Role of Excessive Weight Gain During Gestation in the Risk of ADHD in Offspring of Women With Gestational Diabetes

Abstract

Context

Although attention-deficit/hyperactivity disorder (ADHD) has been associated with gestational diabetes mellitus (GDM) and maternal obesity, excessive weight gain (EWG) during pregnancy has scarcely been evaluated.

Objective

This study aimed to assess the joint effect of maternal weight and EWG on the risk of ADHD in offspring of GDM pregnancies.

Methods

In this cohort study of singleton births >22 weeks of gestation of women with GDM between 1991 and 2008, gestational weight gain above the National Academy of Medicine (NAM) recommendations was classified into EWG. Cox-regression models estimated the effect of maternal pregestational weight and EWG on the risk of ADHD (identified from medical records), adjusted for pregnancy outcomes and GDM-related variables.

Results

Of 1036 children who were included, with a median follow-up of 17.7 years, 135 (13%) were diagnosed with ADHD. ADHD rates according to pregestational maternal weight were 1/14 (7.1%) for underweight, 62/546 (11.4%) for normal weight, 40/281 (14.2%) for overweight, and 32/195 (16.4%) for obesity. Only maternal obesity was independently associated with ADHD (HR_adjusted 1.66 [95% CI, 1.07-2.60]), but not maternal overweight or EWG. On evaluating the joint contribution of maternal weight and EWG, maternal obesity with EWG was associated with the highest risk of ADHD (vs normal weight without EWG; HR_adjusted 2.13 [95% CI, 1.14-4.01]). Pregestational obesity without EWG was no longer associated (HR_adjusted 1.36 [95% CI, 0.78-2.36]).

Conclusion

Among GDM pregnancies, pregestational obesity was associated with a higher risk of ADHD in offspring. Nonetheless, when gestational weight gain was taken into account, only the joint association of obesity and EWG remained significant.

excessive weight gain, gestational weight gain, attention-deficit/hyperactivity disorder, gestational diabetes

Gestational diabetes mellitus (GDM) has been associated with an increased risk of neuropsychiatric disorders in offspring, such as attention-deficit/hyperactivity disorder (ADHD) (1, 2). Hyperglycemia may predispose fetuses to stress, chronic inflammation, hypoxia, and fetal hyperinsulinemia, which, in turn, may interfere with fetal brain development during critical prenatal windows, leading to neurobehavioral disorders later in life (3, 4).

In addition to the deleterious effects of hyperglycemia, in recent decades maternal obesity has emerged as one of the main risk factors for not only neonatal complications such as macrosomia, large for gestational age infants (LGA), prematurity, and perinatal mortality, but also for adverse long-term consequences in offspring mental health (3-5). Several population-based studies have described an association between maternal obesity and ADHD diagnosis in offspring (6-9). This association has also been described in pregnancies complicated with GDM (9).

However, despite the efforts to prevent maternal obesity, nowadays, roughly 30% of women in reproductive age are obese at the first antenatal visit, increasing up to 47% in pregnancies complicated by GDM (10, 11). In this setting, gestational weight gain (GWG) becomes a modifiable risk factor, since excessive GWG (EWG) has been associated with adverse pregnancy outcomes such as LGA, cesarean delivery, and a low Apgar score in both nondiabetic and diabetic populations (11-14). Moreover, in animal studies, maternal overnutrition during pregnancy triggers an inflammatory cascade, resulting in alterations in the fetal serotonergic system (15). Indeed, disturbances in fetal serotonergic system by certain drugs have been associated to neurodevelopment disorders later in life (16). Thus, it seems plausible that in human studies, EWG could also have adverse long-term consequences in mental health for offspring. Nevertheless, in contrast to maternal obesity, the few current studies published have failed to demonstrate an independent relationship between EWG during pregnancy and ADHD in offspring (6, 8), with no studies in a high-risk population such as GDM pregnancies.

With this background, the aim of this study was to investigate whether children of women with GDM exposed to maternal overweight and obesity are more likely to develop ADHD later in life and the role of EWG in these associations.

Methods

Study Population

This cohort study was composed of singleton children born between January 1, 1991, and December 31, 2008, to mothers with GDM in a university hospital (Hospital Universitari Mútua de Terrassa, Barcelona, Spain). Children who migrated outside the hospital catchment area or with no registered medical visits during follow-up, or no data about maternal pregestational body mass index (BMI) or GWG during pregnancy were excluded. The study protocol was conducted according to the principles of the Declaration of Helsinki and approved by the hospital Research Ethics Committee. The study and data analysis were conducted from June 4, 2020, to December 30, 2020.

