American Journal of Medical Genetics Part B: Neuropsychiatric Genetics

Volume 159B, Issue 5 pp. 580-588

Research Article

Full Access

Genetic and environmental causes of variation in perceived loneliness in young people^†

Trine Waaktaar,

Corresponding Author

Trine Waaktaar

[email protected]

Centre for Child and Adolescent Mental Health, and Department for Psychology, University of Oslo, Oslo, Norway

RBUP Eastern and Southern Norway, 4623 Nydalen, 0405 Oslo, Norway.Search for more papers by this author

Svenn Torgersen,

Svenn Torgersen

Centre for Child and Adolescent Mental Health, Oslo, Norway

Department for Psychology, University of Oslo, Oslo, Norway

Search for more papers by this author

Trine Waaktaar,

Corresponding Author

Trine Waaktaar

[email protected]

Centre for Child and Adolescent Mental Health, and Department for Psychology, University of Oslo, Oslo, Norway

RBUP Eastern and Southern Norway, 4623 Nydalen, 0405 Oslo, Norway.Search for more papers by this author

Svenn Torgersen,

Svenn Torgersen

Centre for Child and Adolescent Mental Health, Oslo, Norway

Department for Psychology, University of Oslo, Oslo, Norway

Search for more papers by this author

First published: 16 May 2012

https://doi.org/10.1002/ajmg.b.32064

Citations: 15

^†

How to Cite this Article: Waaktaar T, Torgersen S. 2012. Genetic and Environmental Causes of Variation in Perceived Loneliness in Young People. Am J Med Genet Part B 159B:580–588.

Share a link

Email
Wechat
Bluesky

Abstract

Loneliness is prevalent in adolescence, despite the widespread expectation directed to young people to start building close relationships beyond the nuclear family. The aim of the present study was to explore the causal genetic and environmental structure behind variability in adolescents' perceived loneliness. Seven national cohorts (ages 12–18 years) of Norwegian twins reared together (1,394 twin pairs) participated. Perceived loneliness was measured with five items from the UCLA Loneliness Scale. Data were collected from mothers, fathers, and twins' self ratings by means of a posted questionnaire. Biometric analyses were applied, testing the causal architecture of loneliness within a psychometric model with one common latent factor in addition to specific genetic and environmental sources influencing the scores of each informant. The results showed a heritability (h²) of 75% on the latent perceived loneliness factor, and nonshared environmental effects (e²) explaining the remaining 25% of the latent factor variance. There were also significant rater-specific genetic and nonshared environmental effects. No shared environmental effects were found in the model, and there were no sex differences in the estimates. This study showed that variation in perceived loneliness in adolescents is highly genetic. Additional genetic and nonshared environmental etiological sources are to some extent represented in the scores of the specific rater. © 2012 Wiley Periodicals, Inc.

INTRODUCTION

Adolescence is a period where the young person is expected to move beyond the family to form equal relations with friends and peers. However, youths differ markedly in the ways in which they experience and solve this phase-specific developmental task. Studies indicate that loneliness is prevalent in adolescents [Pinquart and Sörensen, 2003]. Higher prevalences of loneliness have specifically been found in youth with chronic illness [Shute and Walsh, 2005] and developmental disorders [Poulsen et al., 2007; Lasgaard et al., 2010].

Salient predictors of loneliness in young people include gender, internalizing symptoms, shyness, and low self-esteem, as well as more contextual factors such as social support and parental expressiveness [Mahon et al., 2006]. Loneliness in young people has been particularly related to depression, deliberate self-harm, and suicidal behavior [Groholt et al., 2000]. Longitudinal studies indicate that the loneliness–depression relationship goes primarily in the direction of depression leading to loneliness [Lasgaard et al., 2011]. However, loneliness has also been found to act as a mediator between early peer isolation and later depression in adolescents [Hawkley et al., 2010b].

Persistent loneliness has been related to a series of negative health outcomes—physical [Caspi et al., 2006; Aanes et al., 2010; Hawkley et al., 2010a, b] as well as mental [Fontaine et al., 2009; Kim et al., 2009; Qualter et al., 2010]. Interestingly, it seems that the perception of being lonely, which does not necessarily coincide with actual social isolation, is what predicts other problems [de Jong-Gierveld, 1987; Cole et al., 2007; Bernardon et al., 2011; Cacioppo et al., 2011]. Thus, the mediating role of personality in the relationship between perceived loneliness and other mental health difficulties has been documented in recent studies [Kunst and van Bon-Martens, 2011; Vanhalst et al., 2011].

Earlier genetically informative twin studies of perceived loneliness in children have yielded cross sectional additive genetic estimates around 50% [McGuire and Clifford, 2000]. Moderate heritability estimates were also found in two studies of adult twins [Boomsma et al., 2005, 2006; Distel et al., 2010]. An overview of the heritability results reported in former studies on loneliness is shown in Table I.

