**3. Results**

Of the 249 pregnan<sup>t</sup> women with pregestational DM assessed for eligibility, 87 were included in the DASDIA clinical trial, and 70 were included in the present study because there were sufficient data for the analyses: 41 in the traditional diet group and 29 in the DASH diet group (Figure 1).

**Figure 1.** Flowchart of the study (Rio de Janeiro/Brazil, 2016–2020).

The median age was 32 years (IQR 25.7–36.0), and the gestational age at randomization was 15 weeks (IQR 11.1–20.1). DM type 1 was 51.4% (*n* = 36) of the cases. The distribution of the variables was homogeneous among the diet groups (Table 2).

**Table 2.** General characteristics of the participants at baseline (Rio de Janeiro/Brazil, 2016–2020).



Data presented as median (interquartile range) or as absolute and relative frequencies *n* (%). † Estimated for exchange rate of 1 real (BRL) = USD 5.5. \* Mann–Whitney U test or Kruskal–Wallis test to compare medians and chi-square test or Fisher exact test to compare frequencies.

The genotypic frequencies of rs9939609–FTO were TT 40%, AT 48.6%, and AA 11.4%, while for rs17817449–FTO, the values were TT 45.7%, GT 44.3%, and GG 10%, without differences among diet groups (*p* = 0.48 and *p* = 0.73, respectively; Table 3). The MAFs for rs9939609 (A) were 35.7%, while it was 32.1% for rs17817449 (G).

The genotypic frequencies of rs1042713–ADRB2 were GG 35.7%, AG 52.9%, and AA 11.4%, while for rs1042714–ADRB2, the values were CC 50%, CG 44.3%, and GG 5.7%, without differences among diet groups (*p* = 0.35 and *p* = 0.28, respectively; Table 3). The MAFs for rs1042713 (A) were 37.9% and 27.9% for rs1042714 (G). The genotypes of all evaluated polymorphisms were in HWE (*p* > 0.05).


**Table 3.** Genetic background of the participants concerning FTO and ADRB2 polymorphisms (Rio de Janeiro/Brazil, 2016–2020).

FTO, fat mass and obesity-associated gene; ADRB2, adrenoceptor beta 2 gene. Data presented as absolute and relative frequencies *n* (%). Genotypes were in Hardy–Weinberg equilibrium. \* Chi-square test or Fisher's exact test to compare frequencies.

The median time of intervention was 22.50 weeks (IQR 15.50–26.04). Most pregnan<sup>t</sup> women attended six scheduled appointments (*n* = 38, 54.3%) or at least five of them (*n* = 16, 22.9%). Almost 40% of the participants had the highest adherence scores in both groups (39.5% in the traditional diet group and 40.7% in the DASH diet group).

In the overall sample (*n* = 70), 28.6% of the women had a normal pre-pregnancy BMI, 35.7% were overweight, and 35.7% were obese. None of the pregnan<sup>t</sup> women were underweight pre-pregnancy. The median GWG was 13.7 kg (IQR 11.5–17.5), 11.8 kg (IQR 7.5–16.4), and 11.0 (IQR 5.9–14.1) for normal-weight, overweight, and obese women, respectively, without differences between diet groups (Supplementary Table S1). We found no statistically significant interaction between diet and genotype on GWG but a marginal effect for the AA genotype of rs9939609–FTO and GG genotype of rs17817449–FTO (*p* = 0.05 and *p* = 0.08, respectively) on higher GWG comparing to another genotypes, only in the traditional diet group (Supplementary Figure S1).

