Are smart children genetic or "chicken baby"? These twin studies show that it's not simple

In today's era where education is extremely important, parents use all kinds of tricks to improve their children's education (grades), constantly "chicken blood" for their children, jokingly calling them "chicken babies". Some children make rapid progress as a result, while others remain stagnant and are considered by their parents to have average IQs. To what extent is IQ affected by genetics, and how is it affected by environmental factors? This article uses twin studies (twin design) as an example to introduce genetic methods for IQ-related research. We will have a deep understanding of how twin studies can help us understand the inheritance of IQ and what environmental factors can affect IQ levels.
Written by | Shuang Chenyue

The earliest introduction of twins into scientific research can be traced back to British eugenics scientist Francis Galton. Based on scholars' research on twins in the 1870s, Galton advocated the use of twins for comparative studies as new evidence to support biometric identification or eugenics.

Since then, scientists have begun to value twins and regard them as "natural experiments" or "real laboratories", believing that they are a gift from nature to scientific research. This influence is not only reflected in actual observation experiments. Physicist Albert Einstein and philosopher Henri Bergson once used twins to debate the nature of time. In Einstein's thought experiment "Twin Paradox", if one of a pair of twins enters the universe at a speed close to the speed of light, when he returns to Earth, he will be younger than the other twin who stayed at home. Bergson's rebuttal is that the concept of time cannot be so simply understood in a scientific definition. This debate attracted many viewers and made more people realize that twin narratives can be used to explore more truths about human existence [1].

Figure 1 Velvet Twins - Oil painting on linen | Source: Reference [2]

The complex issue of IQ inheritance

Stories about twins always attract attention and trigger deeper thinking. In 2011, when the "best study schedule" became popular on the Internet, the twin sisters who received the "Special Scholarship" from Tsinghua University attracted widespread attention. Many people were curious: Is their excellence due to high IQ in their genes, or is it because of the superior family and school environment they grew up in?

Over the past century, psychologists have conducted many studies on this topic, especially experiments based on twin designs, which have provided a lot of evidence for the "nature or nurture" debate about IQ.

Despite its conceptual simplicity, the heritability of IQ has been the subject of intense debate over the last century. This is because, due to differences in measurement methods and definitions of heritability, the broad-sense and narrow-sense heritability results for IQ are inconsistent [3], with narrow-sense heritability generally being lower than broad-sense heritability. In addition, the genetic impact on IQ varies with age. Overall, most researchers have found that genetic factors contribute approximately 30% to 70% of the variance in IQ scores in the general population over a lifetime. Why is there such a range? Many classic studies have combined twin designs with other behavioral genetics methods to reveal more complex biological mechanisms underlying this issue.

In a classic study published in Science in 1981, the authors summarized 111 IQ heritability surveys from around the world [4] and compared the IQ correlation between biological parents and children and between adoptive parents and adopted children (no genetic relationship). The study found that the intelligence correlation between biological parents and children with a blood relationship is about 1 times higher than that between adopted parents, and the intelligence correlation between biological parents and children is about 0.4, which is at a medium level of correlation. However, subsequent studies using twin samples combined with other behavioral methods have made more specific findings on the impact of genetics and environment on IQ; in particular, the role of many environmental factors in the development of IQ cannot be ignored.

To better understand the relevant research, let us first look at the research method using twins as samples - the twin design, as well as other classic methods related to behavioral genetics.

Twin Design and Genetic Methods

In academia, studies using twins as experimental subjects are often referred to as twin designs, and many of these studies have been a valuable source of information for exploring the genetic basis of complex traits. We know that because relatives share some genetic genes and living environment to varying degrees, their differences in specific aspects may come not only from genes but also from the shared environment in which they live together. Researchers use twin or family studies to estimate the relative influence of genes and environment. The underlying logic of these studies is simple[2]: to the extent that a trait is influenced by a specific genetic architecture, individuals with more corresponding genetic information will have a higher degree of similarity in that trait; this phenomenon is more obvious when relatives with different genetic similarities are in similar environments.

The twin design compares the differences in traits between adult identical twins and fraternal twins to explore whether the family environment directly affects the next generation or whether genetic genes play a greater role [2, 3. 5]. Its basic principle is to compare the similarities of two different types of twins and other relatives in a certain characteristic that is jointly affected by genetics and the environment.

Identical twins (MZ twins, MZ for short) from the same family are derived from the same fertilized egg (100% shared genes), and share exactly the same genes and family environment. Fraternal twins (DZ twins, DZ for short) are the same age and grew up in the same family environment, but only half of their genetic material is the same on average (50% shared genes), and they are essentially the same genetically similar as ordinary siblings. Therefore, identical twins are usually more similar than fraternal twins, and due to the different degrees of similarity in the environment, fraternal twins are often more similar than ordinary siblings.

