the Danish Civil Registration System using the unique personal identification number (CPR), which enables linkage of recorded individual-level information in all registers.22 The cohort consisted of individuals born from 1960 through 2003 who were alive at age 15 years with CPR linkage to both parents, including biological and adoptive. See Figure 1 for details on selection of the population.
Figure 1. Flowchart of the Study Population and the Applied Selection Criteria.
Outcome
First registered MD diagnosis was identified in the Danish Psychiatric Research Register23 or the Danish National Patient Register.24 The Danish Psychiatric Research Register contains information on physician diagnosis at discharge for all admissions to psychiatric hospitals or departments in Denmark from 1969 onwards.23 From 1995, the Danish Psychiatric Research Register became an integrated part of the Danish National Patient Register and information on psychiatric outpatient and emergency department contacts was added.24 From 1975 to 1994, where only inpatient hospital contacts were available, cohort members born from 1960 to 1980 were included and, thus, were aged 15 to 34 years. In the remaining period (1995-2018) where the Danish National Patient Register included both inpatient and outpatient visits, cohort members born from 1960 to 2003 were included and, thus, were aged 15 to 58 years. Since 1994, the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) has been used for registration of diagnostic information. Prior to this, ICD-8 was applied. MD was defined using ICD-8 codes 296.09, 296.29, 298.09, and 300.49 and ICD-10 codes F32.0-F33.9, including both main and secondary diagnoses. Data on race and ethnicity were not reported.
Family History of MD (Exposure)
The main exposure was MD onset in family members (parents, full siblings, or half-siblings). Family members were linked through the Danish Civil Registration System. Through parental CPR numbers, full siblings were identified by shared parental CPR numbers, while half-siblings were identified as paternal (shared paternal CPR number) or maternal (shared maternal CPR number). The family members’ first-time MD diagnosis was identified in the Danish Psychiatric Research Register23 and the Danish National Patient Register24 as described previously.
We constructed a time-updated exposure variable indicating type of family MD dependent on other family members’ MD status, including MD diagnoses in family members prior to birth of the cohort member or the index date. If the cohort member had more than 1 full sibling or half-sibling with MD, only the first registered MD diagnosis within the sibling groups was considered. The variable was categorized as (1) no family members with MD, (2) maternal MD, (3) paternal MD, (4) full sibling with MD, (5) paternal half-sibling with MD, (6) maternal half-sibling with MD, (7) both parents with MD, (8) maternal and full siblings and/or half-sibling with MD, (9) paternal and full siblings and/or half-sibling with MD, and (10) both parents and full siblings and/or half-sibling with MD.
To investigate the association of family members’ age at MD onset with outcomes, we defined variables with the age at the family members’ first MD diagnosis for each affected family member. Age at MD onset for both parents was divided into (1) younger than 40 years, (2) 40 to 49 years, (3) 50 to 59 years, (4) 60 to 69 years, or (5) 70 years or older. For full siblings and half-siblings, age at MD diagnosis was categorized as (1) younger than 25 years, (2) 25 to 34 years, (3) 35 to 44 years, or (4) 45 years or older.
We further created time-updated variables measuring at which age cohort members were exposed to a family member’s MD onset. For both parents, the categorization was (1) no MD, (2) age younger than 1 year, (3) age 1 to younger than 12 years, (4) age 12 to younger than 19 years, (5) age 19 to younger than 30 years, and (6) 30 years or older. For full siblings and half-siblings, the categories younger than 1 year and age 1 to younger than 12 years were combined because there were too few events to provide estimates with sufficient precision.
Covariates
We included highest attained parental educational level at baseline from the Danish Population’s Education Register.25 If both parents had attained education, the highest attained educational level was chosen based on the dominance approach.26 Educational level was categorized as low (primary and lower secondary school), medium (upper secondary school, short- and medium high vocational education), high (longer vocational [university] education), or unknown (missing information). To account for the association of changes in living conditions with outcomes over time, we included calendar time for each observation period in 10-year intervals from 1975 to 2018. Further, age of the individuals categorized with 15 to 19 years as reference by 5-year increments to age 58 years was included.
Statistical Analysis
The incidence rate of MD among cohort members was calculated per 1000 person-years at risk. We conducted time-to-event analyses using multivariable Poisson regressions27 with the logarithm of risk time as offset variable to estimate the adjusted incidence rate ratio (IRR) and 95% CI of the associations between MD in cohort members and family MD, family members’ age at MD, and own age at exposure to family MD. Individuals were followed up from index date (15th birthday) until time of MD, death, emigration, or end of follow-up (December 31, 2018), whichever came first. In the analyses of family members’ age at MD onset as the exposure of MD risk, cohort members were included at the time of exposure. In all analyses, cohort members with no family history of MD were chosen as reference, except in the analyses of family members’ age at MD onset were 70 years or older was chosen among parents and 45 years or older among full siblings and half-siblings.