The two-step approach recommended by the National Diabetes Data Group was used to diagnose GDM in our hospital along the study period (17). Women were screened for GDM with the 50-g 1-hour glucose challenge test. A positive glucose challenge test result was defined as a serum glucose level ≥ 140 mg/dL (7.8 mmol/L). Women with a positive glucose challenge test underwent a 3-hour, 100-g diagnostic oral glucose tolerance test (OGTT). GDM was diagnosed if 2 or more of 4 glucose thresholds were met: fasting plasma glucose ≥ 105 mg/dL (5.8 mmol/L), ≥190 mg/dL at 1 hour (10.6 mmol/L), ≥165 mg/dL at 2 hours (9.2 mmol/L), and ≥ 145 mg/dL at 3 hours (8.1 mmol/L). Gestational age at GDM diagnosis was calculated using the date of the OGTT that met the GDM diagnostic criteria. Treatment for GDM was based on national practice guidelines, using insulin as antidiabetes medication (oral drugs were not used) (18). Capillary glucose treatment goals to determine when insulin initiation or titration is needed were a fasting glucose <95 mg/dL (5.3 mmol/L) and a 1-hour postprandial glucose <140 mg/dL (7.8 mmol/L).

Main Exposure

Maternal weight was prospectively recorded during pregnancy. Anthropometric data were obtained as follows: patients were weighed with calibrated scales in light clothing without shoes to the nearest 0.1 kg. Height was measured to the nearest 0.5 cm. The BMI was calculated as weight in kilograms divided by height in square meters (kg/m²). Pregestational BMI was calculated based on self-reported maternal weight before pregnancy in the first antenatal visit and classified into 4 groups: underweight (BMI < 18.5 kg/m²), normal weight (18.5 kg/m² ≤ BMI < 25 kg/m²), overweight (25 kg/m² ≤ BMI < 30 kg/m²) and obese (BMI ≥ 30 kg/m²). GWG at the end of pregnancy was calculated as: final weight measured at the last antenatal visit − pregestational weight. According to the 2009 Institute of Medicine (currently NAM) guidelines, the rate of GWG was classified into insufficient, adequate, and excessive, if it was below, within, or above the recommendations as follows: 12.5 to 18 kg (underweight), 11.5 to 16 kg (normal weight), 7 to 11.5 kg (overweight), and 5 to 9 kg (obese) (19).

Main Outcome

ADHD was identified from medical records in accordance with International Classification of Diseases (ICD)-10 codes: F90 and F91 for ADHD. These codes include children with and without medical treatment.

Covariates

Maternal and pregnancy covariates were prospectively collected during gestation and included: maternal age at delivery, ethnicity, history of previous GDM, smoking at first antenatal visit, labor induction, preeclampsia (blood pressure 140/90 mmHg plus proteinuria above 300 mg/day (20)), birth year, gestational age at birth, newborn sex, birth weight, cesarean section, Apgar score at 5 minutes.

The birth weight for gestational age of the child was calculated according to Spanish fetal growth charts that take into account sex and gestational age (21) and categorized as small (birth weight below 10th percentile), LGA (birth weight above 90th percentile) and macrosomia (birth weight above 4000 g).

Statistical Analysis

Data are presented as mean ± SD, median [25th and 75th percentiles], or number (percentage) unless otherwise indicated. The Kruskal-Wallis, Pearson’s chi-squared test and ANOVA were performed, as appropriate, for comparisons between BMI categories. The Bonferroni test was used as a post hoc analysis to make pairwise comparisons, correcting for multiple analyses.

Cox proportional hazards modeling was used to estimate the association of exposure to maternal pregestational weight with ADHD. Pregestational weight was included in the models as a categorical variable according to BMI (normal, overweight, and obesity), GWG (adequate, insufficient, or excessive) or BMI according to EWG (normal weight with/without EWG, overweight with/without EWG, and obesity with/without EWG). The normal weight category included children of women with pregestational underweight because of the low number of cases in this group. A first adjusted model was performed, including birth year, smoking, maternal age, birth weight, birth sex, prematurity, ethnicity, and cesarean section. On the other hand, the use of insulin during pregnancy and early GDM diagnosis were related both with maternal weight and neuropsychiatric disorders in offspring. Thus, a second adjusted model was performed, including variables from model 1 plus insulin use during pregnancy and early GDM diagnosis (before 26 weeks of gestation). To overcome a possible selection bias due to the high rates of missing maternal data (pregestational BMI and GWG), multiple imputation in these variables were performed in the models aforementioned. Hazard ratios (HRs) with 95% CI were reported as measures of effect size.

