Interest in the nutritional correlates of attention-deficit/hyperactivity disorder (ADHD) has grown over the last two decades, with substantial attention paid to the role of polyunsaturated fatty acids (PUFAs). Most of this research has focused on whether omega(n)-3 fatty acid levels, either consumed in the diet or measured in circulation, differ between individuals with and without ADHD. Results generally suggest that ADHD is associated with lower plasma n-3 fatty acid concentrations (1–3) however differences in fatty acid concentrations between ADHD and control groups often are small in magnitude (1) and n-3 supplement trials aimed at treating ADHD have reported mixed results (1, 3–5). One proposed reason for the small observed differences between ADHD and control groups (1, 6) is that consideration of circulating levels of n-3 fatty acids alone is not a sufficiently comprehensive indicator of fatty acid levels, as it does not take into account concentrations of n-6 fatty acids, which compete for the same enzymes for elongation and desaturation (7). Thus, absolute levels of n-3 or n-6 fatty acid concentrations may not be as meaningful of a metric as the ratio of n-6 to n-3 fatty acids (n-6:n-3). A greater n-6:n-3 has been linked with increased systemic inflammation and with increased risk for diseases with inflammatory underpinnings (8–10), though few studies have empirically examined blood levels of fatty acids and inflammatory markers together in the context of psychiatric disorders [see (11, 12) for exceptions], and none to our knowledge have investigated these associations among individuals with heightened symptoms of ADHD.

Research looking at the n-6:n-3 in individuals with and without ADHD is more limited than studies of n-3s alone (there have been fewer than 10 studies to date), but these studies also support a picture of altered fatty acid status in ADHD (6, 13, 14). As is true for the n-3 studies, there is some heterogeneity in the sample types used (e.g., plasma vs. serum vs. red blood cells) as well as in which fatty acids were used to compute the n-6:n-3. Most studies report the ratio of all measured n-6s to n-3s (the specific fatty acids considered varies from study to study), while some also report the ratio of arachidonic acid (AA; an n-6) to eicosapentaenoic acid (EPA; an n-3; AA:EPA), selected because they are precursors to important inflammatory mediators (15) and because they are important biologically active n-6 and n-3 fatty acids in the brain (6). With these study design differences aside, the extant studies typically report that the n-6:n-3 is increased in individuals with ADHD, even in the absence of differences in absolute values of n-3s or n-6s [for exceptions see (4, 16)]. For example, Parletta et al. (8), using a sample of children from Melbourne, Australia, found that the AA:EPA and the n-6:n-3 in red blood cells were elevated in children with ADHD relative to controls (mean age = 9.10 years; 80% male). Utilizing data from a sample of children living in Taiwan, Wang et al. (13) found that the n-6:n-3 was elevated in the serum of children with ADHD relative to controls (mean age = 9.20 years; 86% male). In another study, Stevens et al. (17), using a US sample (mean age = 9.10 years; 100% male), found that males with ADHD had greater plasma (but not red blood cell) n-6:n-3s, even though they did not differ in their absolute levels of n-6 fatty acids. Though these studies offer important preliminary information about the n-6:n-3 and ADHD, a number of important questions remain.

First, while the previous studies on this topic are informative, they have been conducted almost exclusively with samples of children or adolescents and have utilized data from samples that are predominantly male. While this focus on male youth is consistent with the known demographics of ADHD [which is more common in males (18) and is typically diagnosed in childhood], these demographics are not representative of all individuals with ADHD, and thus it is unclear if these results generalize to other populations with ADHD (e.g., adult females). ADHD is a neurodevelopmental disorder that is commonly diagnosed in childhood, however in about two-thirds of cases, clinically significant or impairing symptoms persist into adulthood (19, 20), making this an important population to study. Evaluating whether other subgroups of individuals with ADHD also show evidence of altered fatty acid concentrations is important, as this information may give further insight into the etiology and pathophysiology of ADHD, and because this information may help with the design of treatments that have potential to help a wider range of individuals.