Several members from the same family may be included in the models. To account for intragroup (familial) correlation between individuals, we applied a cluster function by maternal CPR number, which relaxes the requirement of independent observations across individuals, affecting the standard errors and variance-covariance matrix. All models were stratified by sex and adjusted by age, calendar time, and highest attained parental education. Further, by use of Wald testing, we investigated whether the exposure groups in each model differed significantly from one another. Two-sided P values were statistically significant at .05. All statistical analyses were performed using Stata version 17.0 (StataCorp). Analysis took place between April 2022 and December 2022.
Sensitivity Analyses
We repeated the analysis of family MD (Figure 2), to investigate the association of potential delayed detection of MD, differential length of follow-up, inclusion of only main diagnoses with MD, nonadjustment for parental education, censoring at bipolar disease, and secular trends in the familial aggregation of MD. Further information is included in the supplementary materials (eMethods in Supplement 1).
Figure 2. Incidence Rate Ratios (IRRs) of Major Depression (MD) Among Individuals Exposed to Familial MD Compared With Individuals With No Exposure, Adjusted for Individual Age, Calendar Time, and Highest Attained Parental Education .
Results
This study included 2 903 430 cohort members followed up for a mean (SD) of 21 (13.0) years of whom 1 486 574 (51.2%) were men. Overall, 108 193 cohort members were registered with an MD diagnosis during follow-up, of which 20 274 (18.7%) were identified by inpatient contacts (2183 prior to 1995) and 87 919 (81.3%) by outpatient contacts from 1995 through 2018. The incidence rate of MD was 1.21 (95% CI, 1.20-1.22) for men and 2.36 (95% CI, 2.34-2.38) for women per 1 000 person-years. Men and women had similar exposure to family MD and educational backgrounds (Table).
Table. Exposure to Family MD, Highest Attained Parental Educational Level, and Year of Inclusion in 2 903 430 Men and Women Born From 1960-2003.
| Characteristic | No. (%)a | No. of MD outcomes, No. (%)b | No. of person-years | Incidence rate per 1000 person-years (95% CI)) | ||||
|---|---|---|---|---|---|---|---|---|
| Men | Women | Men | Women | Men | Women | Men | Women | |
| Total | 1 486 574 (51.2) | 1 416 856 (49.8) | 37 970 | 70 223 | 31 353 526 | 29 757 412 | 1.21 (1.20-1.22) | 2.36 (2.34-2.38) |
| ≥1 Family member with MD | 199 716 (13.4) | 189 412 (13.4) | 6947 (18.3) | 11 984 (17.1) | 4 477 950.6 | 4 259 395.2 | 1.55 (1.51-1.59) | 2.81 (2.76-2.86) |
| Parental MD status | ||||||||
| Father | ||||||||
| With MD | 47 918 (3.2) | 45 443 (3.2) | 1912 (5.0) | 3093 (4.4) | 1041 957.8 | 989 411 | 1.84 (1.75-1.92) | 3.13 (3.02-3.24) |
| Age at MD diagnosis, mean (SD), y | 49.6 (13.8) | 49.7 (14.0) | NA | NA | NA | NA | NA | NA |
| Mother | ||||||||
| With MD | 76 006 (5.1) | 72 007 (5.1) | 3038 (8.0) | 5115 (7.3) | 1 651 148 | 1 570 375.4 | 1.84 (1.78-1.91) | 3.26 (3.17-3.35) |
| Age at MD diagnosis, mean (SD) | 46.3 (14.0) | 46.5 (14.0) | NA | NA | NA | NA | NA | NA |
| Full sibling MD status | ||||||||
| No. of individuals with full siblings | 1 262 376 (84.9) | 1 201 720 (84.8) | NA | NA | NA | NA | NA | NA |
| No full sibling | 224 198 (15.1) | 215 136 (15.2) | 7298 (19.2) | 13 655 (19.4) | 4 520 474.5 | 4 307 585.6 | 1.61 (1.58-1.65) | 3.17 (3.12-3.22) |
