SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Disclosure
  8. REFERENCES

Objective: To search for an association between the Glu27Gln (rs1042714; B27) and the Arg16Gly (rs1042713; B16) polymorphisms of the β2-adrenergic receptor (ADRB2) gene and obesity.

Methods: Meta-analysis of published studies, included if subjects were genotyped at either codon 27 (“B27”) or codon 16 (“B16”) of the ADRB2 gene and both obese and nonobese subjects were selected, based on a reported cutoff BMI limit. Initial selection included 14,444 subjects genotyped at B27 (rs1042714) and 6,825 genotyped at B16 (rs1042713). After testing each control group for Hardy-Weinberg equilibrium, the final selection included 10,404 subjects and 4,328 subjects, respectively. Studies were published before 18 August 2006.

Results: The frequency of Glu27 allele carriers, either homozygous or heterozygous, ranged from 6.71% in Aymara American Indians to 78.29% in a Dutch population. The frequency of Arg16 allele carriers varied from 51.4 to 64.6% in Europeans and from 71.1 to 85.6% in East Asians. The summary odds ratio (OR) from overall analyses showed no association between either rs1042714 or rs1042713 and obesity. In race groups with low Glu27 allele frequency (Asians, Pacific Islanders, and American Indians), ORs revealed a significant obesity risk associated with rs1042714. These results were not found in East Asians for rs1042713.

Discussion: The presence of the Glu27 allele in the ADRB2 gene appears to be a significant risk factor for obesity in Asians, Pacific Islanders, and American Indians, but not in Europeans. Obesity does not appear to be associated with the Arg16 allele.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Disclosure
  8. REFERENCES

Despite intense effort, the genetic pathways underlying obesity remain elusive. A key factor in obesity is the ratio of food intake to energy expenditure, and it is likely that genetic constitution is involved in setting its value. By regulating energy expenditure, catecholamines play a central role in the metabolism of lipids through their action on four adrenergic receptors subtypes (α2, β1, β2, and β3). β-Adrenergic receptors stimulate, while α2-adrenergic receptors inhibit lipolysis in fat (1). Genes encoding these receptors may constitute interesting candidates to explain part of the genetic predisposition to obesity in humans.

The β2-adrenergic receptor (ADRB2) is involved in lipid mobilization, as a major lipolytic receptor in human fat cells, and genetic variation in this receptor gene could theoretically reduce lipolysis and predispose to obesity. Since 1993, when nine different point mutations in the gene encoding for the ADRB2 were described (2) (but only four were found to cause changes in the encoded amino acids at residues 16, 27, 34, and 164), this receptor came under scrutiny, with the hypothesis that these single-nucleotide polymorphisms are associated with obesity and metabolic disorders. The most frequent single-nucleotide polymorphisms occur at codon 16 (rs1042713) and codon 27 (rs1042714) and their possible association with obesity is the focus of this meta-analysis.

Every point mutation at a nucleotide position in the gene generates a single-nucleotide polymorphism. It changes the structure in the corresponding codon (the coding structure is composed of three consecutive nucleotides that can code for only one amino acid) and generates two variants of the same gene called alleles that code for the same amino acids, except one, encoded by the codon where the mutation occurs. Both rs1042713 and rs1042714 mutant alleles are believed to alter the ADRB2 function by altering the amino acid sequence in the extracellular N-terminus of the ADRB2. In the rs1042713 polymorphism, a change occurs at the 16th amino acid position of ADRB2 protein, where the amino acid glycine (Gly16) is replaced with the amino acid arginine (Arg16). In the rs1042714 polymorphism, the change occurs at the 27th amino acid position, where glutamine (Gln27) is replaced with glutamic acid (Glu27). In our article, the alleles are designated using the name of the amino acid that constitutes their end product. The Arg16 allele has been associated with propensity to gain weight in early adulthood, while the Gly16 allele has been associated with lower receptor density and reduced efficiency (3). The Glu27 allele may limit ADRB2 down regulation and thus affect BMI.

The results concerning the contribution of both rs1042713 and rs1042714 toward obesity are contradictory. The first polymorphism was not a major contributing factor for obesity in Japanese men (4,5), but researchers found a significant association in Japanese women (4). In Swedish women, Large et al. (6) found obesity to be associated with rs1042714, but not with rs1042713.

For rs1042714, Lin et al. (7) found that the polymorphism was more frequent in overweight people, while Ehrenborg (8) found it to be associated only with a slightly to moderately elevated BMI. Echwald (9) and Oberkofler (10) found no association.

In Japanese men, Mori (11) found that rs1042714 was associated with obesity due to subcutaneous fat accumulation, while Hayakawa (5) found that it was not a major contributing factor to obesity. Hellstrom (12) found discordant results in women vs. men: there was a positive association in women, but not in men.

Other researchers found only a limited association between rs1042714 and obesity: according to Lange (13), there was a positive association only for the distribution of visceral adipose tissue; according to Gonzalez Sanchez (14), it favored only the accumulation of the visceral fat; according to Corbalan (15), it was associated with abdominal obesity among male subjects.

Some reports suggested that the association was conditional on the level of physical activity. Meirhaeghe (16), reported that the association was manifested at a low level of physical activity; conversely, in another study by Corbalan (17), females who carried the Glu27 allele were more resistant to losing weight when they participated in higher physical activity level.

Because of these conflicting results, many of which came from small studies, we undertook a meta-analysis of the published data on this topic.

Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Disclosure
  8. REFERENCES

The rs1042714 (Glu27Gln) polymorphism

We performed a thorough search of the medical literature in the English, Spanish, Portuguese, French, Italian, German, and Romanian languages. For medical articles in English, we searched PubMed, Embase, CINAHL (Cumulative Index to Nursing and Allied Health literature), the Cochrane database, Current Concepts, Austrom (Meditext), and Medind using the terms Glu27 and obesity. Other MeSH (Medical Subject Heading) terms such as BMI, morbid obesity, overweight, body weight, weight gain, weight loss, body weight changes, body weights, and measures were also used, in combination with Glu27. We used the search engines Google and Scirus to find additional articles, conference abstracts, and government reports suitable for inclusion in our meta-analysis. For Spanish literature, we used Google and the search terms “Glu27” and “obesidad.” For Portuguese literature, we used “Glu27” and “obesidade.” For French literature, we used “Glu27” and “obesité.” For Italian literature, we used “Glu27” and “obesità.” For German literature, we used “Glu27” and “obesitas.” The German terms captured also the Dutch literature. For Romanian medical literature, we used “Glu27” and “obezitate.” We did not search the medical literature in other languages.

As of 18 August 2006, when we finished our search, we identified 39 articles that reported polymorphisms at codon 27 of the ADRB2 gene (rs1042714) and included both obese and nonobese subjects. The English literature search provided 38 of these. The foreign language search provided only a doctoral thesis in Spanish, the results of which also appeared later in English literature. All these studies were selected for meta-analysis, including four with related family pedigrees as well as those using slightly different BMI cut points. Six additional studies performed on three databases (two studies/database in English) and one in Portuguese, that genotyped only obese persons, were excluded.

The broad reach of our terms led us to capture databases used to study other topics, such as the contribution of rs1042714 to various diseases, i.e., diabetes, hypertension, and metabolic disorders of the lipid metabolism, where the authors had calculated the BMI of the patients in addition to genotyping. In some of these databases, the BMI was not reported as a cutoff limit, as needed for our analysis, but rather was reported as a range of values with a standard deviation. In such instances, we corresponded with the authors and asked them to send us their data showing counts of subjects above and below a BMI of 30kg/m2. At this stage, we were able to include 28 of the 39 articles, covering 71.6% of the total number of patients reported. Eleven authors did not respond to our inquiries.

We considered the subjects who were obese (BMI above the cutoff limit) as being the cases, the nonobese subjects as being the controls. The subjects with the mutant allele at codon 27, coding for glutamic acid in the 27th amino acid position of the ADRB2 protein, were considered “exposed” and those with the wild type (Gln27Gln) were considered “nonexposed.” We pooled together the populations of both genders from all studies and compared in three ways those who have the glutamic acid allele at codon 27 (the “exposed”) to the wild type Gln27Gln (the “nonexposed”):

•    the first statistical analysis compared the homozygous Glu27Glu to the homozygous Gln27Gln;

•    the second statistical analysis compared the heterozygous Glu27Gln to the homozygous Gln27Gln; and

•    the third statistical analysis compared the combination of homozygous Glu27Glu and heterozygous Glu27Gln to the homozygous Gln27Gln.

Most of the authors used a cutoff BMI = 30 in their reports (Table 1). Five used BMI = 27 as a cutoff limit, two used BMI = 25, one used BMI = 26, and one a cutoff BMI = 26.4. Two authors characterized the cutoff BMI they used as “extreme” (18,19). We contacted a corresponding author and he stated that the obese subjects had a BMI above the 97th percentile, while the nonobese subjects had a BMI < 20. In the other study, “extreme” is clearly classified as morbid obesity, with a BMI > 40.

Table 1.  Studies included in the initial selection for meta-analysis at codon 27
inline image

As shown in Table 1, the subjects were from the United States, Europe (Sweden, France, Italy, Spain, Austria, the Netherlands, Denmark, and Poland), Asia (Japan, South Korea, and Taiwan) and two isolated populations: Tongans and Aymara American Indians from Chile.

We calculated the frequency of the Glu27 allele at the global level. Then, due to the wide variation in the frequency of Glu27 in these populations, we separated them by race and we performed further subgroup analyses. The subgroups were:

•    whites (all Europeans (6,8,9,10,12,14,16,18,19,20,21,22,23,24,25) and the Anglo-Celtic Australian whites (7)).

•    Asians, Pacific Islanders, and American Indians (Japanese (4,5,11,26,27), Taiwanese Chinese (28), Koreans (29), Tongans (30), and Aymara American Indians (31)).

•    Americans (mixed races (13,32,33) from United States and Canada).

We separated the data by gender (male, female, and both). Some authors provided data only in subjects of one gender, while others did not tabulate subjects based on gender and BMI combined or failed to report gender at all. The latter studies are included only in the tabulations of both sexes combined.

Then we calculated the Hardy-Weinberg equilibrium for every study both in the main group and in the gender-based subgroups. The procedure, including the formulas, is explained elsewhere (34). In order to perform the calculations, we used the Hardy-Weinberg equilibrium calculator software posted at http:www.changbioscience.comgeneticshardy.html. This calculator allowed us to introduce the frequencies of both alleles and their total number at the top and it gave the expected frequencies of genotypes at the bottom. Furthermore, we compared the real and the expected genotype frequencies using the χ2-test. We performed the χ2-test for the whole study and, separately, for both the experimental group and the control group. Also, we performed the test for both genders and then for male and female subgroups separately, if gender-based data were reported. If the χ2-test showed a significant departure from the Hardy-Weinberg equilibrium in the control group, the study was excluded from further analyses. If the departure was significant for the whole study population and/or for the experimental group, but not for the control group, the study was included. Following these calculations, we excluded from the main analyses four studies that have a significant departure from the Hardy-Weinberg equilibrium in the control group. The first study did report gender-based data, but, based on the χ2-test, it was excluded from both the main analyses and the gender-based subgroup analyses in males as well as in females. The second study did not report gender-based data, the third study reported only data in women and a fourth one reported data only in men. A fifth study was excluded because it provided data only for the heterozygotes and for the Gln27 homozygotes, while data for the Glu27 homozygotes were not reported and the Hardy-Weinberg equilibrium could not be calculated. In addition, a study reporting data for both genders was excluded only from the male subgroup, as long as the control group for the male population slightly departed from the Hardy-Weinberg equilibrium. Eventually, 23 studies were included in the main analyses, covering 51.55% from the subjects reported in the medical literature.

