The Institutional Animal Ethical Committee (IAEC), National College, Tiruchirappalli, India, and the European Communities Council Directive for the care and use of laboratory animals were strictly followed during all animal operations and were authorized prior to the start of the study under CPCSEA registration number 1941/GO/RE/S/17/CPCSEA. We purchased adult Wistar rats from the Indian Institute of Science in Bangalore. The focus of this study was only behavior. No animals were slaughtered or intrusive treatments carried out. Following behavioral evaluations, all rats were treated humanely and kept in the institutional animal facility under normal living circumstances in compliance with the Guide for the Care and Use of Laboratory Animals and CPCSEA.

Virgin female Wistar rats (12–14 weeks old, no siblings) were kept under standard circumstances (temperature 22 ± 2°C, humidity 55 ± 5%, 12 h light/dark cycle) with unrestricted access to tap water and standard rat food. After 7 days of acclimatization, males were paired with females periodically until mating was verified by the presence of a vaginal copulation plug (1–4 days later). Females were randomly allocated to four groups (initial n = 14 each):

The CD (Control Diet) followed the usual AIN-93G diet throughout.

The purpose of the RD-P and RD-W groups was to differentiate between the effects of gestational restriction and preconception on children. Non-pregnant females were eliminated after mating (RD-P: 6 excluded → 8 pregnant; RD-W: 4 excluded → 10 pregnant). For behavioral testing, offspring from pregnant mothers (n = 8–10 pups/group, balanced sex ratio) were utilized.

From gestation day 0 to postnatal day 21 (weaning), pregnant rats were placed on either the restricted diet (RD; 50% of all nutrients, both macro and micronutrients) or the control diet (CD; AIN-93G standard, 100% nutritional needs; Reeves et al., 1993) [20]. Diets were obtained from a certified commercial supplier. The RD-P and RD-W groups were transferred to CD during weaning and after giving birth, respectively. Every week, body weight and food intake were recorded (refer to Results, Table 1).

The offspring were evaluated for attention, hyperactivity, and behaviors associated with anxiety or impulsivity between postnatal days 35 and 40 (adolescence/young adulthood) [19–22]. After acclimating to the testing room for one hour, all tests were carried out during the light phase. In order to prevent olfactory signals, the apparatus was cleaned between experiments.

The arms of the Y-maze measured 40 cm in length, 3 cm in bottom width, 13 cm in upper width, and 15 cm in wall height. Rats were positioned in the middle, and video tracking was used to capture their spontaneous alternation for ten minutes.

A l t e r n a t i o n   p e r c e n t a g e = c o r r e c t   a l t e r n a t i o n s   t o t a l   a r m   e n t r i e s × 100

An arena of 50 cm × 50 cm × 50 cm with a center zone 10 cm from walls was utilized for the Open Field Test (locomotor activity/anxiety-like behavior). Ten minutes of time spent in the center and the total distance travelled were noted.

A plus-shaped maze was employed, with closed arms and 20 cm black walls. The arms were 40 cm long, 10 cm wide, and 50 cm high. Rats began by standing in the middle with their arms spread wide; their time there was noted for five minutes, which is the typical amount of time needed to measure anxiety.

All statistical analyses were performed using GraphPad Prism software (Version 10.2.0; GraphPad Software Inc., San Diego, CA, USA). Data are presented as mean ± standard deviation (SD). Comparisons among experimental groups were conducted using One-way ANOVA and Dunnett’s post-hoc test (control vs. each RD group) to examine the data. The mean ± SD is used to show the results. The threshold for significance was p < 0.05. There are 8–10 pups per batch. Every test is carried out as indicated in the Supplementary materials.

Throughout gestation and lactation, maternal food consumption and body weight were recorded weekly (Table 1). The RD and RD-W groups consumed 50% less food than the CD group during restriction periods (p < 0.001). Following the transfer to a control diet, food intake in RD-P normalized after parturition, and in RD-W after weaning. Compared to CD, restricted groups (RD and RD-W) saw substantially lower body weight growth throughout gestation (p < 0.01), indicating decreased calorie and nutrient consumption. After diet normalization, postnatal weight trajectories in the RD-P and RD-W groups improved to some extent. Pearson correlation studies revealed a modest negative relationship between gestational food intake and offspring Elevated Plus Maze (EPM) open-arm time (r = –0.34, p = 0.06), indicating that decreased maternal nutrition intake during pregnancy may lead to less anxiety-like behavior in children. There were no significant relationships between maternal food/body weight and Y-maze alternation or Open Field distance (p > 0.05) (Table 2).

No significant differences in alternation scores across groups (one-way ANOVA, F(3,36) = 1.45, p = 0.24). The RD group showed numerically higher scores compared to CD (mean difference = 18.00, 95% CI: –9.47 to 45.47, p = 0.21). RD-P and RD-W groups did not differ from CD (p = 0.61 and p = 0.95, respectively) (Figure 1).