| Full siblings | ||||||||
| Without MD | 1 198 022 (94.9) | 1 140 836 (94.9) | 28 658 (75.5) | 53 086 (75.6) | 25 264 191 | 23 950 557 | 1.13 (1.12-1.14) | 2.22 (2.20-2.24) |
| With MD | 64 354 (4.3) | 60 884 (4.3) | 2014 (5.3) | 3482 (5.0) | 1 568 860.6 | 1 499 270.1 | 1.28 (1.23-1.34) | 2.32 (2.24-2.40) |
| Age at MD diagnosis, mean (SD) | 31.0 (10.8) | 31.2 (10.9) | NA | NA | NA | NA | NA | NA |
| Paternal half-sibling MD status | ||||||||
| No. of individuals with paternal half-siblings | 241 983 (16.3) | 231 165 (16.3) | NA | NA | NA | NA | NA | NA |
| No paternal half-siblings | 1 244 591 (83.7) | 1 185 691 (83.7) | 29 966 (78.9) | 55 018 (78.3) | 26 810 674 | 25 433 259 | 1.12 (1.11-1.13) | 2.16 (2.15-2.18) |
| Paternal half-siblings | ||||||||
| Without MD | 225 324 (93.1) | 215 417 (93.2) | 7589 (20.0) | 14 424 (20.5) | 4 187 605.9 | 3 989 145 | 1.81 (1.77-1.85) | 3.61 (3.56-3.68) |
| With MD | 16 659 (6.9) | 15 749 (6.8) | 415 (1.1) | 781 (1.1) | 355 245.8 | 335 007.2 | 1.17 (1.06-1.29) | 2.33 (2.17-2.50) |
| Age at MD diagnosis, mean (SD) | 29.2 (9.9) | 29.2 (9.8) | NA | NA | NA | NA | NA | NA |
| Maternal half-sibling MD status | ||||||||
| No. of individuals with maternal half-siblings | 200 986 (13.5) | 190 709 (13.5) | NA | NA | NA | NA | NA | NA |
| No maternal half-siblings | 1 285 588 (86.5) | 1 226 114.7 (86.5) | 30 740 (80.6) | 56 358 (80.3) | 27 643 463 | 26 256 441 | 1.11 (1.10-1.12) | 2.15 (2.13-2.16) |
| Maternal half-siblings | ||||||||
| Without MD | 186 993 (93.0) | 177 433 (93.0) | 6 818 (18.0) | 13 059 (18.6) | 3 416 538.4 | 3 224 543.6 | 2.00 (1.95-2.04) | 4.05 (3.98-4.12) |
| With MD | 13 993 (7.0) | 13 276 (7.0) | 412 (1.1) | 806 (1.2) | 293 524 | 276 427.9 | 1.40 (1.27-1.55) | 2.92 (2.72-3.12) |
| Age at MD diagnosis, mean (SD) | 28.5 (9.5) | 28.4 (9.4) | NA | NA | NA | NA | NA | NA |
| Follow-up, y | 21.1 (12.9) | 21.0 (13.0) | NA | NA | NA | NA | NA | NA |
| Age at start of follow-up | 15 | 15 | NA | NA | NA | NA | NA | NA |
| Highest attained parental educational level | ||||||||
| Low | 280 611 (18.9) | 267 101 (18.9) | 9408 (24.8) | 16 164 (23.0) | 7 539 846.1 | 7 171 027 | 1.24 (1.22-1.27) | 2.25 (2.22-2.30) |
| Medium | 689 339 (46.4) | 657 666 (46.4) | 16 713 (44.0) | 32 718 (45.6) | 13 443 315 | 12 739 761 | 1.24 (1.22-1.26) | 2.57 (2.54-2.60) |
| High | 413 217 (27.8) | 393 590 (27.8) | 8608 (22.7) | 16 761 (23.9) | 6 670 334.4 | 6 288 049.9 | 1.29 (1.26-1.32) | 2.67 (2.63-2.71) |
| Unknownc | 103 407 (7.0) | 98 499 (7.0) | 3241 (8.5) | 4580 (6.5) | 3700 030.9 | 3 558 574.4 | 0.88 (0.85-0.91) | 1.29 (1.25-1.32) |
| Year of inclusion | ||||||||
| 1975-1984 | 386 795 (26.0) | 369 145 (26.1) | 11 912 (31.4) | 16 877 (24.0) | 13 720 528 | 13 152 396 | 0.87 (0.85-0.88) | 1.28 (1.26-1.30) |
| 1985-1994 | 342 796 (23.1) | 328 455 (23.2) | 10 564 (27.8) | 18 856 (26.9) | 9283 564.8 | 8793 984.3 | 1.14 (1.12-1.16) | 2.14 (2.11-2.18) |
| 1995-2004 | 282 848 (19.0) | 270 094 (19.1) | 8502 (22.3) | 19 076 (27.2) | 5067 331.8 | 4749 004.8 | 1.68 (1.64-1.71) | 4.02 (3.96-4.07) |
| 2005-2014 | 342 391 (23.0) | 324 204 (22.9) | 6444 (17.0) | 14 167 (20.2) | 3 018 146.1 | 2 812 469.6 | 2.14 (2.08-2.19) | 5.04 (4.95-5.12) |
| 2015-2018 | 131 744 (8.9) | 124 958 (8.8) | 548 (1.4) | 1247 (1.8) | 263 955.4 | 249 557 | 2.08 (1.91-2.26) | 5.00 (4.73-5.28) |
Abbreviations: MD, major depression; NA, not applicable.