Table I. Heritability Estimates for Perceived Loneliness Reported in Earlier Genetically Informative Studies

Study	Age (in years)	Genetically informative design	Measure of loneliness employed	Types of repeated measures (time/informant)	Heritability coefficient, percentages of total variance (h2)
McGuire and Clifford [2000 ]	9–12	Adoption (biologically related and unrelated siblings)	Loneliness in Children questionnaire [Asher et al., 1984] (8 items)	Longitudinal, self rating	55
	8–14	Twin/sibling		Cross sectional self rating	48
Bartels et al. [2008 ]	7, 10, 12	Classical twin	Child Behavior Checklist [Achenbach, 1991] (2 items)	Longitudinal, maternal rating	45 (total)
					58 (age 7)
					56 (age 10)
					26 (age 12)
Boomsma et al. [2005 ]	18–33	Classical twin	Young Adult Self Report YASR [Achenbach, 1990] (6 items)	Longitudinal, self rating	48
Boomsma et al. [2006 ]	18–33	Twin/sibling	Young Adult Self Report YASR [Achenbach, 1990] (2 items)	Longitudinal, self rating	40
Distel et al. [2010 ]	16–90	Extended twin-family (twins, spouses of twins, siblings, parents)	R-UCLA Loneliness Scale [Russel et al., 1980 ] (3 items)	Cross sectional, self rating	18 (additive)
					19 (non-additive, dominance)

There were no marked effects of sex, age, or sociodemographic variables in the genetic architecture found in these studies. However, the longitudinal twin study by [Bartels et al., 2008] showed a drop in the contribution of genetic factors from childhood to the outset of adolescence, from 58% additive genetic effects in at age 7 years to 26% additive genetic effect at age 12. Thus, there is some indication that there may be a drop in the heritability of loneliness in adolescence. So far, however, there has been a lack of genetically informative studies documenting the causal structure behind loneliness in adolescents.

The aim of the present study was to investigate the relative contribution of genetic and environmental etiological sources explaining variability of perceived loneliness in adolescents. Data were collected in a population-based survey of seven national Norwegian cohorts of monozygotic and dizygotic adolescent twins of both sexes reared together. A multi informant (mothers, fathers, and twins' self ratings) approach was chosen, due to its advantage of offering an error free estimate of the relative impact of the common etiological sources as well as to sources specific for each informant on the trait in focus [Hewitt et al., 1992; Bartels et al., 2007]. The primary hypothesis, based on heritability estimates reported in earlier studies on children and adult samples, was that perceived loneliness in adolescents would be moderately heritable. Previous genetically informative studies have indicated that although there may be sex differences in the prevalence of loneliness, the causal architecture seems not to differ between males and females [Bartels et al., 2008]. Thus, our hypothesis was that there would be no differences in heritability between the sexes.

MATERIALS AND METHODS

Participants and Procedure

The sample consisted of 1,394 twin pairs (25.5% of total twin population in the relevant cohorts, 56.2% of available pairs) and parents who participated in the study based on a mailed invitation to seven full cohorts of twins born between 1988 and 1994. Twin status and addresses were provided by the national Norwegian Medical Birth Registry. Twins, mothers and fathers answered paper-and-pencil inventories that were returned by mail. More details on sample characteristics, including study attrition and procedure, can be found in Waaktaar and Torgersen [2011].

Measures

Perceived loneliness

Perceived loneliness (in the following abbreviated PLON) was measured by 5 items from the original 20 items revised UCLA Loneliness scale by Russel et al. [1980] and Roberts et al. [1993]. The items were: (1) I feel in tune with people around me; (2) I can find companionship when I want it; (3) No one really knows me well; (4) People are around me but not with me (5) I feel alone. Item 1 and 2 were reverse coded. Twins' self-rating forms and parental forms were equal except from the substitution of “I” and “me” in the twins' forms with “The twin” and “his/her” in the parental forms. Items were scored on a five-point scale from “Not typical” to “Very typical.” Final scale inter-item reliability Cronbach's alpha across informants ranged from 0.78 (twins' self ratings) to 0.87 (mothers' ratings).

Zygosity

Zygosity was determined through a combination of a questionnaire testing for twin physical similarity that was answered by the total sample, and DNA secured through cheek swabs from 15% of the sample [Waaktaar and Torgersen, 2011]. Twins who were not tested on DNA were allocated to zygosity group by means of discriminant analysis of the questionnaire data from each twin, mother, and father. The misclassification following this procedure was estimated to be <2%.

Data Analyses

Data preparation

Scale distributions were significantly negatively skewed (the ratio of skewness/standard error of skewness >−2) on several of the PLON scale scores, and square root transformations were therefore performed for these variables before entered into the genetic analyses. Separate linear regression analyses within informant and twin groups showed no significant effects of age on the PLON scores. However, a significant effect of sex on the means structure was later controlled for through the use of residualized scores in all genetic analyses. Preliminary univariate analyses for each informant separately using a correlated factors model showed no deviation from the basic assumption of equality of means and variances across twins and across zygosity groups in any of the informants' scores (results of data preparation analyses are available from the first author upon request).

Means and standard deviations were calculated using the SPSS Statistical Package, version 18.01. All further data analyses were performed using the open source statistical software package R, version 2.13.2 [R Development Core Team, 2008]. All genetic modeling analyses were performed using the OpenMx version 1.1.2 [Boker et al., 2011], a package within the R software for fitting Structural Equation Models (SEM) to observed data [R Development Core Team, 2008].