Thirty-seven pregnan<sup>t</sup> women (52.9%) presented with excessive GWG, and the median gestational age of exceeding GWG was 31.6 weeks (IQR 26.6–35.0). Compared to the traditional diet, the DASH diet did not modify the risk of progression to excessive GWG (aHR 1.32, CI 95% 0.62–2.79; *p* = 0.46) in our sample. Instead, the time of living with DM ≥ 8 years (aHR 1.99, CI 95% 1.01–3.93; *p* = 0.04), pre-pregnancy overweight (aHR 3.15, CI 95% 1.23–8.09; *p* = 0.02) or obesity (aHR 2.87, CI 95% 1.11–7.42; *p* = 0.03) status, previous hypothyroidism (aHR 4.37, CI 95% 1.62–11.77; *p* = 0.00), and yellow color of the skin (aHR 74.40; CI 95% 4.25–1302.72; *p* = 0.00, not shown in the table) were risk factors for earlier GWG. In contrast, age ≥ 32 years was a protective factor (aHR 0.41, CI 95% 0.21–0.80; *p* = 0.01) (Table 4).


**Table 4.** Cox proportional hazard models or time-to-event analyses (from conception to excessive gestational weight gain) of diet groups and general characteristics of the participants (Rio de Janeiro/Brazil, 2016–2020).


**Table 4.** *Cont.*

pY, person-years; CI, confidence interval; DASH, Dietary Approach to stop Hypertension; DM, diabetes mellitus; BMI, body mass index. \* Adjusted HR for skin color, previous chronic diseases, and housing conditions.

Adjusting for the main confounders, the A allele carriers (AT/AA) had a higher risk of earlier exceeding GWG (aHR 2.55; CI 95% 1.14–5.69; *p* = 0.02) than the rs9939609 TT genotype in the FTO gene, which was also found in the comparison of AT vs. TT genotypes (aHR 2.44; CI 95% 1.03–5.78; *p* = 0.04) (Table 5).

**Table 5.** Cox proportional hazard models or time-to-event analyses (from conception to excessive gestational weight gain) stratified by the FTO polymorphisms rs9939609 and rs17817449 (Rio de Janeiro/Brazil, 2016–2020).


pY, person-years; CI, confidence interval; FTO, fat mass and obesity-associated gene. \* Adjusted HR for skin color, previous chronic diseases, and housing conditions.

The A allele carriers for rs1042713 in the ADRB2 gene had a higher risk of earlier exceeding GWG than the GG genotype (aHR 2.37; CI 95% 1.01–5.52; *p* = 0.04), being almost four times higher for the AA carriers (aHR 3.91; CI 95% 1.12–13.70; *p* = 0.03). We found that the genotypes for rs17817449 (FTO) and rs1042714 (ADRB2) had no effect on the risk of excess GWG in our sample (Table 6).


**Table 6.** Cox proportional hazard models or time-to-event analyses (from conception to excessive gestational weight gain) stratified by the ADRB2 polymorphisms rs1042713 and rs1042714 (Rio de Janeiro/Brazil, 2016–2020).

pY, person-years; CI, confidence interval; ADRB2, adrenoceptor beta 2 gene. \* Adjusted HR for skin color, previous chronic diseases, and housing conditions.

Although rs17817449 alone was not associated with the outcome, in the haplotype analysis of rs9939609:rs17817449 (TT/AG/AT), we found a higher risk for earlier excessive GWG among AG carriers: the A allele for rs9939609 and the G allele for rs17817449 (aHR 1.79; CI 95% 1.04–3.06; *p* = 0.02). We found no association between haplotype analysis of the ADRB2 gene in our sample (Table 7).

**Table 7.** Cox proportional hazard models or time-to-event analyses (from conception to excessive gestational weight gain) of the haplotypes ADRB2 rs1042713:rs1042714 and FTO rs9939609:rs17817449 (Rio de Janeiro/Brazil, 2016–2020).


pY, person-years; CI, confidence interval; FTO, fat mass and obesity-associated gene; ADRB2, adrenoceptor beta 2 gene. \* Adjusted HR for skin color, previous chronic diseases, and housing conditions.