Arguably, if there were to be a group of individuals with ADHD that these results would be least likely to extend to, it would be pregnant individuals. Previous research shows that there are sex differences in fatty acid synthesis that are the result of sex hormone-induced alterations in the expression of desaturase and elongase enzymes (21, 22); these differences, however, have typically been studied in non-pregnant populations. They are likely to be exaggerated during pregnancy given the normative and marked increase in sex-hormones that accompanies the pregnant state (23). Indeed, pregnancy is typically characterized by hyperlipidemia (24, 25), which appears to intensify in correspondence with pregnancy-related hormonal changes, and may be responsible for the depletion in long-chain fatty acid concentrations that appears to occur starting in the second trimester (26–28). These pregnancy-specific changes in fatty acid metabolism make our primary research question (i.e., whether the n-6:n-3 differs in pregnant individuals with and without ADHD) a non-trivial extension of previous studies in this area.

Second, there exists ambiguity as to whether ADHD-associated differences in the n-6:n-3 in plasma reflect differences in fatty acid metabolism or if they are due to ADHD-associated differences in dietary intake (either in the form of diet or n-3 supplementation). One hypothesis is that these disparities are the result of metabolic differences (29); this hypothesis is supported by research that suggests that there are ADHD-associated differences in fatty acid desaturase genes and phospholipid metabolism genes (30, 31). However, few studies have examined the n-6:n-3 in ADHD alongside measures of dietary intake, and the existing studies have yielded somewhat mixed results. While two studies (4, 32) suggest that youth with and without ADHD do not differ in their dietary intake of relevant fatty acids, at least one study reported that children with ADHD consumed less healthful foods, including more frequent consumption of high-fat and high-sugar foods, assessed using a food frequency questionnaire (13). This question has not yet been addressed during pregnancy, a time when dietary patterns may change to accommodate the nutritional requirements needed to support the developing child, and that may change in response to physiologic changes associated with food aversions, cravings, nausea, vomiting, tiredness, and heartburn (33). Importantly, the circulating fatty acid concentrations of the pregnant individual has implications for the health of both the pregnant person and the developing fetus (26–28).

To examine dietary-intake differences in individuals with and without ADHD, the gold standard is 24-h recall data conducted by trained dietitians. A second source of information about dietary intake comes from examining blood levels of essential fatty acids (those that cannot be synthesized by the human body, and thus must come from diet). Most relevant here are the essential fatty acids linoleic acid (n-6) and α-linolenic acid (n-3) (34). If n-6:n-3 differences in plasma are observed in the absence of differences in dietary intake (assessed either via recalled dietary intake or by concentrations of essential fatty acids), this would be consistent with the hypothesis that there are ADHD-associated differences in the metabolism rather than the consumption of fatty acids.

The third remaining question revolves around the relations among fatty acid concentrations, systemic inflammation, and ADHD. Increased systemic inflammation has been implicated in the etiology and pathophysiology of ADHD (35–37), yet limited empirical attention has been paid to the association between ADHD and inflammation in humans. The evidence that does exist generally suggests that, relative to controls, children (38–40) with ADHD have higher circulating levels of pro-inflammatory cytokines, including interleukin (IL)-6 and tumor necrosis factor-alpha (TNF-α). Whether these differences in cytokines are also apparent in pregnant individuals with ADHD remains untested, despite evidence that these cytokines are elevated in the context of other psychiatric symptoms during pregnancy (e.g., depression) (41, 42). There are normative changes in immune system functioning that occur during pregnancy to support healthy fetal growth and development (43, 44)), which again makes it challenging to extend the findings of previous research to this population. Additionally, although the n-6:n-3 has been shown to be related to greater inflammation in non-pregnant adults (45–47), its association with inflammation in pregnancy, and in the context of ADHD, has not yet been explored. This is important because it would have implications for both intervention and even for theories of mechanisms of intergenerational transmission.

To address these outstanding questions, the goal of the current study was to test:

We hypothesized that, on average, individuals with heightened symptoms of ADHD would have higher plasma n-6:n3s and higher plasma concentrations of IL-6 and TNF-α. We also hypothesized that a higher n-6:n-3 would be positively correlated with inflammation. Consistent with the hypothesis that there are fatty acid conversion or metabolic differences in ADHD, we hypothesized that there would be no ADHD-associated differences in dietary intake or in essential fatty acid concentrations.