Genetic modeling

A fundamental assumption of the twins reared together design is that monozygotic (MZ) twins have (virtually) all their genetic composition in common, while dizygotic (DZ) pairs on the average share half of their segregating genes. This implies that any phenotypic difference between MZ twins must be due to the effect of unique environmental factors and/or measurement error. Phenotypic MZ correlations exceeding those of DZ twins are indications of genetic effects. DZ correlations more than half the size of the MZ correlations will be caused by shared environmental factors, while DZ correlations smaller than 50% of the MZ correlations indicate an effect of non-additive genetic factors, attributable to either dominance (alleles interacting within a particular locus) or epistasis (alleles interacting across different loci). A further characteristic of the classic twin design is that the effects of non-additive genetic and common environmental parameters cannot be tested within the same model. Thus, non-additive effect models are normally considered when the MZ correlations exceed the DZ correlations by more than 50%. Two further assumptions must be met for the twin model to be valid: (i) no assortative mating for the phenotype measured; and (ii) MZ and DZ twins are equally exposed to the relevant environmental stimuli for the trait studied. There is generally little empirical evidence of violations of these assumptions [Neale et al., 1998; Bulik et al., 2000].

The etiological structure behind the twin associations was analyzed using a biometrical modeling approach [Neale and Cardon, 1992] where covariances based on raw data are fitted to a structural equation model through maximum likelihood estimation (FIML). Alternative models were compared using the Akaike information criterion (AIC) [Akaike, 1987]. This fit statistic takes into account the overall fit as well as the parsimony of the model, and lower numbers signify improved model fit.

Model fitting

The genetic models employed in the present study were all variants of the psychometric model [Rijsdijk and Sham, 2002] (Fig. 1).

Details are in the caption following the image — **Figure 1**
Open in figure viewer PowerPoint

Full psychometric model for both twins (A = additive genetic factors; C = common environmental factors; E = non-shared environmental factors. Subscripts: 1 = twin 1; 2 = twin 2; m = mothers; f = fathers; t = twins. 0.5/1.0 = correlations for dizygotic and monozygotic twins, respectively according to model).

When applied in the multi-informant case, the psychometric model assumes that, in addition to contributing to set of common genetic, common shared environmental and common nonshared environmental sources explaining the variance of a common latent psychometric factor, each informant will constitute a unique source of variation that may meaningfully be partitioned into informant-specific genetic, shared environmental, and non-unshared parts [Baker et al., 2007; Bartels et al., 2007]. The relevance of a model that takes into account that not all variation between informants may be due to measurement error or rater bias but allows for the possibility that mothers, fathers and adolescents provide substantive additional information about an adolescent's behavior, has been demonstrated in several former studies [Scourfield et al., 2004; Hoekstra et al., 2008; Tackett et al., 2009; Waaktaar and Torgersen, 2011].

It is worth noticing that the common latent psychometric factor in this model will be unaffected by informant-specific bias and measurement error. Thus, the nonshared environmental pathway on the common perceived loneliness factor includes only effective environmental influences that make the twins different from each other. Measurement error will in this model be contained together with informant-specific environmental effects within the nonshared environmental pathways on the informant-specific perceived loneliness factors [Neale and Cardon, 1992]. This implies that the common factor effects on the latent psychometric loneliness factor will provide highly reliable estimates of the effect of the different etiological sources on perceived loneliness in adolescents.

RESULTS

Descriptives

Group means and standard deviations

Table II shows sample means and standard deviations of (unresidualized) scale scores by gender, zygosity groups, and informants. There were several significant differences between informants' scores in all zygosity groups, with higher perceived loneliness scores generally endorsed by twins and fathers than by mothers, and with girls scoring higher on perceived loneliness than boys. Based on these and other preliminary univariate analyses (Data preparation section), the further multivariate genetic analyses were performed with scale scores residualized by sex, and the biometric models were fitted without restricting means to be equal across informants.

Table II. Descriptive Data of Informants (Mothers, Fathers, and Twins) Ratings by Zygosity Group

Zygosity group	Informant	N	Mean	SD
MZ Males (222 twin pairs)	Mother	425	3.89	3.23
	Father	317	4.15	3.09
	Twins	415	4.85	3.19
MZ Females (314 twin pairs)	Mother	578	4.79	3.58
	Father	414	4.84	3.07
	Twins	599	5.35	3.77
DZ Males (203 twin pairs)	Mother	382	3.66	2.98
	Father	288	4.24	3.13
	Twins	371	4.83	3.28
DZ Females (250 twin pairs)	Mother	455	4.65	3.62
	Father	330	4.53	3.24
	Twins	475	5.46	3.66
DOS (405 twin pairs)	Mother	781	4.19	3.41
	Father	582	4.54	3.02
	Twins	747	5.09	3,40

MZ, monozygotic; DZ, dizygotic; DOS, dizygotic, opposite sex.

Correlation structure

Within person correlations across informant's scores, as well as cross person within informant and cross informant correlations for each zygosity groups (MZ, DZ) and each sex are shown in Table III. The correlations were generated through ML estimation, and were thus built on the same logic that constitutes the basis for the later the twin models.