Significant group effect on open-arm time (one-way ANOVA, F(3,36) = 12.34, p < 0.0001). All restricted groups spent more time in open arms compared to CD, indicating reduced anxiety-like behavior: RD (mean difference = 28.5 s, 95% CI: 17.55–39.45, p = 0.0002), RD-P (mean difference = 25.5 s, 95% CI: 14.55–36.45, p = 0.0004), RD-W (mean difference = 51.5 s, 95% CI: 40.55–62.45, p < 0.0001) (Figure 2).

Significant increase in distance traveled in RD vs. CD (mean difference = 31.00, 95% CI: 7.99–54.01, p = 0.0118). No significant differences in RD-P (p = 0.92) or RD-W (p = 0.9985) (Figure 3).

During crucial developmental windows, maternal global food restriction caused domain-specific behavioral changes in rat offspring. There was no discernible difference in working memory across groups as measured by spontaneous alternation in the Y-maze (p > 0.05). This result is consistent with a number of earlier studies that found that moderate gestational nutrient restriction had mild or context-dependent effects on attention and executive function. This could be due to compensatory hippocampus mechanisms or the short duration of restriction in the current model [18, 19]. In keeping with studies that attentional/executive deficits are frequently modest or task-specific in such contexts, the lack of a substantial deficit in alternation performance raises the possibility that working memory circuits may be resilient to mild prenatal undernutrition [20, 21].

On the other hand, all restricted groups showed a substantial increase in time spent in the open arms of the EPM, which demonstrated a reduction in anxiety-like behavior (p < 0.001). The highest impact was shown in RD-W offspring (p < 0.0001). This paradoxical decrease in anxiety-like traits is in line with new research that suggests undernutrition during pregnancy can reduce stress responsiveness and anxiety-related behaviors in rodent models. This may be due to changes in frontostriatal maturation, altered HPA-axis programming, or decreased threat sensitivity [22, 23, 24]. The idea that prenatal food restriction may rewire emotional regulatory circuitry toward less anxiety-like responses is supported by the fact that enhanced open-arm exploration is traditionally viewed as decreased anxiety-like behavior rather than increased impulsivity [25, 26].

Only the RD group showed a significant increase in hyperactivity, as shown by the total distance walked in the Open Field Test (p = 0.0118); following postnatal diet normalization, there were no discernible differences between the RD-P or RD-W groups. This pattern implies that postnatal dietary recovery permits partial flexibility in motor and arousal systems, whereas gestational restriction has a more noticeable impact on locomotor circuits [27, 28]. In keeping with differing developmental windows of sensitivity, the partial mitigation shown in the RD-P and RD-W groups further suggests that locomotor hyperactivity could be more responsive to postnatal intervention than anxiety-like traits [19, 29].

Moderately unfavorable trends were found by Pearson correlation analyses between the open-field distance (r = –0.28, p = 0.12) and EPM open-arm time (r = –0.34, p = 0.06) and gestational maternal food consumption. These correlations imply that whereas postnatal recovery of intake largely normalizes these effects, poorer nutritional availability during pregnancy may lead to decreased anxiety-like behavior and increased locomotor activity in offspring. Working memory may be less responsive to the level of maternal constraint in this paradigm, as seen by the lack of substantial associations with Y-maze performance. Dopaminergic synthesis, myelination, and the formation of the frontostriatal circuitry depend on a number of micronutrients, especially iron, zinc, choline, folate, and vitamin B12 [30, 31]. Iron deficiency causes dysregulation of attention and impulse control by impairing dopamine metabolism and D2 receptor function [32, 33]. While folate and B12 deficits impair methylation pathways and raise homocysteine, which impacts brain plasticity and synaptic maturation, zinc and choline promote neurogenesis and cholinergic transmission [34, 35]. Through energy deficit, changed insulin/IGF-1 signaling, and decreased substrate availability for brain development, macronutrient restriction most likely exacerbated these effects [36, 37]. Although the lack of direct neurochemical (dopamine levels, receptor expression) or neuroanatomical (prefrontal cortex histology) measurements in this study means that such interpretations remain conjectural and need further validation, these pathways offer tenable mechanisms for the observed behavioral profile.

This study has a few drawbacks, including lower final sample numbers owing to non-pregnant exclusions. To identify micronutrient vs calorie effects, no negative controls (for example, a pair-fed isocaloric group) were used. The 21-day gestational restriction model represents moderate exposure but may not depict chronic deficit. No molecular, neurochemical, or histological studies were performed. Behavioral paradigms give indirect assessments of ADHD-like features, but rodent models have intrinsic limitations in properly replicating human neuropsychiatric disorders. Future research should precisely distinguish between macronutrient and micronutrient impacts, target specific nutrients (iron, zinc, choline, B-vitamins), and use sophisticated models (patient-derived organoids, humanized mice, multi-omics) to unravel processes. Longitudinal designs with longer restriction periods, neurochemical assessments (e.g., dopamine levels, receptor expression), and brain histology are encouraged.

This study shows that maternal global nutritional restriction during pregnancy causes long-term behavioral changes in children, including reduced anxiety-like behavior and hyperactivity, which may be reversed partially after birth. These findings emphasize the importance of sufficient maternal nutrition in neurodevelopmental outcomes and encourage early nutritional treatments to reduce cognitive and behavioral vulnerabilities.