Lastly, as a sensitivity analysis, we selected the whole cohort to estimate logistic regression models including multiple imputations in missing children (ADHD diagnosis) and maternal data (BMI and GWG). Adjusted models included the same covariates previously explained. Logistic regression models with P values <0.05 were considered statistically significant. All statistical calculations were performed with the STATA 14.0 (College Station, TX) statistical package.

Results

Participant Characteristics

A total of 1036 children were included (Fig. 1). Table 1 shows the maternal and neonatal characteristics at birth. The mean maternal pregestational BMI was 25.9 ± 5.4 kg/m², with 1.4%, 52.7%, 27.1%, and 18.8% of the mothers being classified as underweight, normal weight, overweight, and obese, respectively. The highest-weight categories (both overweight and obesity) were related to poor obstetric and neonatal outcomes such as preeclampsia, labor induction, cesarean section, LGA, and macrosomia. In addition, these groups were more likely to have an early GDM diagnosis (before 26 weeks of gestation) and insulin therapy during pregnancy (Table 1).

Table 1.

Baseline characteristics according to maternal prepregnancy BMI

	Total (n = 1036)	Underweight (n = 14)	Normal weight (n = 546)	Overweight (n = 281)	Obesity (n = 195)
Maternal characteristics
Maternal age (years)	33.0 ± 4.5	30.6 ± 3.5	32.6 ± 4.4	33.9 ± 4.7a,b	33.3 ± 4.5
European descendant	1003 (97)	14 (100)	539 (98.9)	266 (94.7)b	184 (94.9)b
History of GDM	143 (13.8)	1 (7.1)	51 (9.4)	44 (15.7)	47 (24.1)b
Current smoking	202 (19.5)	3 (21.4)	107 (19.6)	53 (18.7)	39 (20)
Birth year
1991 - 1993	34 (3.3)	0	18 (3.3)	14 (5)	2 (1)
1994 - 1996	155 (15)	1 (7.1)	90 (16.5)	42 (15)	22 (11.2)
1997 - 1999	166 (16)	2 (14.2)	85 (15.6)	50 (17.8)	29 (14.9)
2000 - 2002	235 (22.7)	5 (35.7)	115 (21.1)	62 (22.1)	53 (27.2)
2003 - 2005	238 (23)	5 (35.7)	136 (24.9)	52 (18.5)	45 (23.1)
2006 - 2008	207 (20.1)	1 (7.1)	102 (18.7)	61 (21.7)	44 (22.6)
Gestational weight gain
Weight gain (kg)	9.5 (6.5-12.2)	10.6 (8.4-13.7)	10.5 (8.1-12.9)	9 (5.6-12)a,b	6.5 (2.7-10) a,b,c
Inadequate	507 (48.9)	10 (71.4)	332 (60.8)	92 (32.7)	73 (37.4)
Adequate	348 (33.6)	3 (21.4)	175 (32.1)	108 (38.4)	62 (31.8)
Excessive	181 (17.5)	1 (7.1)	39 (7.1)	81 (28.8)b	60 (30.8)b,c
Obstetrical characteristics
Preeclampsia	39 (3.9)	0	6 (1.1)	9 (3.3)	24 (12.6)b,c
Labor induction	207 (23.1)	1 (10)	89 (18.9)	58 (23.9)	59 (33.9)b
Cesarean section	256 (24.8)	3 (21.4)	101 (18.6)	83 (29.5)b	69 (35.4)b
Neonatal characteristics
Female sex	481 (46.4)	8 (57.1)	249 (45.6)	136 (48.4)	88 (45.1)
Birthweight
Birthweight (g)	3240 (2920-3560)	2900 (2790-3400)	3185 (2900-3500)	3250 (2950-3580)	3350 (2970-3660)b
Macrosomia (≥4000 g)	59 (6.0)	0	22 (4.1)	16 (5.7)	21 (10.8)b
Large for GA (>90th centile)	145 (14.1)	0	57 (10.5)	45 (16.1)	43 (22.2)b
Small for GA (<10th centile)	100 (9.7)	2 (15.4)	58 (10.7)	22 (7.9)	18 (9.3)
GA at delivery (wk)	39.7 (38.7-40.6)	39.4 (38.4-40.4)	39.7 (38.7-40.7)	39.6 (38.7-40.4)	39.9 (38.4-40.6)
Preterm delivery (<37 wk)	76 (7.4)	2 (14.3)	32 (5.9)	25 (8.9)	17 (8.7)
Apgar ≤ 7 at 5 min	3 (0.3)	0	2 (0.4)	0	1 (0.5)
GDM-related variables
GA at GDM diagnosis (wk)	28.5 (26.5-30.5)	29 (27.5-30.5)	29.5 (27.5-31.5)	28.5 (27.5-30.5)b	27.5 (23.5-29.5)b,c
GDM diagnosed < 26 wk	181 (17.7)	3 (21.4)	66 (12.2)	46 (16.6)	66 (34.6)b,c
OGTT values
50-g OGTT (mg/dL)	175 (158-197)	165 (152-182)	170 (156-190)	178 (161.5-197)b	186 (168-208)b,c
100-g OGTT (mg/dL) 0’	86 (80-95)	84.5 (75-88)	83 (77-90)	89 (83-96)b	97 (87-106)a,b,c
60’	215 (200-230)	209.5 (186-225)	214 (199-227)	213 (200-227)	219 (205-239.5)b,c
120’	186 (173-200)	188.5 (184-203)	186 (174-200)	183 (170-197)	188 (172-206)
180’	144 (121-162)	165.5 (153-176)	144 (122-162)	143.5 (122-162)	140 (116-164)
Insulin therapy	420 (40.6)	3 (21.4)	156 (28.6)	134 (47.9)b	127 (65.5)a,b,c
GA of insulin therapy initiation (wk)	33.5 (29.5-35.5)	29.5 (29.5-38.5)	34.5 (32.5-36.5)	33.5 (31.5-35.5)	32.5 (27.5-34.5)b,c