For our meta-analysis, we used the “Comprehensive Meta-analysis” software offered by Biostat (Englewood, NJ), at http:www.meta-analysis.com (35). The program provides summary results for both random effects and fixed effects models.

Using these genotype comparisons, we performed subgroup analyses based on race and, separately, on gender, where we included all suitable studies, irrespective of BMI. This way, we performed nine race-based subgroup analyses (three race categories times three categories for the polymorphism) and nine gender-based subgroup analyses (three gender categories—males, females, and both—times three categories for the polymorphism).

We then removed the studies with a BMI different from 30 and, based on race and gender, respectively, we performed two sets of nine subgroup statistical analyses each, that included only the studies that used a cutoff value of BMI = 30.

We performed sensitivity analyses, by first removing the studies that had at least one cell with zero subjects included, and then further removing studies that had at least one cell with less than five subjects included. Separately, by using the CMA software, we performed sensitivity analyses with one study removed.

Finally, we calculated the posttest power using the software posted at http:statpages.orgpostpowr.html.

After the sensitivity analyses, we performed the same operations with all 28 studies included, to detect if there is a loss of power or statistical significance after removing the studies that significantly departed from the Hardy-Weinberg equilibrium or did not provide sufficient data to calculate it. Then, we tested for heterogeneity using the CMA software (35).

In the last step, we ascertained publication bias, using a combination of graphical and numerical methods. As graphical methods, we chose to ascertain the publication bias using a combination of two funnel plots (the funnel plot of precision by log odds ratio (OR) and funnel plot of standard error by log OR, in both fixed effects and random effects model). As numerical methods, we chose to use the Egger's test.

The rs1042713 (Arg16Gly) polymorphism

For this polymorphism we used the same search strategy, by replacing Glu27 with Arg16. As of 18 August 2006, when we finished our search, we identified 27 articles that reported patients genotyped at codon 16 of the ADRB2 gene and included both obese and nonobese subjects, all of them in the English literature. All these studies were selected for meta-analysis, including those with related family pedigrees as well as those using slightly different BMI cut points. Four additional studies, which genotyped only obese persons, were excluded.

As shown in Table 2, the patients were from the United States, Europe (Sweden, Germany, Austria, and Italy), and East Asia (Japan, South Korea, and Taiwan).

Table 2.  Studies included in the initial selection for meta-analysis at codon 16
inline image

There were four studies reporting data only in men and two studies reporting data only in women. Seven studies reported data in both genders, of which only two reported data for men and women separately and could be included in the gender specific analyses.

Most of the authors used a cutoff BMI = 30 in their reports (see Table 2). One author characterized the BMI cutoff as “extreme” (the obese subjects had a BMI above the 97th percentile), while three others used as a cutoff limit BMI = 27. One other author used BMI = 25 and another author a cutoff BMI = 26.

When the BMI was not reported as a cutoff limit, as needed for our analysis, we corresponded with the authors and asked them to send us their data showing counts of subjects above and below a BMI of 30 kg/m2. Eventually, we were able to include 13 of the 27 articles, covering 36.6% of the total number of patients reported. Fourteen authors did not respond to our inquiries (some of them were asked to send data for both polymorphisms, while others only for rs1042713).

We considered the subjects who were obese (BMI above the cutoff limit) as being the cases and the nonobese subjects as being the controls. The subjects with the mutant allele at codon 16, coding for arginine in the 16th amino acid position of the ADRB2 protein, were considered “exposed” and those with the wild type (Gly16Gly) were considered “nonexposed.” The statistical analysis was conducted in the same way for rs1042713 and rs1042714. When we calculated the Hardy-Weinberg equilibrium for every study, both in the main group and in the gender-based subgroups, we removed two studies (both performed in Europeans, one in men and one performed only in women) that did not have the control group in Hardy-Weinberg equilibrium. Thus, from 13 studies initially selected, only 11 studies underwent further analysis. These two studies were also removed from gender-based and race-based analyses. The sensitivity analyses were also performed and posttest power was calculated the same way as for the B27 meta-analysis. Heterogeneity was studied and the publication bias was ascertained by using the CMA software.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Disclosure
  8. REFERENCES

We present the overall results, followed by race-based and gender-based subgroup analyses separately for rs1042713 and rs1042714 polymorphisms.

The rs1042714 (Glu27Gln) polymorphism

Overall. Among the 23 populations studied, rs1042714 was a risk factor for obesity in 15 of them (4,5,7–9,11,14,19,20,22,24–26,30,31), but reached significance only in two (4,11). The highest value OR was 3.4 in the Mori study (11).

In eight studies (10,12,13,16,18,21,29,33), the polymorphism had a protective effect, but this effect was significant only in the group of patients of both genders studied by Meirhaeghe (16), with an OR = 0.566 and a 95% confidence interval (CI) of 0.381 to 0.839. The overall OR varied from 1.076 when comparing the homozygous Glu27Glu to the homozygous Gln27Gln subjects in fixed effects (data not shown) to 1.130, when comparing the homozygous Glu27Glu and heterozygous Glu27Gln combined to the homozygous Gln27Gln in random effects (Figure 1), but the point estimate did not reach statistical significance. In these analyses, random effects models nearly always gave a higher OR and wider confidence limits compared to fixed effects models.

image

Figure 1. Odds ratios for obesity among subjects of both genders homozygous for Glu27Glu or heterozygous for Glu27Gln vs. subjects homozygous for Gln27Gln.