### **4. Discussion**

In a sample of 70 Brazilian pregnan<sup>t</sup> women with pregestational DM, we found that the A allele carriers for rs9939609 (FTO gene) and rs1042713 (ADRB2 gene) had more than twice the risk of earlier exceeding GWG compared to TT and GG genotypes, respectively. The haplotype rs9939609:rs17817449 (AG) was also a risk factor, increasing 1.8 times in terms of the progression to excessive GWG. Time of living with DM of ≥8 years, pre-pregnancy overweight or obesity, and previous hypothyroidism were risk factors for earlier excessive GWG. However, age ≥ 32 years old was a protective factor. We found no effect of the DASH diet on the risk for progression to excessive GWG, but our results of non-association need to be interpreted with caution because of our limited statistical power.

The allele frequencies in our sample were close to the frequencies in global databases (www.ncbi.nlm.nih.gov/snp (accessed on 18 January 2022)). Common polymorphisms (>5% allele frequency) are expected to be shared across different geographical regions and populations [36]. However, the Brazilian population is highly admixed and underrepresented in genomic studies as a potential source of new phenotype-associated genetic variants [37].

The A allele for the rs9939609 polymorphism in the FTO gene was previously associated with higher BMI before and after pregnancy [38,39] but not with excessive GWG [15] in Brazilian women. Studies from the USA [12,13] and Spain [14] found a higher risk of having higher GWG among A-allele carriers, contradicting the results from Turkey [40] and Mexico [41], both without association. The other evaluated polymorphisms (FTO rs17817449, ADRB2 rs1042713, and rs1042714) have not been previously investigated for GWG. The haplotype rs9939609:rs17817449 (AT) was found to increase the risk of obesity in the Brazilian adult population [11], but we identified the role of AG in the risk of progression to GWG.

The effect of the DASH diet on weight gain during pregnancy is controversial. Van Horn et al. (2018) noted that overweight and obese pregnan<sup>t</sup> women (*n* = 280, EUA) gained less weight and had less excessive GWG using the DASH diet than by using a control diet [42], whereas Fulay et al. (2018) reported that high adherence to the DASH diet was associated with more weight gain during pregnancy in obese women (*n* = 1760, EUA) [43].

Genetics may partially explain the interindividual variation in body weight in response to nutritional intervention [44]; thus, it was hypothesized that the FTO and ADRB2 polymorphisms could modify the effect of diet on GWG. We found a marginal association of the FTO polymorphisms on the GWG according to the type of diet: women with the AA genotype for rs9939609 and women with the GG genotype for rs17817449 had higher GWG in the traditional diet group. This result could represent some benefit of the DASH diet on limiting GWG for women with these genotypes, but it was not confirmed in the analysis adjusted for the main confounders, as we found no effect of diet on the risk for progression to excessive GWG in our sample.

Martins et al. (2018) reported that the A allele for rs9939609 was associated with an increase in the total energy intake and increase in the percentage of energy from ultraprocessed foods during pregnancy in a cohort of Brazilian women [45]. In our study, we calculated the individualized meal plan for all participants in both groups, considering the appropriate GWG, and we found similar high adherence to diet in the groups (40%). Indeed, the evaluation of dietary intake in details deserves further analyses to clarify if the genetic polymorphisms affect the GWG by dietary characteristics other than the dietary pattern (traditional or DASH), such as the level of food processing, for example.

Pregestational overweight and obesity were risk factors for progression to excessive GWG in our sample. This result agrees with the study by Brandão et al. (2021), which involved a large cohort of healthy Brazilian women and found that excessive GWG was observed in 30.1%, 30.7%, 56.4%, and 46.2% of underweight, normal weight, overweight, and obese women, respectively [2]. According to the IOM guidelines, GWG recommendations decrease when BMI increases. Thus, obese women should gain less weight than overweight

and normal weight women, but the guidelines do not provide specific recommendations for women with pregestational DM [6].