Data are expressed as mean ± SD, median (Q1-Q3), or n (percentage) as appropriate. GWG was classified into inadequate, adequate, and excessive if it was below, within, or above the recommendations of the 2009 Institute of Medicine (currently NAM) guidelines: 12.5 to 18 kg (underweight), 11.5 to 16 kg (normal), 7 to 11.5 kg (overweight), and 5 to 9 kg (obesity).

Abbreviations: GA, gestational age; GDM, gestational diabetes mellitus; OGTT, oral glucose tolerance test; wk, weeks.

^aP < 0.05 vs underweight;

^bP < 0.05 vs normal weight;

^cP < 0.05 vs overweight.

Figure 1.

Flowchart of the cohort of singleton births born to women with GDM included in the analysis. Abbreviations: BMI, body mass index; GDM, gestational diabetes mellitus; GWG, gestational weight gain.

Maternal Prepregnancy BMI and EWG

Despite lower GWG in overweight and obesity pregnancies (9 and 6.5 kg, respectively), this weight gain was above NAM recommendations in 28.8% of overweight and 30% of obesity pregnancies compared with roughly 7% observed in under and normal weight pregnancies (Table 1). EWG during pregnancy was related to higher rates of adverse pregnancy outcomes in obesity pregnancies, compared to those with normal weight without EWG, such as preeclampsia (18.6 vs 0.8%, P < 0.001), cesarean section (40 vs 17.4%, P = 0.001), labor induction (39.6 vs 18.1%, P = 0.010), macrosomia (15.3 vs 3.1%, P = 0.004), and LGA (27.1 vs 9.7%, P = 0.004). Conversely, only the rates of cesarean section significantly differed in pregnancies with pregestational overweight and EWG compared to normal weight pregnancies (35.8 vs 17.4%, P = 0.003; Table 2).

Table 2.