Download figure to PowerPoint

The ORs for the main meta-analyses failed to reach statistical significance. The P values varied between 0.067 for the combined homozygous Glu27Glu and heterozygous Glu27Gln in fixed effects (Figure 1) and 0.406 for the homozygous Glu27Glu in random effects (data not shown). As a result, the power of these tests was modest and it varied from 45% in the combined homozygous Glu27Glu and heterozygous Glu27Gln group in fixed effects (Table 3) to 13% in the homozygous Glu27Glu group in random effects (data not shown). The heterogeneity was significant, at least in combined homo- and heterozygote group (Q value = 39.6, 22 degrees of freedom, P value = 0.0121, I2 = 44.4, τ = 0.218).

Table 3.  Comparison of power and significance of the main group and subgroups of studies, as selected before and after the Hardy-Weinberg equilibrium testing, in the combined group of subjects, homozygous for the variant allele and heterozygous
inline image

Race-based subgroup analyses. The frequency of the Glu27 allele was 30.3% at the global level. There was a striking difference in the Glu27 allele frequency in study populations according to their race (Table 1): in whites, the frequency ranged from 44.74% in Polish (21) to 78.29% in Dutch (24), with an average of 40.7%; it ranged from 39.16% (Lange (13)) to 63.89% (Phares (33)) in the North American mixed races, and it varied from 6.71% in the Aymara American Indians (31) to 20.51% in Koreans (29), with an average of 7% in the group of Asians, Pacific Islanders, and American Indians.

The rs1042714 polymorphism was very rare in isolated populations: only three Aymaras (31) from 149 and only one from 1,022 Tongan subjects (30) were homozygous for this polymorphism.

The summary OR for Glu27Glu and Glu27Gln combined varied from 0.920 in North American mixed populations (both in fixed and random effects-two studies only—Table 3) to 1.566 in Asians, Pacific Islanders, and American Indians (Figure 2). The OR reached significance in both fixed and random effects (OR = 1.505, 95% CI = 1.169, 1.939 and OR = 1.566, 95% CI = 1.071, 2.288, respectively) only in the latter group (Table 3), where the ORs were driven up by the higher and significant values recorded in the four Japanese studies (OR = 2.082, 95% CI = 1.451, 2.988—both in fixed effects and in random effects models—data not shown).

image

Figure 2. Odds ratios for obesity among Asians, Pacific Islanders, and American Indians of both genders homozygous for Glu27Glu or heterozygous for Glu27Gln vs. subjects homozygous for Gln27Gln.

Download figure to PowerPoint

The power was variable among race-based subgroups. It was just 10% in the group of combined homozygous Glu27Glu and heterozygous Glu27Gln Americans (mixed races) (Table 3) who were homogeneous (the ORs and P values were equal and the power was 10% in both fixed and random effects, τ = 0). In Europeans and white Australians, the power varied between 4% in Glu27Glu homozygotes and 27% in Glu27Gln heterozygotes, the test did not reach statistical significance and heterogeneity was not significant (Q value = 20.4, 22 degrees of freedom, P value = 0.0853, I2 = 36.4, τ = 0.17). In Asians and Pacific Islanders, the power was higher and it varied from 43% in Glu27Glu homozygotes (data not shown) in random effects to 89% in combined homo- and heterozygotes in fixed effects (Table 3). The heterogeneity was not significant (Q value = 10.9, 6 degrees of freedom, P value = 0.091, I2 = 45, τ = 0.32).

Gender-based subgroup analyses. The results of the gender subgroups statistical analyses showed that ORs were higher in men than in women. For this reason, we chose to present the results for both genders combined (Figure 1) and for men only (Figure 3).

image

Figure 3. Odds ratios for obesity in males homozygous for Glu27Glu or heterozygous for Glu27Gln vs. males homozygous for Gln27Gln.

Download figure to PowerPoint

The ORs varied among studies and groups. From 18 studies performed in men, 13 studies showed ORs above the null (4,5,8,9,11,14,19,20,24,26,30,31,33). However, the ORs were significant only in two of them (11,24). The highest OR was 3.4 with a wide 95% CI (1.5–7.5) and it was described by Mori (11) in a group of Japanese patients (Figure 3). This group carried the polymorphism in only 10.4% of its members (Table 1).

Five studies showed a protective effect of rs1042714 for obesity (12,13,16,21,25). This effect, with an OR = 0.417 was significant (95% CI = 0.23, 074) only in a group of men from France described by Meirhaeghe (16) (Figure 3).

The fixed effects OR varied between 1.071 in women who were heterozygous for rs1042714 and 1.185 in homozygous men (data not shown). The random effects ORs were a little higher and varied between 1.072 in women who were heterozygous for rs1042714 and 1.305 in men homozygous for rs1042714 (data not shown). The ORs for overall analysis, including both genders, had intermediate values (see above).

The ORs for the gender-based meta-analyses failed to reach statistical significance. In males, the P values varied between 0.174 for the combined homozygous Glu27Glu and heterozygous Glu27Gln in random effects (Table 3) and 0.267 for the homozygous Glu27Glu in fixed effects (data not shown). As a result, the power of these tests was modest and it varied from 27% in the combined homozygous Glu27Glu and heterozygous Glu27Gln group in random effects (Table 3) to 20% in the homozygous Glu27Glu group in fixed effects (data not shown). In females, the P values varied between 0.267 for the combined homozygous Glu27Glu and heterozygous Glu27Gln in fixed effects (Table 3) and 0.536 for the homozygous Glu27Glu in random effects (data not shown). The power of these tests was even lower in women, from 20% in the combined homozygous Glu27Glu and heterozygous Glu27Gln group in fixed effects (Table 3) to 9% in the homozygous Glu27Glu group in random effects (data not shown). Heterogeneity in combined homo- and heterozygotes was significant in males (Q value = 39.2, 17 degrees of freedom, P value = 0.00167, I2 = 56.6, τ = 0.43), but not significant in females (Q value = 15.7, 13 degrees of freedom, P value = 0.26, I2 = 17.4, τ = 0.13).

The results of the other seven gender-based statistical analyses and of the nine statistical analyses that included only subjects with a cutoff BMI = 30 are quite similar and are not presented in the tables.