The total number of men and women was used as the denominator for calculating the proportion of cohort members in strata. A total of 4816 men (0.4%) and 6189 women (0.3%) were registered with bipolar disorder during follow-up.
The total number of men and women with MD onset was used as the denominator for calculating proportion in strata. A total of 522 men (1.4%) and 829 women (1.2%) were registered with bipolar disorder prior to MD during follow-up. See eMethods and eFigure 5 in Supplement 1 for the incidence rate ratios of MD with censoring at time of bipolar disorder diagnosis.
Individuals with education completed before 1974 and immigrants with no educational record in Denmark.
Figure 2 and eTable 1 in Supplement 1 show the adjusted IRRs of MD among men and women exposed to familial MD compared with nonexposed individuals. For men, exposure to maternal or paternal MD was associated with 2.10 (95% CI, 2.02-2.19) and 2.04 (95% CI, 1.94-2.14) times higher IRR of MD, respectively. The corresponding IRRs for exposure to full sibling, paternal half-sibling, or maternal half-sibling with MD was 2.08 (95% CI, 1.97-2.19), 1.66 (95% CI, 1.48-1.86), and 1.89 (95% CI, 1.68-2.12), respectively. The IRRs of MD were similar for women. The IRRs were higher when more family members were affected. MD in both parents were associated with an IRR of 3.31 (95% CI, 2.85-3.85) for men and 2.76 (95% CI, 2.44-3.12) for women. If both parents and at least 1 full sibling or half-sibling had MD, the IRRs were 5.95 (95% CI, 4.37-8.09) for men and 3.48 (95% CI, 2.59-4.64) for women. Similar patterns of associations were seen in the sensitivity analyses (eFigure 1-6 in Supplement 1).
Figure 3 and eTable 2 in Supplement 1 display the association between the family members’ age at first MD onset with MD risk. For men, the family members’ age at MD onset was not associated with MD. For women, maternal MD onset at 69 years or younger were associated with higher IRRs of MD (onset at age <40 years: IRR, 1.64 [95% CI, 1.28-2.10]; 40-49 years: IRR, 1.62 [95% CI, 1.27-2.07]; 50-59 years: IRR, 1.56 [95% CI, 1.22-2.00]; and 60-69 years: IRR, 1.67 [95% CI, 1.28-2.16]) compared with women with maternal MD onset later in life (age ≥70 years) (Wald test: P < .001). Further, in women, only paternal MD at age 40 to 49 years was associated with a higher IRR for MD (1.31 [95% CI, 1.04-1.65]; Wald test: P = .05). There was no statistically significant association for full sibling or half-sibling age at MD onset.
Figure 3. Incidence Rate Ratios (IRRs) of Major Depression (MD) in Index Persons by Age at MD Diagnosis in Family Members, Stratified by Sex.
Models were adjusted for age, calendar time, and highest attained parental education.
Figure 4 and eTable 3 in Supplement 1 show the associations between the cohort member’s age at exposure to familial MD for each type of family member. Overall, the IRRs of MD were similar in men and women and were highest when exposed to familial depression during childhood and adolescence (younger than 19 years). The IRRs were slightly higher when exposed at age 19 to 30 years compared with individuals without family risk but only when exposed to maternal, paternal, and full sibling MD (and maternal half-sibling in women). In more detail, men exposed to maternal MD when younger than 30 years had increased IRRs compared with nonexposed men (age <1 year: IRR, 1.95 [95% CI, 1.70-2.25]; 1 to <12 years: IRR, 2.31 [95% CI, 2.16-2.47]; 12 to <19 years: IRR, 2.18 [95% CI, 2.03-2.35]; and 19 to <30 years: IRR, 1.42 [95% CI, 1.32-1.53]), while exposure to maternal MD in men at 30 years or older was associated with a lower IRR (0.77 [95% CI, 0.70-0.8]; Wald test: P = .002). This pattern was similar in men exposed to paternal MD younger than 30 years. Similarly, exposure to full sibling or half-sibling with MD in men younger than 19 years was associated with higher IRRs, while exposure at age 30 years or older was associated with lower IRRs. Men exposed to a full sibling with MD between age 19 to 30 years also had a higher IRR of MD compared with nonexposed individuals. The patterns of association were similar for women.
Figure 4. Incidence Rate Ratios (IRR) of Major Depression (MD) Among Individuals by Age of Exposure to Familial Depression Stratified by Sex and Adjusted for Age, Calendar Time, and Highest Attained Parental Education.