In this context, Siegel et al. (2015) found no difference between BMI classes to gain less, within, or above the IOM recommendations in a sample of women with pregestational DM but noticed a higher risk for macrosomia (aOR 4.02; CI 95% 1.16–13.9) and large-forgestational-age infants (aOR 3.08; CI 95% 1.13–8.39) in women who gained excessive weight compared to women who gained weight within the recommended amounts, even after adjusting for pregestational BMI [46]. The results from the study by Egan et al. (2014) were similar, reporting that excessive GWG in women with pregestational DM was a risk factor for macrosomia (aOR 3.58; CI 95% 1.77–7.24) and large-for-gestational-age infants (aOR 3.97; CI 95% 1.85–8.53), but they found more women with overweight or obesity presenting with excessive GWG than non-excessive GWG (44% vs. 27% for overweight and 37% vs. 25% for obesity; *p* < 0.01) [3].

We found that the number of years living with DM was a risk factor for progression to excessive GWG, but older age (≥32 years) was a protective factor, which appears controversial. The studies by Egan (2014) and Siegel (2015) did not find any association between age and excessive GWG [3,46], but in a cohort of 8184 healthy Brazilian women, the GWG decreased as the age increased [2].

A longer time of living with DM is expected in women with type 1 DM, who usually have lower pregestational BMI compared to those with type 2 DM [47]. Given our results for the effect of pregestational BMI on excessive GWG, it seems contradictory. However, the type of DM was not a risk factor for earlier excessive GWG in our sample. Therefore, we sugges<sup>t</sup> that longer years of living with DM may impact the metabolic and behavioral factors influencing GWG not covered in the present study.

Only one participant had yellow skin. This woman was overweight before pregnancy and had the highest GWG in our sample (28.2 kg). Therefore, even though we had found the yellow color of the skin to be a risk factor for progression to excessive GWG, we considered that it cannot be properly interpreted or discussed, as it was an isolated case with a very wide confidence interval. Three women who had excessive GWG in our sample reported the color of the skin as unknown.

A meta-analysis comparing the prevalence of excessive GWG among racial/ethnic groups found that White women were more likely to exceed the IOM guidelines than their Asian and Hispanic counterparts, but White and Black women had a similar prevalence of excessive GWG [48]. Differences in the GWG regarding the color of the skin are often related to socioeconomic discrepancies [48]. We found a marginal effect of inadequate housing conditions on the risk of earlier excessive GWG, but it was not statistically significant.

Of the nine women with hypothyroidism included in our sample, eight had excessive GWG, with a risk more than four times higher than that of women without hypothyroidism (aHR 4.37, CI 95% 1.62–11.77; *p* = 0.00). Collares et al. (2017) found that higher maternal TSH and lower free T4 levels in early pregnancy were associated with a higher GWG [49]. Hypothyroidism is a common endocrine disorder that occurs during pregnancy [50]. In our sample, all women diagnosed with hypothyroidism were treated with oral repositioning of the T4 hormone and had adequate hormone levels at the time of inclusion in the study; however, monitoring adherence to treatment and hormone levels during pregnancy was outside the scope of this study.

Identifying risk factors for earlier exceeding GWG and implementing a dietary intervention to mitigate it may help decrease the related adverse outcomes. The sooner the excess weight is present, the more that the metabolic effects can harm the mother and fetus [51]. Of particular interest for pregnan<sup>t</sup> women with DM is that an increase in maternal fat mass during early pregnancy can increase insulin resistance and thus worsen glycemic control [52]. As excess GWG does influence offspring obesity over the short- and long-term [53], faster fist and second trimester GWG but not third trimester was associated with higher mid-childhood adiposity in a cohort of 979 mother–child pairs from whom children were evaluated between 6.6 and 10.9 years of age [54].

Our study is novel in terms of several relevant characteristics. First, we investigated candidate genetic risk factors for progression to excessive GWG, and we did not find previous studies with this objective. Second, we investigated the influence of obesityrelated polymorphisms in Brazilian women with pregestational diabetes since studies enrolling Brazilian pregnan<sup>t</sup> women in this field are scarce and were not designed for women with DM. Additionally, women were administered two distinct types of diets, and we noticed that it had no effect on the risk of progression to excessive GWG in our sample.