In the simplest twin design, two identical (MZ) or fraternal (DZ) twins from the same family share family and other background environmental influences (such as lifestyle, neighbors, etc.), so there should be no major differences in the degree of environmental similarity between the two twin types. Based on this, researchers define the genetic influence on a behavioral trait as the degree to which the correlation between identical (MZ) twins on that phenotype is higher than the correlation between fraternal (DZ) twins. Through such comparisons, researchers can estimate the heritability (h2) of a particular trait, that is, the proportion of population variability caused by genetic influences.

Figure 2 Examples of commonly used models for twin family data[5]

Behavioral genetics studies the influence of genetics and the environment on individual differences in people's behavior. For more than half a century, studies using twins as samples have occupied a central position in behavioral genetics. Since the late 1980s, the main goal of behavioral genetics research has focused more on the molecular basis of heritability. Many researchers have realized the value of large-scale data collection for molecular genetics research and have begun to introduce samples such as siblings, parents, and offspring of twins into their research projects [2, 5].

In particular, in the past decade, the combination of genome-wide association studies (GWASs) and twin studies has ushered in a new era of scientific discovery and methodological development in behavioral genetics [5]. The GWAS method constructs a polygenic score (PGS) in large-scale data to assess the genetic predisposition of an individual or a certain characteristic, or the risk of a certain disease. It is a direct quantitative tool that can reveal the molecular basis of genetic processes shared across generations (from parents to offspring). Twin studies can meet statistical requirements with smaller sample sizes and conduct heritability and genetic correlation analyses; they have greater diversity in the assessment of existing phenotypes and can usually explore more subtle and detailed hypotheses than GWAS data. The genetic differences revealed by twin/family studies can provide information for the design and interpretation of GWAS. For example, in twin/family studies, behavioral measurement variables with very high genetic correlations (a typical example is IQ and education level) can be "lumped" together to increase the validity of the GWAS, while measurement variables with low genetic correlations can be "split" to make the results of the GWAS more specific.

In recent years, due to the different advantages of twin/family designs and GWAS, and their large degree of complementarity, the combination of these two methods has enabled researchers to tease out more precise genetic effects. The twin design based on twins combined with methods similar to GWAS in genetics and many different fields such as behavioral research has provided scientists with a lot of evidence for the genetic-environment debate on differences between people.

Twin studies linking IQ

Many researchers believe that part of the reason why IQ heritability is so controversial is that the definition of IQ heritability is not well characterized and does not use a unified measurement method [3]. For example, the commonly used direct analysis method and indirect analysis method, due to the different ways of dealing with the influencing factor of "environment", they produce significantly different estimates of the heritability of IQ. The former usually produces significantly higher estimates of IQ heritability. A classic study published in Nature [3] hypothesized that this contradiction can be largely attributed to the failure to take into account the influence of maternal-related factors (such as the prenatal uterine environment) on IQ. Therefore, the separation of maternal effects may allow for a more precise exploration of the genetic mechanism of IQ to solve the mystery of IQ.

This study evaluated an alternative model that replaces the three different environments assumed in the traditional twin design - twins, ordinary siblings, and specific environments for parents and offspring - with two maternal womb environments (one for twins and the other for siblings) and a common family environment. Based on a meta-analysis of 204 correlations among 50,470 different pairs, including zero-degree relatives (identical twins), first-degree relatives (i.e., parents, children, and siblings, including fraternal twins), or their adopted partners, the researchers found that the "mother effect" model fits the data better than the previous "family environment" model and can explain 20% of the covariance between twins and 5% of the covariance between siblings [3]. The results show that the broad heritability of IQ may be 48%; the narrow heritability is about 34%. These results not only suggest that we may have a new and more complex model of the influence of genes and environment on cognitive function; they also indicate that interventions to improve the prenatal environment may help significantly improve the IQ level of the next generation.

Many studies have demonstrated the importance of the maternal uterine environment [3]. The brain's nervous system grows rapidly during the embryonic period in the uterus, and the brain reaches 70% of its final mass within one year after birth. Other related studies have also shown that IQ can be affected by the prenatal environment: IQ is positively correlated with birth weight, and a large-scale study of twins in the mid-20th century showed that their IQ was 4 to 5 points lower than that of only children; some dietary supplements during pregnancy can increase a child's IQ; and the use of alcohol, drugs, and cigarettes during pregnancy may reduce the IQ of the future child; lead exchange between the mother and the fetus may reduce IQ, and so on.