Baseline characteristics according to maternal pepregnancy BMI and EWG

		Normal weight (BMI < 25 kg/m2)		Overweight (BMI 25-29 kg/m2)		Obesity (BMI ≥ 30 kg/m2)		P between groups
		non-EWG (n = 520)	EWG (n = 40)	non-EWG (n = 200)	EWG (n = 81)	non-EWG (n = 135)	EWG (n = 60)
Maternal characteristics
Maternal age (years)		32.6 ± 4.4	32.1 ± 4.5	33.8 ± 4.6	34.0 ± 4.8	33.6 ± 4.7	32.6 ± 4.2	<0.001
European descendant		513 (98.8)	40 (100)	186 (93)a	80 (98.8)	125 (93.3)a	59 (98.3)	<0.001
History of GDM		48 (9.3)	4 (10)	36 (18)a	8 (9.9)	36 (26.7)a,d	11 (18.3)	<0.001
Smoking		101 (19.4)	9 (22.5)	34 (17)	19 (23.5)	25 (18.5)	14 (23.3)	0.779
Gestational weight gain (kg)		10.1 (8-12.2)	18.3 (16.5-20)a	7.2 (4.6-9.4)a,b	14 (12.5-16)a,b,c	4.3 (1.4-6.5)a,b,c,d	12 (10.6-13.3)a,b,c,d,e	<0.001
Obstetrical characteristics
Preeclampsia		4 (0.8)	2 (5.1)	8 (4.1)	1 (1.3)	13 (9.9)a	11 (18.6)a,c,d	<0.001
Labor induction		81 (18.1)	9 (28.1)	39 (22.8)	19 (26.4)	38 (31.4)a	21 (39.6)a	0.001
Cesarean section		90 (17.4)	14 (35)	54 (27)	29 (35.8)a	45 (33.3)a	24 (40)a	<0.001
Neonatal characteristics
Female sex		237 (45.6)	20 (50)	98 (49)	38 (46.9)	62 (45.9)	26 (43.3)	0.950
Birthweight
Birthweight (g)		3170 (2900-3460)	3490 (3105-3650)	3210 (2925-3550)	3287 (3050-3670)	3300 (2950-3660)	3430 (3120-3710)a	<0.001
Macrosomia (≥4000 g)		16 (3.1)	6 (15)	9 (4.5)	7 (8.8)	12 (8.9)	9 (15.3)a	<0.001
Large for GA (>90th centile)		50 (9.7)	7 (17.5)	29 (14.5)	16 (20)	27 (20)a	16 (27.1)a	<0.001
Small for GA (<10th centile)		57 (11.1)	3 (7.5)	20 (10)	2 (2.5)	14 (10.4)	4 (6.8)	0.243
GA at delivery (wk)		39.7 (38.7-40.7)	39.9 (38.7-40.4)	29.4 (38.6-40.4)	39.7 (38.9-40.4)	39.7 (38.9-40.4)	39.9 (38.5-40.7)	0.879
Preterm delivery (<37 wk)		31 (5.9)	3 (7.5)	18 (9)	7 (8.6)	11 (8.2)	6 (10)	0.666
Apgar ≤ 7 at 5 min		2 (0.4)	0	0	0	0	1	0.338
GDM-related variables
GA at GDM diagnosis (wk)		29.5 (27.5-31.5)	29.5 (27.5-33)	28.5 (26.5-30.5)	29.5 (27.5-31.5)	26.5 (22.5-29.5)a,c,d	27.5 (24.5-30.5)a	<0.001
GDM diagnosed < 26 wk		63 (12.2)	6 (15)	33 (16.7)	13 (16.4)	48 (36.4)a,c,d	18 (30.5)a	<0.001
OGTT values
50-g OGTT (mg/dL)		170 (156-190)	169 (153-194)	179 (160-197)	176.5 (161.5-200.5)	186.5 (168-206.5)a	184.5 (165-215)a	<0.001
100-g OGTT (mg/dL)	0’	83 (77-90)	83.5 (76-90.5)	89 (83-90)a	89 (83-101)a,b	95 (87-106)a,b,c	100 (87-107)a,b,c,d,e	<0.001
	60’	214 (199-228)	205 (190-222)	213 (201-227)	216 (196-231)	219 (205-239)a,b	220 (203-240)	0.009
	120’	186 (175-200)	187 (175-201)	182 (171-195)	187 (169-204)	189(172.5-206.5)	187 (168-204)	0.044
	180’	144 (124-164)	145 (115-159)	140 (121-158)	149 (122-168)	140 (115-166)	139 (120-158)	0.391
Insulin therapy		142 (27.4)	17 (42.5)	86 (43.2)a	49 (59.3)a	86 (64.2)a,c	41 (68.3)a,c	<0.001
GA at insulin therapy initiation (wk)		34.5 (32.5-36.5)	35.5 (34.5-36.5)	33.5 (29.5-35.5)	34.5 (32-36.5)	32.5 (25.5-34.5)a,c	33.5 (28.5-35.5)	<0.001

Data are expressed as mean ± SD, median [Q1-Q3], or n (percentage) as appropriate. Normal weight included n = 14 with weight < 18.5 kg/m².

Abbreviations: BMI, body mass index; EWG, excessive weight gain; GA, gestational age; GDM, gestational diabetes mellitus; OGTT, oral glucose tolerance test; wk, weeks.

^aP < 0.05 vs normal weight w/o EWG;

^bP < 0.05 vs normal weight w/ EWG;

^cP < 0.05 vs overweight w/o EWG;

^dP < 0.05 vs overweight w/ EWG;

^eP < 0.05 vs obesity w/o EWG._,

Risk of ADHD According to Maternal Prepregnancy BMI

Children were followed a median of 17.7 years [14.8-21.7], with 135 (13%) incident cases of ADHD. The rates of ADHD according to maternal BMI were 1/14 (7.1%) for underweight women, 62/546 (11.4%) for normal weight, 40/281 (14.2%) for overweight, and 32/195 (16.4%) for women with obesity (P = 0.250).