Discussion
In this cohort of 2 903 430 individuals followed up from age 15 years, family history of MD was associated with higher risks of MD in both men and women. Overall, the risk of MD was approximately 2 times higher among individuals with a family history of MD, independent of which family member was affected. The risk increased when more family members were affected. For men, having both parents and a full sibling or half-sibling with MD entailed the highest risk of MD, while this was less pronounced for women. In men, there was no difference in risk of MD regarding age of the family members at family MD onset. In women, an elevated risk of MD was present if maternal age at time of maternal MD was 69 years or younger. There was no association between other family members’ age at first MD and onset of MD in cohort members. Last, our results indicate that both men and women were at higher risk of MD if exposed to familial MD during childhood and early adulthood while the risk of MD was lower when exposed to familial MD at 30 years or older.
Our results confirm prior findings from smaller studies of families, twins, and siblings4,6,11,13,14,28,29 of a link between family history of MD and the individual’s own risk of MD. No prior study has examined the cumulative risk of MD from parents and siblings, but studies on the role of family history across 3 generations11,12,13,30 have indicated that the risk of MD accumulates with the number of affected generations. Our findings suggest an equally sized IRR of each affected family member. This shows the importance of family environment as risk estimates did not vary for family members who shared either 50% or 25% of genes eg, maternal MD vs sibling MD. This finding was further supported by the finding of a substantially higher risk of MD among men and women who were exposed to familial MD during childhood and adolescence. These patterns existed irrespective of which family member was affected and is in line with prior research.2,12,31 Childhood and adolescence may be sensitive periods because of the derivative family and socioeconomic consequences of MD while also cohabiting with the affected family member.31,32,33
Although we did not test it statistically, our findings suggest a similar relative association of parental MD history with risk of MD in men and women. This is contrary to a US study of 637 adolescents aged 18 to 19 years, where women with parental MD history were at higher risk of elevated depressive symptoms than men.21 Siblings and twin studies have shown inconsistency in sex differences in heritability of MD.15,20,28,34 These differing findings could be due to methodological issues, as our study population was sampled in an unselected population using registry data, while the US study recruited participants based on volunteering, which often causes lack of representativeness to the general population. The findings could also be affected by the measurement of MD varying from diagnoses by physicians registered in hospital records to assessments using, for example, the Beck Depression Inventory II, which may be subject to different levels of detection related to sex.35,36
Prior research suggests that early-onset MD is associated with more severe MD symptoms37 and an increased risk of familial aggregation.16,17,37 Thus, it was surprising that we did not find any clear association of the age at the family members’ first MD with outcomes. To ensure that this was not due to a lack of information on parental depression in the registers, we conducted a sensitivity analysis including only individuals of parents born from 1954 onward, which ensured that onset of parental MD could be detected from age 15 years. A possible explanation to our diverging findings could be that our analyses are restricted to MD diagnosed in the secondary health care system. Hence, our findings may not be directly comparable with prior studies as individuals in general may be treated for MD in the secondary health care system later in life compared with individuals with milder MD treated in primary care or identified through assessments in the general population.
Strengths and Limitations
The major advantage of this study is the possibility to link family members for a national cohort born from 1960 through 2003. The linkage is almost complete for persons born after 1955 and provided us with a large, unselected study population with family data recorded with no reporting errors. We had almost complete follow-up of MD diagnoses through the Danish National Patient Register and the Danish Psychiatric Research Register since 1969, which is reported to be of high quality.38 However, using patient registry data, we were limited to hospital diagnoses that may not include patients with milder MD treated in primary care, which may bias our results. Furthermore, the analyses of age at onset may not provide the precise age at exposure as some individuals may not be diagnosed with MD until years after their first onset of symptoms. This could cause the findings of age to be staggered in the sense that onset at age 30 to 39 years may represent exposure at age 20 to 29 years. However, as we expect the delay in diagnosis to be present at all age groups, the age trends identified should be valid. A further limitation is a potential discrepancy between legal parents and biological parents in the Danish Civil Registration System, yet, at least for maternal linkage, there was conformity between the information in CPR and in the Danish Medical Birth Register.39 Finally, the study design cannot separate the impact of environmental and genetic influences on MD. In this regard, exposure to family MD was considered broadly by whether a family member was diagnosed with MD, regardless of whether the family members lived in the same household during episodes of MD. Furthermore, as both inpatient and outpatient diagnoses were included, some may have been hospitalized during at least some part of the depressive episode, and thus, we do not distinguish between various degrees of physical exposure to family MD. Our findings illustrate this broader definition of exposure to family MD. It is possible that physical exposure may show larger associations.