Table III. Estimated Twin Correlations Across Sex for the Different Informant Scores and the Two Zygosity Groups

	Mo	Fa	Tw		Male pairs			Female pairs			Opposite sex pairs
	Within person correlations				Cross person correlations: Within informant (diagonal) and cross informant (off diagonals)
	Mo	Fa	Tw		Mo	Fa	Tw	Mo	Fa	Tw	Mo	Fa	Tw
				Mother
				MZ	0.636			0.685			—
				DZ	0.374			0.343			0.291
Father				Father
Males	0.446			MZ	0.333	0.714		0.289	0.638		—	—
Females	0.439			DZ	0.095	0.350		0.172	0.324		0.115	0.344
Twins				Twins
Males	0.361	0.256		MZ	0.307	0.264	0.419	0.366	0.247	0.526	—	—	—
Females	0.434	0.321		DZ	0.208	0.091	0.290	0.199	0.174	0.286	0.071	0.059	0.059

Mo, mother's rating; Fa, father's rating; Tw, Twins' self rating.

The correlation matrix shows that the within pair MZ correlations were higher than DZ correlations in all three informants and for both sex groups. Thus, genetic variance was indicated for all informant scores. The MZ twin correlations were generally <50% above the size of the DZ correlations, which could signify that the ACE model would be more adequate than the ADE model in the subsequent genetic modeling analyses. There were some minor differences in the size of the MZ/DZ correlations between same-sex boys and same-sex girls. However, apart from the markedly lower correlations for the twins scores in the opposite sex group in relation to all informants, the correlation between opposite sex DZ twins were generally not of lesser scale than correlations for same sex DZ twins. Thus, based on the cursory view of the correlation matrix, any sex-specific effects were not likely to be of large scale. Although the cross informant-cross person correlations (off diagonal entries of Table III) were generally lower than the within informant-cross person correlations (diagonal entries), the relative MZ/DZ pattern from the cross informant-within person correlations was maintained in the cross informant-cross person correlations. This suggests that that there may be genetic contributions to associations between informants. The cross informant-within person correlations (left side of table off diagonal correlations) were in the range of medium to low, indicating informant specific as well as common effects.

Multivariate Model Testing

The full ACE psychometric measurement model with heterogeneity (allowing for sex differences in the estimates) constituted the baseline model from which to test the fit of more stringent models. The ACE model with no sex limitation did not result in significantly worse fit (Δ-2LL = 28.07, Δdf = 15, P = 0.02) compared to the full multivariate measurement model heterogeneity ACE model. Thus, the psychometric ACE homogeneity model was chosen as the starting point for the subsequent testing of submodels. The results of these analyses are shown in Table IV.

Table IV. Fit statistics for Different Models Under the Psychometric ACE Model Without Sex Differences

#	Latent			Mothers			Fathers			Twins			Δχ²	Δdf	P-value	ΔAIC
#	A	C	E	A	C	E	A	C	E	A	C	E	Δχ²	Δdf	P-value	ΔAIC
I	+	+	+	+	+	+	+	+	+	+	+	+	—	—	—	—
II	−	+	+	+	+	+	+	+	+	+	+	+	56.77	1	0	54.76
III	+	+	+	−	+	+	−	+	+	−	+	+	57.05	3	0	51.05
IV	−	+	+	−	+	+	−	+	+	−	+	+	57.05	4	0	142.17
V	+	−	+	+	+	+	+	+	+	+	+	+	0	1	1	−2
VI	+	+	+	+	−	+	+	−	+	+	−	+	0.91	3	0.82	−5.09
VII	+	−	+	+	−	+	+	−	+	+	−	+	0.91	4	0.92	−7.09a
VIII	+	+	−	+	+	+	+	+	+	+	+	+	56.77	1	0	54.76
IX	−	−	+	+	+	+	+	+	+	+	+	+	210.42	2	0	206.42
X	−	−	−	+	+	+	+	+	+	+	+		1992.49	3	0	1986.48

a Best fitting model.

In the sequence of submodel testing shown in Table IV, each of the common and specific etiological factors were dropped individually and then together from full ACE model. Based on this strategy, Common C and specific C could both be removed without any significant drop in fit compared to the full psychometric model (Table IV, Models V–VII). Model VI (the psychometric ACE model with one common latent factor and no common and no specific C) was the most parsimonious model as judged by the lowest AIC value, and thus ended up as the preferred model.

Figure 2 shows the standardized and squared path estimates for all factors in the final model, with confidence intervals. “Perceived loneliness” was here modeled by means of a common latent psychometric factor with a strong additive genetic component (3/4 of the common factor variance) and with the remaining 1/4 of the common factor variance attributable to nonshared environment. The shared environmental effect was non-significant. The figure also shows that the effect of the common latent Perceived Loneliness factor was largest in the mothers' ratings, and somewhat smaller in the scores of the fathers and the twins self ratings. The informant-specific genetic effects were strongest in the fathers' scores, while the informant-specific environmental effects were strongest in the twins' scores.

A further visualization of these results is given in Table V, showing the proportion of the variance in each informants' scores that was explained by common and rater-specific effects.

Table V. Variance Estimates in Final Model by Informant

	Total heritability	Proportion (%) due to Ac	Proportion (%) due to As	Total environment	Proportion (%) due to Ec	Proportion (%) due to Es
Mothers	0.67	68	32	0.33	46	54
Fathers	0.69	36	64	0.31	27	73
Twins	0.44	46	54	0.56	12	88

Ac = common additive genetic factor; As = informant-specific additive genetic factor; Ec = common nonshared environmental factor; Es = informant-specific nonshared environmental factor.