After adjustment for well-known risk factors such as birth year, smoking, maternal age, birth weight, birth sex, prematurity, ethnicity, and cesarean section, offspring of women with obesity showed a higher risk of ADHD (adjusted HR 1.59 [95% CI, 1.05-2.41]) (Table 3). When the model included GDM-related variables, such as time of GDM diagnosis or insulin use during pregnancy, the association between maternal obesity and ADHD remained largely unchanged (adjusted HR 1.66 [95% CI, 1.07-2.60]). Conversely, maternal overweight was not related to the incidence of ADHD in offspring in either crude or adjusted models (Table 3).

Table 3.

Risk of ADHD in offspring of pregnancies complicated by gestational diabetes mellitus according to maternal prepregnancy BMI and GWG

	ADHD		Unadjusted model		Adjusted model 1		Adjusted model 2
	n/N (%)	Risk difference (%) (95% CI)	HR (95% CI)	P	HR (95% CI)	P	HR (95% CI)	P
Maternal weight by BMI
Normala (<25 kg/m2)	63/560 (11.3)	1 (reference)	1 (reference)		1 (reference)		1 (reference)
Overweight (25-29 kg/m2)	40/281 (14.3)	2.89 (−1.83 to 7.61)	1.30 (0.89-1.91)	0.181	1.32 (0.89-1.97)	0.167	1.33 (0.88-2.00)	0.160
Obesity (≥30 kg/m2)	32/195 (16.4)	5.35 (−0.27 to 1.10)	1.57 (1.05-2.36)	0.030	1.59 (1.05-2.41)	0.029	1.66 (1.10-2.67)	0.025
Maternal weight by GWG
Adequate	44/348 (12.6)	1 (reference)	1 (reference)		1 (reference)		1 (reference)
Inadequate	61/507 (12)	−0.6 (−5.10 to 3.8)	0.92 (0.63-1.37)	0.715	0.92 (0.62-1.37)	0.688	0.93 (0.63-1.39)	0.729
Excessive	30/181 (16.6)	3.9 (−2.5 to 10.4)	1.30 (0.81-2.09)	0.274	1.27 (0.79-2.07)	0.315	1.22 (0.75-1.99)	0.430
Maternal weight by BMI and GWG
Normal weighta w/o EWG	60/520 (11.5)	1 (reference)	1 (reference)		1 (reference)		1 (reference)
Normal weight w/ EWG	3/40 (7.5)	−4.0 (−12.7 to 4.6)	0.67 (0.21-2.14)	0.446	0.67(0.21-2.16)	0.503	0.70 (0.11-2.26)	0.554
Overweight w/o EWG	27/200 (13.5)	1.96 (−3.5 to 7.4)	1.20 (0.74-1.93)	0.473	1.23 (0.76-2.01)	0.397	1.28 (0.79-2.08)	0.311
Overweight w/ EWG	13/68 (16.1)	4.5 (−3.9 to 13.0)	1.39 (0.77-2.54)	0.311	1.41 (0.76-2.61)	0.274	1.37 (0.72-2.60)	0.338
Obesity w/o EWG	18/135 (13.3)	1.8 (−4.6 to 8.2)	1.21 (0.72-2.06)	0.528	1.23 (0.72-2.09)	0.443	1.36 (0.78-2.37)	0.280
Obesity w/ EWG	14/60 (23.3)	11.8 (0.7 to 22.8)	2.14 (1.19-3.84)	0.012	2.09 (1.15-3.80)	0.016	2.13 (1.14-4.01)	0.018

Data are n/N (%) and hazard ratio (HR) and 95% CI. GWG was classified into inadequate, adequate, and excessive if it was below, within, or above the recommendations of the 2009 Institute of Medicine (currently NAM) guidelines: 12.5 – 18 kg (underweight), 11.5 – 16 kg (normal), 7 – 11.5 kg (overweight), and 5 - 9 kg (obesity).

Abbreviations: ADHD, attention-deficit/hyperactivity disorder; BMI, body mass index; EWG, excessive weight; GWG, gestational weight gain; w/, with; w/o, without.

^aNormal weight included n = 14 with weight < 18.5 kg/m².

Model 1 included birth year, smoking, maternal age, birth weight, birth sex, prematurity, ethnicity, and cesarean section.

Model 2 included model 1 plus insulin use during pregnancy and early GDM diagnosis (before 26 weeks of gestation)

Risk of ADHD According to GWG

The median maternal GWG was 9.5 kg [6.5-12.2], with no significant association with ADHD in offspring in either the crude (HR 0.99 [95% CI, 0.96-1.02], P = 0.574) or the fully adjusted model (adjusted HR 0.98 [95% CI, 0.95-1.02], P = 0.335). Neither was any significant association found when GWG was included in the analysis as a categorial variable (appropriate, inadequate, or excessive) (Table 3).