Conclusions
The risk of MD was approximately 2 times higher among both men and women with a family history of MD, irrespective of which family member was affected. The risk of MD increased with more family members affected and was substantially higher in individuals exposed to family MD at age younger than 30 years. There was no clear association between age of family member at first onset of MD and risk of MD.
eMethods.
eTable 1. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to family MD compared with individuals with no exposure with 95% confidence intervals (CI)
eTable 2. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to family MD by age at MD diagnosis in family members with 95% confidence intervals (CI) stratified by sex
eTable 3. The incidence rate ratios (IRR) of major depression (MD) among individuals by age of exposure to family MD stratified by sex
eFigure 1. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to family MD compared with individuals with no exposure, adjusted for age, calendar time and highest attained parental education with 95% confidence intervals (CI) among individuals with parents born from 1954 onwards
eFigure 2. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to family MD compared with individuals with no exposure, adjusted for age, calendar time and highest attained parental education with 95% confidence intervals (CI) with follow-up restricted to 10 years
eFigure 3. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to familial MD compared with individuals with no exposure, adjusted for age, calendar time and highest attained parental education with 95% confidence intervals (CI) restricted to include only individuals with a main diagnosis of MD
eFigure 4. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to familial MD compared with individuals with no exposure, adjusted for age and calendar time with 95% confidence intervals (CI) (Non-adjustment of highest attained parental educational level)
eFigure 5. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to familial MD compared with individuals with no exposure, adjusted for age, calendar time and highest attained parental education with 95% confidence intervals (CI) with censoring at time of bipolar disorder diagnosis, MD diagnosis, death, emigration, or end of follow up (31st of December 2018), whichever came first
eFigure 6. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to familial MD compared with individuals with no exposure, adjusted for age, calendar time and highest attained parental education with 95% confidence intervals (CI) stratified by birth cohorts 1960-1969, 1970-1979 and 1980-1989
Data sharing statement
References
- 1.Weissman MM, Gammon GD, John K, et al. Children of depressed parents: increased psychopathology and early onset of major depression. Arch Gen Psychiatry. 1987;44(10):847-853. doi: 10.1001/archpsyc.1987.01800220009002 [DOI] [PubMed] [Google Scholar]
- 2.Klein DN, Lewinsohn PM, Rohde P, Seeley JR, Olino TM. Psychopathology in the adolescent and young adult offspring of a community sample of mothers and fathers with major depression. Psychol Med. 2005;35(3):353-365. doi: 10.1017/S0033291704003587 [DOI] [PubMed] [Google Scholar]
- 3.Kendler KS, Davis CG, Kessler RC. The familial aggregation of common psychiatric and substance use disorders in the National Comorbidity Survey: a family history study. Br J Psychiatry. 1997;170:541-548. doi: 10.1192/bjp.170.6.541 [DOI] [PubMed] [Google Scholar]
- 4.Musliner KL, Trabjerg BB, Waltoft BL, et al. Parental history of psychiatric diagnoses and unipolar depression: a Danish National Register-based cohort study. Psychol Med. 2015;45(13):2781-2791. doi: 10.1017/S0033291715000744 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Sullivan PF, Neale MC, Kendler KS. Genetic epidemiology of major depression: review and meta-analysis. Am J Psychiatry. 2000;157(10):1552-1562. doi: 10.1176/appi.ajp.157.10.1552 [DOI] [PubMed] [Google Scholar]
- 6.Kendler KS, Ohlsson H, Lichtenstein P, Sundquist J, Sundquist K. The genetic epidemiology of treated major depression in Sweden. Am J Psychiatry. 2018;175(11):1137-1144. doi: 10.1176/appi.ajp.2018.17111251 [DOI] [PubMed] [Google Scholar]
- 7.Low NCP, Dugas E, Constantin E, Karp I, Rodriguez D, O’Loughlin J. The association between parental history of diagnosed mood/anxiety disorders and psychiatric symptoms and disorders in young adult offspring. BMC Psychiatry. 2012;12:188. doi: 10.1186/1471-244X-12-188 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Weissman MM, Wickramaratne P, Nomura Y, Warner V, Pilowsky D, Verdeli H. Offspring of depressed parents: 20 years later. Am J Psychiatry. 2006;163(6):1001-1008. doi: 10.1176/ajp.2006.163.6.1001 [DOI] [PubMed] [Google Scholar]