Other researchers have explored from the perspective of longitudinal development and found that the influence of genes increases with age. A survey of 11,000 pairs of twins from four countries found that the heritability of general cognitive ability increased significantly linearly from 41% in childhood (9 years old) to 55% in adolescence (12 years old), and reached 66% in early adulthood (17 years old) [6]. Why does the "arrow of fate" in life seem to be increasingly determined by genes? Researchers believe that for many characteristics, the influence of genes and the environment is often not dominated by a single factor or black and white. Results from thousands of twin studies show that most behavioral traits are partially heritable, but are also significantly affected by the environment.

A classic finding is that scientists have observed that genetic predispositions are often also correlated with environmental factors, a phenomenon collectively known as gene-environment correlation (rGE)—that is, as children grow up, they increasingly choose, change their behaviors, and even, to some extent, create their own experiences based on their genetic predispositions. Parents’ genes may work together to influence the traits of their offspring through parental characteristics (such as socioeconomic status, IQ, and education) [6, 7]. For example, children with musical talent may inherit both genes and environments from their parents that are more conducive to the development of musical ability relative to other children—such as musical instrument performance talent and exposure to a musical environment.

With the development of gene-related technologies, a study in 2019 used the polygenic score (PGS) method of genome-wide association study (GWAS) in combination with twin design to test the influence of parental genes on the education and cognition of offspring[8]. The study included the latest GWAS data on education level of 1.1 million people. The polygenic score (PGS) constructed based on this large sample was able to predict 11%-13% of the difference in education level and 7%-10% of the difference in cognitive ability[9, 10], and further proved that education level is a feasible indicator for examining cognitive ability. In the results of this study, the three factors with the greatest impact on the IQ score of offspring were first the parental IQ score (correlation of 0.48), followed by the parental education level (correlation of 0.33) and socioeconomic status (SES, correlation of 0.30). This shows that in addition to the inheritance of parental genes, other important environmental factors, such as the parental education level and SES, also have a great impact on the IQ level of the next generation. In addition, other studies in the United States have found that the heritability of IQ and academic achievement appears to be higher under high SES conditions [10, 11].

Therefore, we can see that for parents who want their children to be successful, if they want to raise an excellent child with high IQ, it is not enough to just spend money on "chickening" their children to enroll in classes or buy school district housing. The IQ of a child is closely related to the following factors: First, the IQ genes of the parents have a moderate impact on the development of the child's IQ and academic performance; second, the influence of the mother on the child will be higher than that of the father. The IQ, education and nutritional status of women during pregnancy have a huge impact on the child's future IQ potential; finally, it is better to "chicken yourself" first. The correlation between the parents' education level and socioeconomic status and the child's IQ level is second only to the parents' IQ score. When the socioeconomic status is better, children are more able to exert their inherited intellectual potential in academics and IQ. From a more macro perspective, it is also extremely important to promote gender equality and improve women's educational opportunities and academic qualifications for the future talent training in our country's development strategy.

Other Methods and Applications of Twin Studies

Moving beyond the simplest twin design, many subsequent studies have evolved into more complex twin models based on their different research purposes and experimental data [2, 12], and scientists have also made many interesting discoveries.

For example, multivariate designs, which generalize univariate twin studies to multiple variables and analyze more than one phenotype per subject, can investigate associations between traits and causes of comorbidity [2, 5]. Questions that might be asked in a multivariate design include: Do changes in exercise lead to changes in depression? Do different traits cluster together because they are influenced by a common set of genes?

In addition, the CoT method (children-of-twins) has become a classic in recent years: a research paradigm that uses twin samples that have already started a family and raised the next generation, and can separate genetic influences from environmental influences [12]. CoT studies have become increasingly popular in the past decade or so, as more and more twin samples have reached the age of starting a family and having their own children, and can provide more comprehensive information with a longer time span. A retrospective review of the CoT method shows that the current CoT studies cover a wide range of phenotypes, from mental illness, substance abuse, parenting, and marital difficulties, to distinguish the genetic and environmental effects between parents and children.

The CoT method can control family confounding factors and thus evaluate potential causal relationships in parent-child relationships [12]. One of the most important findings in CoT studies in recent years is the existence of evocative gene-environment correlations (evocative rGE) in parent/child behavior relationships, that is, the way parents interact with their children is also regulated by the children's innate behavioral tendencies caused by genetics. For example, adoption studies and twin studies have found that genetically influenced child behavior can influence parenting styles, just as parents can influence their children's behavior. These findings may help us better understand the dynamic changes in parent-child relationships, or different parent-child relationships in complex family environments.

Many of the large twin samples of the past have gradually had children of their own, which will provide researchers with more possibilities to explore new and interesting topics in the phenotypic associations between parents and children in the future [12].