Lastly, the joint contribution of maternal prepregnancy weight and GWG in the risk of ADHD was evaluated, grouping children according to maternal BMI and EWG. Fig. 2 depicts the crude cumulative incidences of ADHD by maternal BMI categories and EWG. Maternal obesity was associated with a higher risk of ADHD in offspring of women with EWG compared to those with normal weight without EWG (adjusted HR 2.13 [95% CI, 1.14-4.01]), unlike obesity without EWG (adjusted HR 1.36 [95% CI, 0.78-2.36]). These observations were uniform in the unadjusted and adjusted models. As mentioned previously, maternal overweight was not associated with ADHD (adjusted HR 1.37 [95% CI, 0.72-2.60]), even after including EWG in the model (Table 3). In addition, a sensitivity analysis was performed including the whole cohort with multiple imputation in missing data. After imputation, the results remained largely unchanged (Supplemental table (22)).

Figure 2.

Crude cumulative incidence of attention-deficit/hyperactivity disorder by maternal pregestational BMI and EWG. Normal weight included n = 14 with weight < 18.5 kg/m². Abbreviations: BMI, body mass index; EWG, excessive weight gain; w/, with; w/o, without.

Discussion

In the present study of prospectively collected clinical data from pregnancies complicated by GDM, children exposed to maternal obesity were found to be more likely diagnosed with ADHD. Nonetheless, although maternal weight gain was not independently related to ADHD, a higher risk of ADHD in offspring was observed if EWG was present in women with pregestational obesity. Indeed, maternal obesity without EWG was no longer associated with ADHD in offspring. To the best of our knowledge, no previous study has evaluated the long-term consequences of GWG and maternal weight (independently and together) on mental health in the offspring of GDM pregnancies.

The incidence of neurodevelopmental disorders has increased in the last years, suggesting that maternal factors could play a role. In this regard, exposure to maternal hyperglycemia (both pregestational diabetes and GDM) in the prenatal period increases the likelihood of the diagnosis of ADHD (2, 9, 23-25). However, as mentioned previously, it is not only glycemic control that plays a role in neonatal health, especially in GDM pregnancies. Maternal obesity and EWG have been associated with a higher risk of metabolic disturbances in offspring later in life, including childhood obesity and a higher risk of cardiovascular disease (26-28). However, data about the role of maternal weight in the risk of neurodevelopmental disorders in offspring have mainly been focused on pregestational maternal weight. A large nationwide Finnish cohort study including 649 043 newborns found a HR for ADHD in children of 1.15 (95% CI, 1.01-1.30) among women with GDM compared with the nondiabetic population, increasing up to 1.64 (95% CI, 1.42-1.88) in women with pregestational obesity (9). Nevertheless, some GDM characteristics strongly related to both obesity and neurodevelopment disorders were not taken into account, such as early GDM diagnosis and insulin use during gestation (29, 30). Our results showed a HR for ADHD in children of 1.67 (95% CI, 1.04-2.65) in women with obesity compared with the normal-weight group. The inclusion of GDM-related variables in the fully adjusted model reinforces the long-term impact of maternal obesity during the prenatal period on the fetal brain.

Interestingly, we observed that maternal weight gain was not independently related to ADHD, but when it was evaluated jointly with maternal obesity, the presence of both entities was related to the highest risk of ADHD in offspring (HR 2.14; 95% CI, 1.14-4.02). Indeed, maternal obesity without EWG was no longer associated with ADHD in offspring. Nowadays, an adequate GWG has been established as one of the main goals during gestation. The U.S. Preventive Services Task Force recommends that clinicians should offer to pregnant women effective behavioral counseling interventions aimed at promoting healthy weight gain and preventing EWG in pregnancy (31). Those recommendations were based on large evidence from systematic review and meta-analysis that antenatal lifestyle interventions were associated not only with a reduction in GWG, but also with a reduction of adverse maternal outcomes (32, 33). Nonetheless, to date, few studies have evaluated long-term consequences of EWG on mental health. In a cohort study including 331 children aged 2 to 6 years, GWG (crude or adjusted for maternal BMI) was not related to ADHD symptoms (6). Conversely, Pugh et al observed that the offspring of overweight mothers with high GWG had a greater number of impulsivity errors compared with their counterparts with average GWG (8). Unlike our study, the primary outcome of previous studies was dimensional measures of ADHD symptoms and executive function behaviors, which may be related to a number of psychiatric conditions, not only ADHD. Moreover, we selected older children (median age of 17 years), in whom neurodevelopment disorders were fully developed. Altogether, these findings reinforce the important role of an adequate GWG as previously observed on maternal and neonatal outcomes, as well as the hypothesis observed in animal models that maternal overnutrition during pregnancy triggers an inflammatory cascade, resulting in alterations in the fetal serotonergic system (closely related to behavioral disorders) (15).