The results of the sensitivity analyses (not presented here) were quite similar to those of the main analyses, due to the small contribution of studies with empty or sparse cells, which were excluded.

Comparing the groups of studies in combined Glu27Glu homozygotes and Glu27Gln heterozygotes, as selected before and after applying to the Hardy-Weinberg equilibrium testing in the control group (Table 3), it resulted that it was an increase in the posttest power after the Hardy-Weinberg equilibrium testing-based selection in both the main group (from 35 to 45% in fixed effects), and in the gender-based subgroups: it increased in fixed effects from 17 to 24% in males and from 19 to 20% in females. In random effects, the change was smaller in the main group and in males and it did not change in females. In race-based analyses, it was an increase in the group of Asians and Pacific Islanders and a decrease in both the group of Europeans and white Australians and in the group of Americans (mixed races). The differences were greater in random effects model (Table 3). Trends were quite the same, with only minor variations, when the Glu27Glu homozygotes and the Glu27Gln heterozygotes were considered separately (data not shown). The variations in P values were inversely related to the variations in the posttest power in every selection of studies. (Table 3).

As expected, the heterogeneity was a little higher in the initial selection of studies. The Q test showed a value of 49 for 27 degrees of freedom and a P value = 0.00583. Other heterogeneity tests showed I2= 44.96 in initial selection vs. I2= 39.57 in final selection, but τ = 0.217 in both selections.

In the final selection, the visual inspection of the funnel plot of precision by log OR, both in fixed effects (Figure 4) and in random effects shows a symmetrical plot, with a regression slope that crosses the x-axis at a point located very close to zero. The funnel plot of standard error by log OR, both in fixed effects and in random effects is symmetrical (figure not shown). The Egger's regression intercept is 0.815 (95% CI = −0.52, 2.15) with a two-tailed P value of 21.7%.

image

Figure 4. Final B27 selection: funnel plot of precision on log odds ratio in fixed effects.

Download figure to PowerPoint

The rs1042713 (Arg16Gly) polymorphism

Overall. Among the 11 populations studied, rs1042713 was reported as a risk factor for obesity in eight of them (4,5,10,18,26,27,29,32), but in none of the studies did the association achieve statistical significance. The highest OR value was 2.6 in the Hayakawa study (5).

In three studies (8,25,28), the polymorphism had a protective effect, but this effect was significant only in the Bengtsson study (25) (Figure 5).

image

Figure 5. Odds ratios for obesity among subjects of both genders homozygous for Arg16Arg or heterozygous for Gly16Arg vs. subjects homozygous for the Gly16Gly.

Download figure to PowerPoint

The overall OR varied from 1.021 when comparing the homozygous Arg16Arg or heterozygous Arg16Gly to the homozygous Gly16Gly subjects in fixed effects (Figure 5) to 1.047, when comparing the homozygous Arg16Arg to the homozygous Gly16Gly in random effects (data not shown), but the point estimate did not reach statistical significance. In these analyses, random effects models nearly always gave a higher OR and wider confidence limits compared to fixed effects models. Heterogeneity was not significant (data not shown).

Race based. There was some variation in the frequency of the arginine allele among populations. The frequency of the variant allele ranged from 51.4 to 64.6% in Europeans and from 71.1 to 85.6% in East Asians (Table 2).

As long as it was only one study in American mixed races included in our selection, we performed a statistical analysis only in Europeans and in East Asians. Studies in white Australians, Pacific Islanders, and American Indians provided data only for rs1042714, but not for rs1042713.

The summary OR for Arg16Arg and Arg16Gly combined varied from 0.848 in Europeans in fixed effects to 1.142 in East Asians, in both fixed and random effects (Table 3) and it did not reach significance.

The power was variable among race-based subgroups. It varied between 6% in the initial selection in random effects in the group of homozygote Arg16Arg and heterozygote Arg16Gly Europeans and 30% in fixed effects in the final selection of the same group, with intermediate and constant values for East Asians (Table 3). Heterogeneity was not significant in Europeans (data not shown), while the East Asian group was homogeneous (τ = 0).

Gender-based subgroup analyses. The results of the gender subgroups statistical analyses showed that summary ORs had values closer to 1 and did not reach significance. They were a little higher in women than in men (1.03 vs. 0.96 for homozygous and heterozygous groups combined in random effects—Table 3). The highest OR value was 1.194 in heterozygous women (random effects) and the lowest OR was 0.869 in homozygous men (in both fixed and random effects—data not shown). For this reason, we chose to present the results for both genders (Figure 5) and for women only (Figure 6) who were either homozygous or heterozygous for the arginine allele.

image

Figure 6. Odds ratios for obesity among women homozygous for Arg16Arg or heterozygous for Gly16Arg vs. women homozygous for Gly16Gly.

Download figure to PowerPoint

Among four studies providing data on women, two studies showed ORs above the null (4,10). The highest OR value was 3.31 in a study of Japanese women (4) and it achieved statistical significance, but with a wide CI (95% CI = 1.29, 8.52). Alternatively, a study in Swedish women (25) showed a significant protective effect of the B16 ADRB2 variant for obesity (OR = 0.509, 95% CI = 0.335, 0.771—see Figure 6).

The studies were relatively homogeneous. Only studies that included women heterozygous for the arginine allele were significantly heterogeneous (data not shown).

The results of the other gender-based statistical analyses and of the nine statistical analyses that included only patients with a cutoff BMI = 30 are quite similar, but the results are nonsignificant and they are not presented in the tables.

As a sensitivity analysis, we also removed studies reporting protective effect (data not shown) and the overall results were quite similar to the main analyses.