In testing our hypotheses, we controlled for whether the pregnant individual reported that they were currently consuming a fish oil or fatty acid supplement. In models that considered inflammation, we also controlled for pre-pregnancy adiposity. These variables were selected as covariates a priori, given presumed effects of fatty acid supplementation on plasma fatty acid concentrations, and given the large literature linking adiposity and pro-inflammatory cytokine concentrations (48, 49).

Utilizing data from a unique sample of pregnant individuals, the current study examined whether there are ADHD-associated differences in polyunsaturated fatty acid or pro-inflammatory cytokine concentrations during pregnancy, a developmental period when maternal fatty acid status and inflammation have implications for the health of both the pregnant person and the developing child (27, 56–59). Results suggest that pregnant individuals with heightened ADHD symptoms, on average, had higher plasma n-6:n-3s and higher plasma concentrations of TNF-α, when compared to controls. The two groups did not differ in their dietary intake of fatty acids or other relevant nutrients, nor did they differ in concentrations of essential fatty acids (those that can only come from diet). While not conclusive, these results are consistent with the hypothesis that the observed differences in the plasma n-6:n-3 are the result of ADHD-associated differences in fatty acid metabolism, rather than simply reflecting differences in dietary intake.

The current study found that pregnant individuals with ADHD differed from controls in the n-6:n-3 using two metrics that have been used in previous research, the sum of total n-6s divided by the sum of total n-3s and the ratio of arachidonic acid to eicosapentaenoic acid concentrations (AA:EPA). Unsurprisingly, results were similar across these metrics. The observed ADHD group differences in the n-6:n-3 persisted even after controlling for whether the participant was taking a fish oil or fatty acid supplement or not. Despite n-3 supplementation being well-studied as a potential complementary treatment for ADHD (1, 60), few previous studies examining ADHD associated differences in n-3s or the n-6:n-3 have controlled for supplementation in their analyses.

The ADHD and control groups did not differ reliably in their self-reported 24-h dietary recalls of PUFA intake (total PUFAs and total n-3 fatty acids) or on other relevant nutrients, including total calories, total fat, MUFAs, SFAs and the PUFA:SFA ratio. There also were no reliable differences in plasma concentrations of conjugated linoleic acid and α-linolenic acid, fatty acids that can only come from dietary intake. Though not conclusive, our observation that there are differences in the plasma n-6:n-3 in the absence of differences in dietary intake is consistent with the purported ADHD-associated differences in the metabolism of fatty acids or in the conversion of essential fatty acids to longer-chain fatty acids (29). For example, single nucleotide polymorphisms in the fatty acid desaturase enzyme which have been shown to modulate circulating n-3 levels (61) have been associated with ADHD (30).

This study also examined whether pro-inflammatory cytokine concentrations differed between pregnant individuals with and without ADHD. We found that pregnant individuals with heightened ADHD symptoms, on average, had higher plasma concentrations of TNF-α, relative to controls. Previous research has shown that heightened gestational inflammation is associated with other types of psychiatric symptoms during pregnancy (e.g., anxiety, depression) (42, 51), but this is the first study to report that pregnant individuals with heightened ADHD symptoms have greater concentrations of pro-inflammatory cytokines. We did not see differences in plasma IL-6 between pregnant individuals with and without heightened ADHD symptoms, though this study may not have been adequately powered to detect such an effect. This is in contrast to previous research which reports differences in IL-6 between children with and without an ADHD diagnosis (38–40), but is consistent with findings from Corominas-Roso et al. which reports that adults with and without ADHD did not differ from one another in serum IL-6 concentrations (though, of note, these authors also did not find a differences in serum TNF-α) (62).

Consistent with the assertion that a greater n-6:n-3 is indicative of a more pro-inflammatory state, we found a positive association between the AA:EPA ratio and plasma concentrations of both IL-6 and TNF-α in models that did not include covariates. However, these effects did not survive controlling for pre-pregnancy BMI. Interestingly, in their study of pregnant individuals with prenatal depression, Chang et al. (12) similarly did not find a correlation between the n-6:n-3 and TNF-α, despite finding that the n-6:n-3 and TNF-α were increased in depression. The fact that individuals with ADHD are at heightened risk for obesity [see (63) for a systematic review]—which could be the long-term consequence of an increased n-6:n-3 rather than simply a confound–makes it challenging to fully interpret these findings.