The total variance estimates for each informant's scores indicate that there was a difference in the relative strength of additive genetic sources and nonshared environmental sources in the twins' scores compared to those of the mothers' and fathers'. While between two thirds and three quarters of the mothers' and fathers scores were explained by additive genetic factors and some remaining non-shared environmental factors, the additive genetic and nonshared environmental sources of variance in the twins' Perceived Loneliness scores were almost of equal strength.

Furthermore, a very high proportion of the additive genetic variance in the mothers' scores was due to additive genetic variance in the common factor. The twins' additive genetic variance was about half due to variance in the common factor, half due to additive factors specific for the twins' scores. The additive genetic variance in the fathers' ratings was mainly rater specific. The main proportion of the total non-shared environmental factors was due to factors specific for each informant. The specific environmental variance was strongest in the twins' scores, somewhat smaller in the fathers' scores and smallest in the mothers' scores. In interpreting this result one must keep in mind that any measurement error in this model will be allocated to, and cannot be separated from, the effect of the specific non-shared environment factor.

DISCUSSION

The aim of this multi-informant twin study was to provide reliable estimates of the relative impact of genetic and environmental causes of variation in perceived loneliness in adolescents. Biometric analyses were based on an ACE measurement model approach assuming that the covariation between the different informants' scores for both sexes could be modeled by means of a latent psychometric factor, where additive genetic and unique environmental factors influenced the different informants' scores through the same mechanisms. The model also allowed for additive genetic, shared, and nonshared environmental sources influencing each informant's scores specifically.

The results showed that additive genetic factors explained 75% of the variation in the common latent psychometric Perceived Loneliness factor. The remaining variation (25%) in the latent Perceived Loneliness factor was explained by unique, nonshared environmental factors. The impact of shared environmental factors was negligible. There were no significant sex differences in the estimates in the model.

This is the first twin study to our knowledge to report heritability estimates of perceived loneliness in adolescents. We found higher genetic effects on loneliness in adolescents than has been reported in earlier studies based on children and on adults [McGuire and Clifford, 2000; Boomsma et al., 2005; Distel et al., 2010]. The heritability estimate of perceived loneliness found in the present study also exceeds what could be expected based on the results from an earlier longitudinal study by Bartels et al. [2008], which suggested a possible decline in heritability from childhood to adolescence. One likely explanation is that it may be due to the multirater design, which is able to produce more reliable estimates than those based on any one single rater. The lack of significant sex differences in the estimates of the psychometric factor model of perceived loneliness in this study is in accordance with previous the results reported in the studies.

The heritability estimates in the present study varied between raters (from explaining 2/3 of the variance in the mothers' and fathers' ratings to less than half of the twins' self ratings). All informants' scores loaded substantially on the latent common genetic factor; thus there was support for an additive genetic factor that was represented in all informant's ratings of the adolescents' perceived loneliness. In addition, there was evidence of genetic effects contributing to differences between each informants' ratings. These effects were most pronounced in the fathers' and twins' scores, where more than half of the genetic variance was explained by informant-specific genetic effects. Rater-specific genetic effects contributed to only less than one-third of the total genetic variance in the mothers' scores.

The main part of the genetic variation in the latent perceived loneliness factor came from additive genetic sources, that is, contributions of genes that are independent of each other. Non-additive effects have been reported in other personality traits in adolescents [Rebollo et al., 2006; Rettew et al., 2008; Waaktaar and Torgersen, 2011]. Although dominant genetic effects were not specifically tested in this study due to the confounding of the shared environment effects in the classic twin reared together design [Rijsdijk and Sham, 2002], the correlations matrix did not indicate any effect of different genes interacting with each other.

The results of the present study showed that only one quarter of the total variation in loneliness was attributable to environmental factors. With no significant shared environmental effects in the present data, all environmental variation in the latent perceived loneliness factor must have originated from sources that are not shared between the twins within the same family (the error-free common E in the model). This is in accordance with what is generally found for most complex human traits [Plomin et al., 2001]. However, among such potential influences within the family, only those that in effect would make twins within the same family different from one another on perceived loneliness could be of significance.

Limitations

Some limitations should be kept in mind when evaluating the results of this study.

The twins reared together design generally yields common environmental estimates that are lower than those reported in adoption studies [Buchanan et al., 2009].

As published in an earlier article based on the same sample [Waaktaar and Torgersen, 2011], a comparison between level of education in the participating families to age equivalent levels for women and men within the total population indicated a possible selection bias in the participating families based on socioeconomic indicators. Differential heritability estimates across socioeconomic groups have been reported in genetically informative studies have yielded [Tuvblad et al., 2006; South and Krueger, 2011] as well as in groups with different exposure to various stressful conditions [Distel et al., 2011]. Thus, the possibility that this type of bias may have influenced the heritability estimates in the present study cannot be ruled out.