In the final selection, the visual inspection of the funnel plot of precision by log OR, both in fixed effects and in random effects (Figure 7) showed a symmetrical plot, with a regression slope that crosses the x-axis at a point located very close to zero. The funnel plot of standard error by log OR, both in fixed effects and in random effects was symmetrical (figure not shown). The Egger's regression intercept was 0.578 (95% CI = −1.33, 2.48) with a two-tailed P value of 50.4%.

image

Figure 7. Final B16 selection: funnel plot of precision on log odds ratio in random effects.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Disclosure
  8. REFERENCES

The function of the ADRB2, a major lipolytic receptor in human fat cells, may be impaired due to genetic polymorphism, leading to obesity. These meta-analyses were undertaken because of divergent results regarding the effect of the most frequent polymorphisms in the ADRB2 gene (rs1042713 and rs1042714) on obesity. We were successful in getting data on 36.6 and 71.6%, respectively, of subjects genotyped at codons 16 and 27 of the ADRB2 gene that were identified in the world literature.

We considered the same cutoff BMI value of 30 in all ethnic groups, all over the world. The subjects above this value were considered obese and were included in the experimental group, while the subjects with a BMI under the cutoff value were considered normal and used as controls. However, research shows that, at higher BMI levels, Polynesians are significantly leaner than Europeans (36), implying the need for separate BMI definitions of overweight and obesity for Polynesians in order to describe the same health risks. This variation can be explained by body type. At the other end, a recent WHO expert consultation (37) recognized that Asian populations may require a lower BMI to indicate that an individual is at same risk as an European. This is reflected in our selection, where four studies out of five at codon 27 and four studies out of six at codon 16 that were performed in Asians and entered in the final selection used a BMI < 30. Eventually, our sensitivity analyses showed that this variation in BMI did not have a sizable influence on the results.

Using this definition, we chose a population-based case-control study design. In case-control studies, the assessment of the association between a binary disease outcome and a single-nucleotide polymorphism is based on comparing the observed genotype distribution for the cases against that for the controls.

We pooled together the populations of both genders from all studies and compared the “exposed” groups having either the Arg 16 or the Glu27 allele, respectively, to the “nonexposed” group, namely, the wild type (herein either Gly16Gly or Gln27Gln, respectively) in three separate analyses for each polymorphism. We did not choose to assume a specific genetic model for our meta-analysis, because there is no clear biological evidence available to substantiate the choice of a certain genetic model. In a polygenic disease such as obesity, with a strong environmental influence, genotype is only one factor on the causal pathway to the disease and gene-gene and gene-environment interactions can influence the final association genotype/disease and the genetic mode of action.

By choosing to remove the studies that departed from the Hardy-Weinberg equilibrium assumptions in the control group (as much as 28% from our initial B27 database and 16.3% from our initial B16 database), we were concerned that the power and the significance would be affected, but it was not the case. Comparing the groups of studies, as selected before and after applying to the Hardy-Weinberg equilibrium testing in the control group (see Table 3), it resulted that it was an increase in power and significance after the Hardy-Weinberg equilibrium testing based selection with one notable exception, in Europeans and white Australians genotyped at codon 27, where the power and significance of this group of studies actually decreased; however, this variation (higher in fixed effects than in random effects) was small. An interesting finding was that the estimate of between-study variance (“the amount of heterogeneity”), also known as τ2, was equal to 0.0477 in both the initial and the final B27 selections, while the proportion of variability that was due to true heterogeneity rather than sampling error (represented by a new index known as I2) was lower in the final B27 selection.

Prior work has shown disparate results among a number of small studies. The problem of concern is that, usually, small studies may have more significant results than larger trials. This is also the case with our final B27 selection, which shows that two (11,31) out of the three smallest studies have more significant results (Figure 1).

The meta-analytic process had the advantage of combining the small studies into a single-larger analysis, that included 14,444 subjects of both genders, from every continent, except Africa (we had only African Americans included) and of many races (whites, blacks, Hispanics, Asians, Aymara American Indians, and Pacific islanders from Tonga) in the B27 meta-analysis and 6,931 subjects from North America, Europe, and East Asia in the B16 meta-analysis; however, by pooling all these populations together, we did not find a strong and significant association between either rs1042713 or rs1042714 and obesity, but we think that many genetic influences on obesity may have context dependent effects, influencing body weight in specific populations or groups, despite the lack of evidence for an association in a meta-analysis that combines multiple ethnic groups. The ORs showed an increased risk for obesity in subjects having the polymorphism in certain populations, while in others it was not apparent. In our opinion, the strength of the association between either rs1042713 or rs1042714 and obesity may be variable in different populations, depending on the intervention of competing component causes on obesity, despite the lack of evidence for an association in a meta-analysis that combined multiple ethnic groups. As an example, we found the association between rs1042714 and obesity to be significant in the subgroup of Asians and Pacific Islanders, but not in Europeans.

The heterogeneity in the B27 genotyped populations may be due to a wide variation in the frequency of the Glu27 allele among populations subgroups, drawn from six continents, with the highest percentage among whites (40.7%) and the smallest percentage among populations living around the Pacific Ocean and on its islands (7% among Japanese, Taiwanese Chinese, Aymara American Indians, and Tongans). This was in line with another meta-analysis about the effect of the Glu27Gln polymorphism in asthma (38) that found a frequency of the Glu27 allele of 40.3% in Europeans, 8.2% in “oriental” populations and 5% in Polynesians.

Also, the environmental factors such as diet and lifestyle are very different in particular locations, with many genetic influences on obesity having context dependent effects that may interfere with the genetic effect and influence the body weight only in specific populations or groups. A third reason for heterogeneity could reside in the quality and the design of studies, with the studies performed in populations living around the Pacific Ocean and on its islands having more power than those in their European counterparts.