Several strengths of this study should be highlighted. We describe the longest follow-up (almost 20 years) evaluating the long-term effects of maternal weight in the prenatal period. Moreover, maternal weight was collected prospectively during gestation, thereby avoiding memory bias. Second, ICD-10 codes were selected for ADHD diagnosis because of the complexity of the diagnosis. The robustness of these codes is supported by recently published data of a sample of 6834 students aged 5 to 17 in Spain reporting an overall prevalence of ADHD using Diagnostic and Statistical Manual of Mental Disorders criteria, comparable to a previous study from our group using ICD-10 codes (2, 34). In addition, only children undertaking regular visits with a pediatrician/physician were included in the analyses to minimize ascertainment bias. Third, the same diagnostic criteria for GDM and treatment targets were applied throughout the data collection period. Finally, although protocols of obstetric and neonatal management have changed over time, the birth year was included in the adjusted models to overcome this plausible bias.

Nonetheless, we also acknowledge some limitations. First, potential confounding owing to maternal neuropsychiatric disorders, paternal risk factors, and socioeconomic status could not be evaluated because of the lack of data. However, maternal and delivery risk factors (including GDM-related variables such as insulin use or early GDM diagnosis) were taken into account in the multivariate models. In fact, prematurity and low birthweight have consistently been associated with ADHD, with family studies suggesting that these effects cannot be explained by genetic confounding (35). Second, there was no control group without diabetes, therefore these results cannot be extrapolated to pregnancies without this complication. Third, pregestational BMI was based on self-reported weight, which could lead an underestimation of maternal BMI (36). Nonetheless, other anthropometric measures, such as final weight and maternal height have been prospectively collected. Further, previous data from our cohort using self-reported BMI confirm the association between maternal obesity and adverse pregnancy outcomes (37). Fourth, NAM guidelines may have some limitations in clinical application due to the fact that total GWG is strongly correlated with gestational age. To overcome this bias, different GWG z-score charts have been created (38-40); however, none of them in Spanish population. Thus, NAM guidelines with broader evidence in GDM pregnancies (11, 12) and recommended for different societies were applied (19, 31). Moreover, rates of prematurity were not different between BMI categories, and to rule out a possible bias, preterm birth has been included in the adjusted models. Fifth, despite that the study cohort had ongoing follow-up, specific data about which physician (pediatrician or psychiatrist) was doing the follow-up assessment, as well as the requirement of medical treatment, were lacking. Lastly, this was an observational study, and therefore, causal inferences cannot be drawn.

In conclusion, the results of this study suggest that the negative repercussions of EWG on children within the setting of a high-risk population (GDM with obesity) were not only observed during the prenatal period but also years later with a development of ADHD. In addition, the loss of association between maternal obesity and ADHD in offspring when NAM targets were not exceeded, highlights that promoting healthy gain during pregnancy should be a priority in the current management of gestation. Nonetheless, future studies with larger sample sizes in broader populations are needed to confirm these results.

Abbreviations

Abbreviations
 
• ADHD

  attention-deficit/hyperactivity disorder

 
• BMI

  body mass index

 
• EWG

  excessive weight gain

 
• GDM

  gestational diabetes mellitus

 
• GWG

  gestational weight gain

 
• HR

  hazard ratio

 
• ICD-10

  International Classification of Diseases, Tenth Revision

 
• LGA

  large for gestational age

 
• NAM

  National Academy of Medicine

 
• OGTT

  oral glucose tolerance test

Acknowledgments

We are grateful to Donna Pringle for help in the writing and editing of the manuscript.

Funding Sources

V.P. received a research grant from Fundació Docència i Recerca MútuaTerrassa, “Beca FMT d’Intensificació per a professionals de la Salud MT 2021”

Author Contributions

All authors have seen and approved the final version of the manuscript. A.S.-S., C.Q., N.A.-C., M.V., and X.U. acquired data and reviewed the manuscript. E.L. and M.J.B. reviewed/edited the manuscript. A.J.A., reviewed/edited the manuscript and contributed to the discussion. V.P. contributed to the study concept and design, supervised the study, researched data, participated in data analysis and interpretation, wrote the manuscript, and had final responsibility for the decision to submit for publication. V.P. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Disclosures

The authors report no potential conflicts of interest.

Data Availability

All datasets generated during and analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.