- 9.Klein DN, Lewinsohn PM, Seeley JR, Rohde P. A family study of major depressive disorder in a community sample of adolescents. Arch Gen Psychiatry. 2001;58(1):13-20. doi: 10.1001/archpsyc.58.1.13 [DOI] [PubMed] [Google Scholar]
- 10.Lieb R, Isensee B, Höfler M, Pfister H, Wittchen HU. Parental major depression and the risk of depression and other mental disorders in offspring: a prospective-longitudinal community study. Arch Gen Psychiatry. 2002;59(4):365-374. doi: 10.1001/archpsyc.59.4.365 [DOI] [PubMed] [Google Scholar]
- 11.van Dijk MT, Murphy E, Posner JE, Talati A, Weissman MM. Association of multigenerational family history of depression with lifetime depressive and other psychiatric disorders in children: results from the Adolescent Brain Cognitive Development (ABCD) study. JAMA Psychiatry. 2021;78(7):778-787. doi: 10.1001/jamapsychiatry.2021.0350 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Josefsson A, Vikström J, Bladh M, Sydsjö G. Major depressive disorder in women and risk for future generations: population-based three-generation study. BJPsych Open. 2019;5(1):e8. doi: 10.1192/bjo.2018.83 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Weissman MM, Berry OO, Warner V, et al. A 30-year study of 3 generations at high risk and low risk for depression. JAMA Psychiatry. 2016;73(9):970-977. doi: 10.1001/jamapsychiatry.2016.1586 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Li X, Sundquist K, Hemminki K, Sundquist J. Familial risks for depression among siblings based on hospitalizations in Sweden. Psychiatr Genet. 2008;18(2):80-84. doi: 10.1097/YPG.0b013e3282f08ac9 [DOI] [PubMed] [Google Scholar]
- 15.Wium-Andersen MK, Villumsen MD, Wium-Andersen IK, et al. Familial risk and heritability of depression by age at first diagnosis in Danish twins. Acta Psychiatr Scand. 2020;142(6):446-455. doi: 10.1111/acps.13238 [DOI] [PubMed] [Google Scholar]
- 16.Tozzi F, Prokopenko I, Perry JD, et al. Family history of depression is associated with younger age of onset in patients with recurrent depression. Psychol Med. 2008;38(5):641-649. doi: 10.1017/S0033291707002681 [DOI] [PubMed] [Google Scholar]
- 17.Nierenberg AA, Trivedi MH, Fava M, et al. Family history of mood disorder and characteristics of major depressive disorder: a STAR*D (sequenced treatment alternatives to relieve depression) study. J Psychiatr Res. 2007;41(3-4):214-221. doi: 10.1016/j.jpsychires.2006.02.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Parker G, Roy K, Hadzi-Pavlovic D, Mitchell P, Wilhelm K. Distinguishing early and late onset non-melancholic unipolar depression. J Affect Disord. 2003;74(2):131-138. doi: 10.1016/S0165-0327(02)00002-2 [DOI] [PubMed] [Google Scholar]
- 19.Kendler KS, Prescott CA. A population-based twin study of lifetime major depression in men and women. Arch Gen Psychiatry. 1999;56(1):39-44. doi: 10.1001/archpsyc.56.1.39 [DOI] [PubMed] [Google Scholar]
- 20.Kendler KS, Gatz M, Gardner CO, Pedersen NL. A Swedish national twin study of lifetime major depression. Am J Psychiatry. 2006;163(1):109-114. doi: 10.1176/appi.ajp.163.1.109 [DOI] [PubMed] [Google Scholar]
- 21.Morris BH, McGrath AC, Goldman MS, Rottenberg J. Parental depression confers greater prospective depression risk to females than males in emerging adulthood. Child Psychiatry Hum Dev. 2014;45(1):78-89. doi: 10.1007/s10578-013-0379-5 [DOI] [PubMed] [Google Scholar]
- 22.Pedersen CB, Gøtzsche H, Møller JO, Mortensen PB. The Danish Civil Registration System: a cohort of eight million persons. Dan Med Bull. 2006;53(4):441-449. [PubMed] [Google Scholar]
- 23.Mors O, Perto GP, Mortensen PB. The Danish Psychiatric Central Research Register. Scand J Public Health. 2011;39(7)(suppl):54-57. doi: 10.1177/1403494810395825 [DOI] [PubMed] [Google Scholar]
- 24.Lynge E, Sandegaard JL, Rebolj M. The Danish National Patient Register. Scand J Public Health. 2011;39(7)(suppl):30-33. doi: 10.1177/1403494811401482 [DOI] [PubMed] [Google Scholar]
- 25.Jensen VM, Rasmussen AW. Danish education registers. Scand J Public Health. 2011;39(7)(suppl):91-94. doi: 10.1177/1403494810394715 [DOI] [PubMed] [Google Scholar]
- 26.Erikson R. Social class of men, women and families. Sociology. 1984;18(4):500-514. doi: 10.1177/0038038584018004003 [DOI] [Google Scholar]
- 27.Andersen PK, Borgan Ø, Gill RD, Keiding N. Statistical Models Based on Counting Processes. Springer US; 1993, doi: 10.1007/978-1-4612-4348-9. [DOI] [Google Scholar]
- 28.Olino TM, Lewinsohn PM, Klein DN. Sibling similarity for MDD: evidence for shared familial factors. J Affect Disord. 2006;94(1-3):211-218. doi: 10.1016/j.jad.2006.04.008 [DOI] [PubMed] [Google Scholar]