In order to ascertain the publication bias, we used a combination of graphical and numerical methods. As a graphical method, we used two types of funnel plots. At visual inspection, both the funnel plot of precision by log OR (Figure 4) and the funnel plot of standard error by log OR are symmetrical. The Egger's test (a numerical method) shows that there are less publication bias in the final B27 selection vs. the initial selection (intercept = 0.815 and P two-tailed value =21.7% vs. intercept = 1.098 and P two-tailed value = 7.25%) and that bias do not reach significance. However, these results concerning the publication bias should be interpreted with caution, because there is still a significant amount of heterogeneity and the posttest power in the main analyses (45%) still is low. Because of this heterogeneity, the use of random effects model is preferred.

While this meta-analysis was under review, Gjesing et al. (39) published another meta-analysis about a possible association between the ADRB2 polymorphism at amino acid positions 16 and 27 and obesity. They included in their meta-analysis 13 of our 39 B27 selected studies and their huge database of 7,808 European subjects. Eventually, they obtained similar results, but they did not provide details about methodology and nor did they detail their results. Using different statistical methods on a database with significant heterogeneity, these authors failed to identify any association between the presence of glutamic acid at the 27th amino acid position of the β2-receptor protein and obesity (Glu/Glu OR = 0.92 (0.81–1.04), Gln/Glu OR = 1.07 (0.98–1.18)). Their findings in the Danish population confirm our results concerning a lack of association between rs1042713 and rs1042714 and obesity in Europeans, obtained on a database of 7,761 B27 genotyped and 2,830 B16 genotyped subjects from eight European countries.

Our meta-analysis also had some limitations:

•    we performed a thorough literature search; however, some studies in foreign languages, in medical journals with little circulation, might have been missed;

•    as meta-analysts, we did not enroll patients, but we depended on the studies' authors to provide data;

•    we did not receive data from all the authors, but we believed that this was not a major source of bias, as long as the authors who did not respond were not different from those who responded. Of note is that we cannot make further comments, as long as we do not know if the populations from these studies departed from the Hardy-Weinberg equilibrium assumptions in their control groups.

•    In case-control studies, population admixture may be a problem of concern; however, only two studies in ethnically and racially mixed populations from the American continent met the selection criteria and were included in the final selection.

In conclusion, both rs1042713 and 1042714 polymorphisms in the ADRB2 gene do not appear to be associated with obesity defined by a BMI > 30. The results across studies clearly show heterogeneity of the effect according to race in the B27 selection, since rs1042714 is a significant risk factor for obesity only in Pacific populations, defined by a low frequency of the glutamic acid in the 27th amino acid position of the β2-receptor protein.