Measurement invariance may be an issue in studies where different groups are being compared on measures of complex traits. We could argue that one type of measurement non-invariance is actually demonstrated in the multiple raters scores in the present study, expressed through the differences in latent factor loadings between informants assessing twins' perceived loneliness. In twin studies, measurement non-invariance at item level between zygosity groups may impact the estimates in studies based on sum scores [Neale et al., 2005]. The problem of item level non-invariance would be highest in the case of binary items, and the direction of the bias on the variance estimates would depend upon which of the MZ or DZ scores were the best indicators of the latent trait in question. Preliminary exploratory factor analyses (PCA) on the items constituting the loneliness sum scores in the present sample showed preference for a one factor solution for both zygosity groups and across all three informants ratings (factor loadings of each item typically ranging from 0.60 to 0.85). A full multivariate simultaneous analysis of measurement model and variance decomposition analytic approach would provide an even more effective handling of the measurement invariance issue. There are indications from earlier studies on problem behaviors in young people that this kind of approach could yield even higher heritability estimates [van den Berg et al., 2007]. For the target phenotype of perceived loneliness, the usefulness of analyses at this level of precision is questionable.

CONCLUSION

Perceived loneliness in adolescents as measured by three informants (mothers, fathers, and the twins' self ratings) was modeled by means of a common latent factor, with additional genetic, shared environmental factors as well as non-shared environmental factors that were specific for each informants' scores. The results showed no significant sex differences in the estimates within the model. Perceived loneliness as a latent factor was highly genetically determined, with additive genetic effects explaining three quarters of the variance. There were no significant shared environmental factors in perceived loneliness at any level within the model. The heritability of the mothers' and fathers' ratings were large, while in the twins ratings genetic factors only explained less than half of the variance. There were significant informant-specific contributions to the additive genetic effect in all three informants' scores. While measurement error in the psychometric model is absent from the nonshared environmental source in the common latent factor, such error may explain some of the informant-specific nonshared environmental effect. This study demonstrates that a multiple rater approach is useful in the search for the etiological basis of perceived loneliness in young people.