However, this study had some limitations. First, we had a limited sample size to include in these analyses because the clinical trial was originally designed to analyze the effect of the two types of diets on perinatal outcomes without making use of the nutrigenetics approach. Given the increasing evidence regarding the effects of genetic characteristics and gene-diet interactions on BMI and obesity predisposition, we considered that it should gain more attention in the field of maternal and child nutrition. Therefore, we believe that this study will contribute to paving this way.

Furthermore, we had to make adaptations to maintain the study when the COVID-19 pandemic began in 2020. We used telemedicine to complete the follow-up of six women (8.6% of the sample) who were already enrolled in the study at the beginning of the pandemic quarantine. The same study protocol was used for the present visits. Prenatal visits to the physicians were maintained at the local level of the study, maintaining the measurements of weight at each visit. We also asked the participants to send a photograph of the prenatal card containing this information using a popular free smartphone application. Once the pandemic quarantine began, we did not include more participants in the study.

### **5. Conclusions**

In this study, we investigated obesity-related polymorphisms in FTO and ADRB2 genes as candidate genetic risk factors for excessive GWG in pregnan<sup>t</sup> women with pregestational DM using traditional or DASH diets.

Regardless the type of diet, the AT carriers of rs9939609 (FTO gene) had more than twice the risk of earlier exceeding GWG compared to the TT genotype, and the AA carriers of rs1042713 (ADRB2 gene) had almost four times higher risk than the GG carriers. The frequencies of these genotypes in our study population were 48.6% and 11.4%, respectively.

We found no effect of the genotypes of rs17817449 (FTO gene) and rs1042714 (ADRB2 gene) on the risk of progression to excessive GWG; however, the AG carriers for FTO haplotype rs9939609:rs17817449 had almost twice the risk of earlier exceeding GWG compared to TT carriers. Non-genetic characteristics associated with the risk of progression to GWG were time living with DM ≥ 8 years, pre-pregnancy overweight or obesity, and previous hypothyroidism. In contrast, age ≥ 32 years old was a protective factor.

Identifying women at a higher risk for exceeding GWG earlier may help improve nutritional interventions to mitigate this risk. The next step in advancing personalized nutrition is to understand which diet patterns may protect these women against excessive GWG and to investigate other genes and potential gene-diet-environment interactions with effects on GWG.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/nu14051050/s1, Table S1: Comparison of gestational weight gain (kg) between diet groups according to pregestational BMI (Rio de Janeiro/Brazil, 2016–2020); Figure S1: Gestational weight gain according to genotypes and dietary patterns.

**Author Contributions:** Conceptualization, K.d.S., E.L.R. and C.S.; data curation, K.d.S., G.P.B. and L.B.G.d.S.; formal analysis, K.d.S. and M.R.-A.; funding acquisition, E.L.R. and C.S.; investigation, K.d.S., A.C.P.d.F., G.P.B., L.B.G.d.S. and V.M.Z.; project administration, C.S.; writing—original draft, K.d.S.; writing—review and editing, E.L.R., J.A.M., and C.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received financial support from the Conselho Nacional de Desenvolvimento Científico e Tecnológico–CNPq (National Council for Scientific and Technological Development, gran<sup>t</sup> number: 409032/2016-6), the Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro—FAPERJ (Carlos Chagas Filho Foundation to Research in the State of Rio de Janeiro, gran<sup>t</sup> numbers: E-26/202.972/2016-6 and E-26/201.193/2021), and the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior–CAPES–Brasil (Coordination for the Improvement of Higher Education Personnel). This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior-Brasil (CAPES)—Finance Code 001.

**Institutional Review Board Statement:** The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Maternidade Escola da UFRJ (CAAE– 46913115.0.0000.5275; 15 July 2015). The study was registered in the Brazilian Registry of Clinical Trials (RBR-4tbgv6).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Acknowledgments:** Lenita Zajdenverg, Marcus Miranda, Karina Bilda de Castro Rezende, Rita Bornia, Joffre Amin Jr., and Jorge Rezende Filho for their support during the study.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