REFERENCES

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Disclosure
  8. REFERENCES
  • 1
    Lafontan M, Sengenes C, Galitzky J et al. Recent developments on lipolysis regulation in humans and discovery of a new lipolytic pathway. Int J Obes Relat Metab Disord 2000; 24 (Suppl 4):S47S52.
  • 2
    Reihsaus E, Innis M, MacIntyre N, Liggett SB. Mutations in the gene encoding for the β2-adrenergic receptor in normal and asthmatic subjects. Am J Respir Cell Mol Biol 1993; 8: 334339.
  • 3
    Ellsworth DL, Coady SA, Chen W et al. Influence of the β2-adrenergic receptor Arg16Gly polymorphism on longitudinal changes in obesity from childhood through young adulthood in a biracial cohort: the Bogalusa Heart Study. Int J Obes Relat Metab Disord 2002; 26: 928937.
  • 4
    Ishiyama-Shigemoto S, Yamada K, Yuan X et al. Association of polymorphisms in the β2-adrenergic receptor gene with obesity, hypertriglyceridaemia, and diabetes mellitus. Diabetologia 1999; 42: 98101.
  • 5
    Hayakawa T, Nagai Y, Kahara T et al. Gln27Glu and Arg16Gly polymorphisms of the β2-adrenergic receptor gene are not associated with obesity in Japanese men. Metabolism 2000; 49: 12151218.
  • 6
    Large V, Hellstrom L, Reynisdottir S et al. Human β−2 adrenoceptor gene polymorphisms are highly frequent in obesity and associate with altered adipocyte β−2 adrenoceptor function. J Clin Invest 1997; 100: 30053013.
  • 7
    Lin RCY, Ericsson JO, Benjafield AV, Morris BJ. Association of β2-adrenoceptor Gln27Glu variant with body weight but not hypertension. Am J Hypertens 2001; 14: 12011204.
  • 8
    Ehrenborg E, Skogsberg J, Ruotolo G et al. The Q/E27 polymorphism in the 2-adrenoceptor gene is associated with increased body weight and dyslipoproteinaemia involving triglyceride-rich lipoproteins. J Intern Med 2000; 247: 651656.
  • 9
    Echwald SM, Sorensen TI, Tybjaerg-Hansen A, Andersen T, Pedersen O. Gln27Glu variant of the human β2-adrenoreceptor gene is not associated with early-onset obesity in Danish men. Diabetes 1998; 47: 16571658.
  • 10
    Oberkofler H, Esterbauer H, Hell E, Krempler F, Patsch W. The Gln27Glu polymorphism in the β2-adrenergic receptor gene is not associated with morbid obesity in Austrian women. Int J Obes Relat Metab Disord 2000; 24: 388390.
  • 11
    Mori Y, Kim-Motoyama H, Ito Y et al. The Gln27Glu β2-adrenergic receptor variant is associated with obesity due to subcutaneous fat accumulation in Japanese men. Biochem Biophys Res Commun 1999; 258: 138140.
  • 12
    Hellstrom L, Large V, Reynisdottir S, Wahrenberg H, Arner P. The different effects of a Gln27Glu β2-adrenoceptor gene polymorphism on obesity in males and in females. J Intern Med 1999; 245: 253259.
  • 13
    Lange LA, Norris JM, Langefeld CD et al. Association of adipose tissue deposition and β−2 adrenergic receptor variants: the IRAS family study. Int J Obes 2005; 29: 449457.
  • 14
    Gonzalez Sanchez JL, Proenza AM, Martinez Larrad MT et al. The glutamine 27 glutamic acid polymorphism of the β2-adrenoceptor gene is associated with abdominal obesity and greater risk of impaired glucose tolerance in men but not in women: a population-based study in Spain. Clin Endocrinol (Oxf) 2003; 59: 476481.
  • 15
    Corbalan MS, Marti A, Forga L, Martinez-Gonzalez MA, Martinez JA. β2-adrenergic receptor mutation and abdominal obesity risk: effect modification by gender and HDL-cholesterol. Eur J Nutr 2002; 41: 114118.
  • 16
    Meirhaeghe A, Helbecque N, Cottel D, Amouyel P. Impact of polymorphisms of the human β2-adrenoceptor gene on obesity in a French population. Int J Obes Relat Metab Disord 2000; 24: 382387.
  • 17
    Corbalan MS, Marti A, Forga L, Martinez-Gonzalez MA, Martinez JA. The 27Glu polymorphism of the β2-adrenergic receptor gene interacts with physical activity influencing obesity risk among female subjects. Clin Genet 2002; 61: 305307.
  • 18
    Tafel J, Branscheid I, Skwarna B et al. Variants in the human β1-, β2-, and β3-adrenergic receptor genes are not associated with morbid obesity in children and adolescents. Diabetes Obes Metab 2004; 6: 452455.
  • 19
    Kortner B, Wolf A, Wendt D, Beisiegel U, Evans D. Lack of association between a human β−2 adrenoceptor gene polymorphism (gln27glu) and morbid obesity. Int J Obes Relat Metab Disord 1999; 23: 10991100.
  • 20
    Martinez JA, Corbalan MS, Sanchez-Villegas A et al. Obesity risk is associated with carbohydrate intake in women carrying the Gln27Glu β2-adrenoceptor polymorphism. J Nutr 2003; 133: 25492554.
  • 21
    Malczewska-Malec M, Wybranska I, Leszczynska-Gołabek I et al. An analysis of the link between polymorphisms of the β2 and β3 adrenergic receptor gene and metabolic parameters among Polish Caucasians with familial obesity. Med Sci Monit 2003; 9: CR277CR286.
  • 22
    Carlsson M, Orho-Melander M, Hedenbro J, Groop LC. Common variants in the β2-(Gln27Glu) and β3-(Trp64Arg)—adrenoceptor genes are associated with elevated serum NEFA concentrations and type II diabetes. Diabetologia 2001; 44: 629636.
  • 23
    Galletti F, Iacone R, Ragone E et al. Lack of association between polymorphism in the β2-adrenergic receptor gene, hypertension, and obesity in the Olivetti Heart Study. Am J Hypertens 2004; 17: 718720.
  • 24
    van Tilburg JH, Wijmenga C, van Bakel H et al. Relationship of β2-adrenergic receptor polymorphism with obesity in type 2 diabetes. Diabetes Care 2003; 26: 251252.
  • 25
    Bengtsson K, Orho-Melander M, Melander O et al. β2-adrenergic receptor gene variation and hypertension in subjects with type 2 diabetes. Hypertension 2001; 37: 13031308.
  • 26
    Iwamoto N, Ogawa Y, Kajihara S et al. Gln27Glu β2-adrenergic receptor variant is associated with hypertriglyceridemia and the development of fatty liver. Clin Chim Acta 2001; 314: 8591.
  • 27
    Kato N, Sugiyama T, Morita H et al. Association analysis of β2-adrenergic receptor polymorphisms with hypertension in Japanese. Hypertension 2001; 37: 286292.
  • 28
    Chang TJ, Tsai MH, Jiang YD et al. The Arg16Gly polymorphism of human β2-adrenoreceptor is associated with type 2 diabetes in Taiwanese people. Clin Endocrinol (Oxf) 2002; 57: 685690.
  • 29
    Kim SH, Kim DJ, Seo IA et al. Significance of β2-adrenergic receptor gene polymorphism in obesity and type 2 diabetes mellitus in Korean subjects. Metabolism 2002; 51: 833837.
  • 30
    Duarte NL, Colagiuri S, Palu T, Wang XL, Wilcken DE. Obesity, type II diabetes and the β2 adrenoceptor gene Gln27Glu polymorphism in the Tongan population. Clin Sci (Lond) 2003; 104: 211215.
  • 31
    Santos JL, Pérez-Bravo F, Martínez JA et al. No evidence for an association between genetic polymorphisms of 2- and 3-adrenergic receptor genes with body mass index in Aymara natives from Chile. Nutrition 2002; 18: 255258.
  • 32
    Garenc C, Perusse L, Chagnon YC et al. Effects of β2-adrenergic receptor gene variants on adiposity: the HERITAGE Family Study. Obes Res 2003; 11: 612618.
  • 33
    Phares DA, Halverstadt AA, Shuldiner AR et al. Association between body fat response to exercise training and multilocus ADR genotypes. Obes Res 2004; 12: 807815.
  • 34
    Emigh TH. A Comparison of Tests for Hardy-Weinberg equilibrium. Biometrics 1980; 36: 627642.
  • 35
    Borenstein M, Hedges L, Higgins J, Rothstein H. Comprehensive Meta-analysis Version 2, Biostat, Englewood NJ (2005).
  • 36
    Swinburn BA, Ley SJ, Carmichael HE et al. Body size and composition in Polynesians. Int J Obes 1999; 23: 11781183.
  • 37
    WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 2004; 363: 157163.
  • 38
    Thakkinstian A, McEvoy M, Minelli C et al. Systematic review and meta-analysis of the association between β2-adrenoceptor polymorphisms and asthma: a HuGE review. Am J Epidemiol 2005; 162: 201211.
  • 39
    Gjesing AP, Andersen G, Burgdorf KS et al. Studies of the associations between functional β2-adrenergic receptor variants and obesity, hypertension and type 2 diabetes in 7,808 white subjects. Diabetologia 2007; 50: 563568.