REFERENCES

Aanes MM, Mittelmark MB, Hetland J. 2010. Interpersonal stress and poor health: The mediating role of loneliness. Eur Psychol 15(1): 3– 11.
10.1027/1016-9040/a000003
Web of Science® Google Scholar
Akaike H. 1987. Factor analysis and AIC. Psychometrika 52(3): 317– 332.
10.1007/BF02294359
Web of Science® Google Scholar
Baker LA, Jacobson KC, Raine A, Lozano DI, Bezdjian S. 2007. Genetic and environmental bases of childhood antisocial behavior: A multi-informant twin study. J Abnorm Psychol 116(2): 219– 235.
10.1037/0021-843X.116.2.219
PubMed Web of Science® Google Scholar
Bartels M, Boomsma DI, Hudziak JJ, Van Beijsterveldt TCEM, Van den Oord EJCG. 2007. Twins and the study of rater (Dis)agreement. Psychol Methods 12(4): 451– 466.
10.1037/1082-989X.12.4.451
PubMed Web of Science® Google Scholar
Bartels M, Cacioppo JT, Hudziak JJ, Boomsma DI. 2008. Genetic and environmental contributions to stability in loneliness throughout childhood. Am J Med Genet Part B 147(3): 385– 391.
10.1002/ajmg.b.30608
PubMed Web of Science® Google Scholar
Bernardon S, Babb KA, Hakim-Larson J, Gragg M. 2011. Loneliness, Attachment, and the Perception and Use of Social Support in University Students. Can J Behav Sci 43(1): 40– 51.
10.1037/a0021199
Web of Science® Google Scholar
Boker S, Neale M, Maes H, Wilde M, Spiegel M, Brick T, Spies J, Estabrook R, Kenny S, Bates T, Mehta P, Fox J. 2011. OpenMx: An open source extended structural equation modeling framework. Psychometrika 76(2): 306– 317.
10.1007/s11336-010-9200-6
PubMed Web of Science® Google Scholar
Boomsma D, Willemsen G, Dolan C, Hawkley L, Cacioppo J. 2005. Genetic and environmental contributions to loneliness in adults: The Netherlands Twin Register Study. Behav Genet 35(6): 745– 752.
10.1007/s10519-005-6040-8
PubMed Web of Science® Google Scholar
Boomsma DI, Cacioppo JT, Slagboom PE, Posthuma D. 2006. Genetic linkage and association analysis for loneliness in Dutch twin and sibling pairs points to a region on chromosome 12q23-24. Behav Genet 36(1): 137– 146.
10.1007/s10519-005-9005-z
PubMed Web of Science® Google Scholar
Buchanan JP, McGue M, Keyes M, Iacono WG. 2009. Are there shared environmental influences on adolescent behavior? evidence from a study of adoptive siblings. Behav Genet 39(5): 532– 540.
10.1007/s10519-009-9283-y
PubMed Web of Science® Google Scholar
Bulik CM, Sullivan PF, Wade TD, Kendler KS. 2000. Twin studies of eating disorders: A review. Int J Eat Disord 27(1): 1– 20.
10.1002/(SICI)1098-108X(200001)27:1<1::AID-EAT1>3.0.CO;2-Q
CAS PubMed Web of Science® Google Scholar
Cacioppo JT, Hawkley LC, Norman GJ, Berntson GG. 2011. Social isolation. Ann N Y Acad Sci 1231(1): 17– 22.
10.1111/j.1749-6632.2011.06028.x
PubMed Web of Science® Google Scholar
Caspi A, Harrington H, Moffitt TE, Milne BJ, Poulton R. 2006. Socially isolated children 20 years later: risk of cardiovascular disease. Arch Pediatr Adolesc Med 160(8): 805– 811.
10.1001/archpedi.160.8.805
PubMed Web of Science® Google Scholar
Cole SW, Hawkley LC, Arevalo JM, Sung CY, Rose RM, Cacioppo JT. 2007. Social regulation of gene expression in human leukocytes. Genome Biol 8(9): R189.
10.1186/gb-2007-8-9-r189
CAS PubMed Web of Science® Google Scholar
de Jong-Gierveld J. 1987. Developing and testing a model of loneliness. J Pers Soc Psychol 53(1): 119– 128.
10.1037/0022-3514.53.1.119
CAS PubMed Web of Science® Google Scholar
Distel MA, Rebollo-Mesa I, Abdellaoui A, Derom CA, Willemsen G, Cacioppo JT, Boomsma DI. 2010. Familial resemblance for loneliness. Behav Genet 40(4): 480– 494.
10.1007/s10519-010-9341-5
PubMed Web of Science® Google Scholar
Distel MA, Middeldorp C, Trull T, Derom C, Willemsen G, Boomsma D. 2011. Life events and borderline personality features: the influence of gene–environment interaction and gene–environment correlation Psychol Med 41(4): 849– 860.
10.1017/S0033291710001297
CAS PubMed Web of Science® Google Scholar
Fontaine RG, Yang C, Burks VS, Dodge KA, Price JM, Pettit GS, Bates JE. 2009. Loneliness as a partial mediator of the relation between low social preference in childhood and anxious/depressed symptoms in adolescence. Dev Psychopathol 21(2): 479– 491.
10.1017/S0954579409000261
PubMed Web of Science® Google Scholar
Groholt BMD, Ekeberg OMD, Wichstrom LPD, Haldorsen TORMS. 2000. Young suicide attempters: a comparison between a clinical and an epidemiological sample. J Am Acad Child Adolesc Psychiatry 39(7): 868– 875.
10.1097/00004583-200007000-00015
CAS PubMed Web of Science® Google Scholar
Hawkley LC, Preacher KJ, Cacioppo JT. 2010a. Loneliness impairs daytime functioning but not sleep duration. Health Psychol 29(2): 124– 129.
10.1037/a0018646
PubMed Web of Science® Google Scholar
Hawkley LC, Thisted RA, Masi CM, Cacioppo JT. 2010b. Loneliness predicts increased blood pressure: 5-year cross-lagged analyses in middle-aged and older adults. Psychol Aging 25(1): 132– 141.
10.1037/a0017805
PubMed Web of Science® Google Scholar
Hewitt JK, Silberg JL, Neale MC, Eaves LJ, Erickson M. 1992. The analysis of parental ratings of childrens behavior using lisrel. Behav Genet 22(3): 293– 317.
10.1007/BF01066663
PubMed Web of Science® Google Scholar
Hoekstra RA, Bartels M, Hudziak JJ, Van Beijsterveldt TCEM, Boomsma DI. 2008. Genetic and environmental influences on the stability of withdrawn behavior in children: A longitudinal, multi-informant twin study. Behav Genet 38(5): 447– 461.
10.1007/s10519-008-9213-4
PubMed Web of Science® Google Scholar
Kim J, LaRose R, Peng W. 2009. Loneliness as the cause and the effect of problematic Internet use: The relationship between Internet use and psychological well-being. Cyberpsychol Behav 12(4): 451– 455.
10.1089/cpb.2008.0327
PubMed Google Scholar
Kunst MJ, van Bon-Martens MJ. 2011. Examining the link between domestic violence victimization and loneliness in a Dutch community sample: A comparison between victims and nonvictims by Type D personality. J Fam Violence 26(5): 403– 410.