- 29.Olino TM, Pettit JW, Klein DN, Allen NB, Seeley JR, Lewinsohn PM. Influence of parental and grandparental major depressive disorder on behavior problems in early childhood: a three-generation study. J Am Acad Child Adolesc Psychiatry. 2008;47(1):53-60. doi: 10.1097/chi.0b013e31815a6ae6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Hancock KJ, Mitrou F, Shipley M, Lawrence D, Zubrick SR. A three generation study of the mental health relationships between grandparents, parents and children. BMC Psychiatry. 2013;13:299. doi: 10.1186/1471-244X-13-299 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Jaffee SR, Sligo JL, McAnally HM, Bolton AE, Baxter JM, Hancox RJ. Early-onset and recurrent depression in parents increases risk of intergenerational transmission to adolescent offspring. J Child Psychol Psychiatry. 2021;62(8):979-988. doi: 10.1111/jcpp.13356 [DOI] [PubMed] [Google Scholar]
- 32.Gilman SE, Kawachi I, Fitzmaurice GM, Buka SL. Socioeconomic status in childhood and the lifetime risk of major depression. Int J Epidemiol. 2002;31(2):359-367. doi: 10.1093/ije/31.2.359 [DOI] [PubMed] [Google Scholar]
- 33.Copeland WE, Wolke D, Shanahan L, Costello EJ. Adult functional outcomes of common childhood psychiatric problems: a prospective, longitudinal study. JAMA Psychiatry. 2015;72(9):892-899. doi: 10.1001/jamapsychiatry.2015.0730 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.van Sprang ED, Maciejewski DF, Milaneschi Y, et al. Familial resemblance in mental health symptoms, social and cognitive vulnerability, and personality: a study of patients with depressive and anxiety disorders and their siblings. J Affect Disord. 2021;294:420-429. doi: 10.1016/j.jad.2021.06.072 [DOI] [PubMed] [Google Scholar]
- 35.Salokangas RKR, Vaahtera K, Pacriev S, Sohlman B, Lehtinen V. Gender differences in depressive symptoms: an artefact caused by measurement instruments? J Affect Disord. 2002;68(2-3):215-220. doi: 10.1016/S0165-0327(00)00315-3 [DOI] [PubMed] [Google Scholar]
- 36.Bulhões C, Ramos E, Severo M, Dias S, Barros H. Measuring depressive symptoms during adolescence: what is the role of gender? Epidemiol Psychiatr Sci. 2019;28(1):66-76. doi: 10.1017/S2045796017000312 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Zisook S, Lesser I, Stewart JW, et al. Effect of age at onset on the course of major depressive disorder. Am J Psychiatry. 2007;164(10):1539-1546. doi: 10.1176/appi.ajp.2007.06101757 [DOI] [PubMed] [Google Scholar]
- 38.Kessing L. Validity of diagnoses and other clinical register data in patients with affective disorder. Eur Psychiatry. 1998;13(8):392-398. doi: 10.1016/S0924-9338(99)80685-3 [DOI] [PubMed] [Google Scholar]
- 39.Bliddal M, Broe A, Pottegård A, Olsen J, Langhoff-Roos J. The Danish medical birth register. Eur J Epidemiol. 2018;33(1):27-36. doi: 10.1007/s10654-018-0356-1 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods.
eTable 1. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to family MD compared with individuals with no exposure with 95% confidence intervals (CI)
eTable 2. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to family MD by age at MD diagnosis in family members with 95% confidence intervals (CI) stratified by sex
eTable 3. The incidence rate ratios (IRR) of major depression (MD) among individuals by age of exposure to family MD stratified by sex
eFigure 1. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to family MD compared with individuals with no exposure, adjusted for age, calendar time and highest attained parental education with 95% confidence intervals (CI) among individuals with parents born from 1954 onwards
eFigure 2. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to family MD compared with individuals with no exposure, adjusted for age, calendar time and highest attained parental education with 95% confidence intervals (CI) with follow-up restricted to 10 years
eFigure 3. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to familial MD compared with individuals with no exposure, adjusted for age, calendar time and highest attained parental education with 95% confidence intervals (CI) restricted to include only individuals with a main diagnosis of MD
eFigure 4. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to familial MD compared with individuals with no exposure, adjusted for age and calendar time with 95% confidence intervals (CI) (Non-adjustment of highest attained parental educational level)
eFigure 5. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to familial MD compared with individuals with no exposure, adjusted for age, calendar time and highest attained parental education with 95% confidence intervals (CI) with censoring at time of bipolar disorder diagnosis, MD diagnosis, death, emigration, or end of follow up (31st of December 2018), whichever came first
eFigure 6. The incidence rate ratios (IRR) of major depression (MD) among individuals exposed to familial MD compared with individuals with no exposure, adjusted for age, calendar time and highest attained parental education with 95% confidence intervals (CI) stratified by birth cohorts 1960-1969, 1970-1979 and 1980-1989
Data sharing statement