10.1007/s10896-011-9374-4
PubMed Web of Science® Google Scholar
Lasgaard M, Nielsen A, Eriksen ME, Goossens L. 2010. Loneliness and social support in adolescent boys with autism spectrum disorders. J Autism Dev Disord 40(2): 218– 226.
10.1007/s10803-009-0851-z
PubMed Web of Science® Google Scholar
Lasgaard M, Goossens L, Elklit A. 2011. Loneliness, depressive symptomatology, and suicide ideation in adolescence: Cross-sectional and longitudinal analyses. J Abnorm Child Psychol 39(1): 137– 150.
10.1007/s10802-010-9442-x
PubMed Web of Science® Google Scholar
Mahon NEYA, Yareski TJ, Cannella BL, Hanks MM. 2006. A meta-analysis of predictors of loneliness during adolescence. Nurs Res 55(5): 308– 315.
10.1097/00006199-200609000-00003
PubMed Web of Science® Google Scholar
McGuire H, Clifford J. 2000. Genetic and environmental contributions to loneliness in children. Psychol Sci 11(6): 487– 491.
10.1111/1467-9280.00293
CAS PubMed Web of Science® Google Scholar
Neale MC, Cardon LR. 1992. Methodology for genetic studies of twins and families. Dordrecht, The Nertherlands: Kluwer Academics.
10.1007/978-94-015-8018-2
Web of Science® Google Scholar
Neale MC, Kendler KS, Hewitt JK, Silberg JL, Foley DL, Meyer JM. 1998. Assortative mating for major psychiatric diagnoses in two population-based samples. Psychol Med 28(6): 1389– 1401.
10.1017/S0033291798007326
PubMed Web of Science® Google Scholar
Neale MC, Lubke G, Aggen SH, Dolan CV. 2005. Problems with using sum scores for estimating variance components: Contamination and measurement noninvariance. Twin Res Hum Genet 8(6): 553– 568.
10.1375/twin.8.6.553
PubMed Web of Science® Google Scholar
Pinquart M, Sörensen S. 2003. Risk factor for loneliness in adulthood and old age—A meta-analysis. In: SP Shohov, editor. Advances in psychology research. Hauppauge, NY: Nova Science Publishers. p 111– 143.
Google Scholar
Plomin R, Asbury K, Dunn J. 2001. Why are children in the same family so different? Nonshared environment a decade later. Can J Psychiatry 46(3): 225– 233.
10.1177/070674370104600302
CAS PubMed Web of Science® Google Scholar
Poulsen AA, Ziviani JM, Cuskelly M, Smith R. 2007. Boys with developmental coordination disorder: Loneliness and team sports participation. Am J Occup Ther 61(4): 451– 462.
10.5014/ajot.61.4.451
PubMed Web of Science® Google Scholar
Qualter P, Brown SL, Munn P, Rotenberg KJ. 2010. Childhood loneliness as a predictor of adolescent depressive symptoms: An 8-year longitudinal study. Eur Child Adolesc Psychiatry 19(6): 493– 501.
10.1007/s00787-009-0059-y
PubMed Web of Science® Google Scholar
R Development Core Team. 2008. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL: http://www. R-project.org.
Google Scholar
Rebollo IM, Boomsma DI. 2006. Genetic and environmental influences on Type A behavior pattern: Evidence from twins and their parents in The Netherlands Twin Register. Psychosom Med 68(3): 437– 442.
10.1097/01.psy.0000204631.76684.28
PubMed Web of Science® Google Scholar
Rettew DC, Rebollo-Mesa I, Hudziak JJ, Willemsen G, Boomsma DI. 2008. Non-additive and additive genetic effects on extraversion in 3314 Dutch Adolescent Twins and Their Parents. Behav Genet 38(3): 223– 233.
10.1007/s10519-008-9192-5
PubMed Web of Science® Google Scholar
Rijsdijk F, Sham P. 2002. Analytic approaches to twin data using structural equation models. Brief Bioinform 3(2): 119– 133.
10.1093/bib/3.2.119
CAS PubMed Google Scholar
Roberts RE, Lewinsohn PM, Seeley JR. 1993. A brief measure of loneliness suitable for use with adolescents. Psychol Rep 72(3): 1379– 1391.
10.2466/pr0.1993.72.3c.1379
CAS PubMed Web of Science® Google Scholar
Russel D, Peplau LA, Cutrona CE. 1980. The revised UCLA Loneliness Scale: Concurrent and discriminant validity evidence. J Pers Soc Psychol 39: 472– 480.
10.1037/0022-3514.39.3.472
PubMed Web of Science® Google Scholar
Scourfield J, Van den Bree M, Martin N, McGuffin P. 2004. Conduct problems in children and adolescents - A twin study. Arch Gen Psychiatry 61(5): 489– 496.
10.1001/archpsyc.61.5.489
PubMed Web of Science® Google Scholar
Shute RH, Walsh C. 2005. Adolescents with chronic illnesses: school absenteeism, perceived peer aggression, and loneliness. Sci World J 5: 535– 544.
10.1100/tsw.2005.68
PubMed Web of Science® Google Scholar
South S, Krueger R. 2011. Genetic and environmental influences on internalizing psychopathology vary as a function of economic status. Psychol Med 41(1): 107– 117.
10.1017/S0033291710000279
CAS PubMed Web of Science® Google Scholar
Tackett JL, Waldman ID, Lahey BB. 2009. Etiology and measurement of relational aggression: A multi-informant behavior genetic investigation. J Abnorm Psychol 118(4): 722– 733.
10.1037/a0016949
PubMed Web of Science® Google Scholar
Tuvblad C, Grann M, Lichtenstein P. 2006. Heritability for adolescent antisocial behavior differs with socioeconomic status: gene-environment interaction. J Child Psychol Psychiatry 47(7): 734– 743.
10.1111/j.1469-7610.2005.01552.x
CAS PubMed Web of Science® Google Scholar
van den Berg SM, Glas CA, Boomsma DI. 2007. Variance decomposition using an IRT measurement model. Behav Genet 37(4): 604– 616.
10.1007/s10519-007-9156-1
PubMed Web of Science® Google Scholar
Vanhalst J, Klimstra TA, Luyckx K, Scholte RH, Engels RC, Goossens L. 2011. The interplay of loneliness and depressive symptoms across adolescence: Exploring the role of personality traits. J Youth Adolesc DOI: 10.1007/s10964-011-9726-7 [Epub ahead of print].
10.1007/s10964‐011‐9726‐7
Google Scholar
Waaktaar T, Torgersen S. 2011. Genetic and environmental causes of variation in trait resilience in young people. Behav Genet DOI: 10.1007/s10519-011-9519-5 [Epub ahead of print].
10.1007/s10519‐011‐9519‐5
PubMed Web of Science® Google Scholar

Citing Literature

Volume159B, Issue5

July 2012

Pages 580-588

Genetic and environmental causes of variation in perceived loneliness in young people^†

Abstract

INTRODUCTION