The precise mechanistic connection between ADHD, PUFAs and inflammation is not fully elucidated, but see (64) for a comprehensive review on the association between PUFAs and inflammation. Fatty acids are critical nutrients in maintaining neuronal signaling (65). Additionally, the n-3 PUFA membrane content has been shown to influence neurotransmitter receptor functionality in dopaminergic neurons (66) which has implications for behaviors related to ADHD. In glial cells, levels of n-3 PUFAs can influence microglial phagocytosis (67) as well as astrocyte differentiation (68). Notably, microglia are the main cell type able to detect and respond to inflammatory signals in the brain. Further, microglia express several inflammatory receptors that have also been suggested to be sensitive to lipids (69). Animal models show that n-3 PUFA supplementation can inhibit microglial activation and the expression of pro-inflammatory cytokines, including TNF-a and IL-6 (70). It is plausible that altering microglial functional states via the n-6:n-3 in the brain may provide a connection between ADHD, PUFAs, and inflammation.

This study had a number of strengths. This is the first study to examine fatty acid or cytokine differences in pregnant individuals with and without heightened ADHD symptoms. Though differences in n-6:n-3 and pro-inflammatory cytokine concentrations have been observed in children with ADHD (6, 38, 39), no previous study has examined these associations during pregnancy, despite the profound implications that fatty acids and cytokines during pregnancy have for both the health of the pregnant person and for fetal development (26–28, 71). Methodological strengths of this study include assessment of fatty acids in plasma, the use of 24-h recall data collected by trained dietitians, and our inclusion of covariates that may impact fatty acid status and inflammation (most studies in this area have not considered covariates in their analyses).

This study also had limitations, including our small sample size. Though samples of this size are common in this literature, we may have lacked sufficient power to detect some effects, for example a difference in plasma IL-6 based on ADHD status. Thus, the null findings in this study should be viewed as inconclusive. In this first study investigating these associations in pregnant individuals, we focused on plasma concentrations of fatty acids, which reflect current fatty acid levels. Future studies on this topic should also examine fatty acid levels in red blood cell membranes, which reflect more chronic (i.e., the previous 3 months) levels of fatty acids. Examination of other plasma lipid levels (i.e., triglycerides) during pregnancy and may also yield important insights. While the unique nature of our sample (of pregnant individuals living in the US) is a strength of this study, it may limit generalizability of our findings. Future research should examine these associations utilizing data from participants residing in other countries as well as in non-pregnant populations. We examined two pro-inflammatory cytokines in our analyses (selected because they were shown to be related to ADHD in previous research) that were assessed at one timepoint in pregnancy. Future research should examine a wider range of relevant immune factors, assessed at multiple times in pregnancy. While our study included measurement of ADHD symptomatology assessed using a well-validated and widely used scale with established norms, we did not assess ADHD diagnosis, which also may limit the generalizability of our findings. Last, while metrics such as the n-6:n-3 offer a more comprehensive picture of fatty acid status than absolute values of n-3s or n-6s, the n-6:n-3 is also not a perfect or complete measure of fatty acid status. For example, trans fatty acids interfere with the desaturation and elongation of both omega-6 and omega-3 fatty acids (10). Future research should include other such metrics (including measures of SFAs, MUFAs, and measures of PUFAs not examined here, including docosapentaenoic acid) in their analyses.

This study found that, on average, pregnant individuals with heightened ADHD symptoms had higher plasma n-6:n-3s and higher TNF-α concentrations in plasma relative to controls. The heightened ADHD symptoms and control groups did not differ statistically on self-reported dietary intake (assessed using 24-h recalls), nor on essential fatty acid concentrations in blood. Though not conclusive, these findings are consistent with a picture of altered fatty acid metabolism in ADHD.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

All study procedures were reviewed and approved by the Oregon Health & Science University Institutional Review Board. Written informed consent to participate in this study was provided by the participants.

ES and JN conceived the project. HG, ES, JN, and JL designed the research. HG, ES, and JL analyzed the data. AM data created the plots. HG and ES wrote the manuscript, with contributions from JN, JL, GD, AM, and KH. All authors discussed the data. All authors contributed to the article and approved the submitted version.

Research reported in this publication was supported by the Abracadabra Foundation, the Bob and Charlee Moore Institute for Nutrition and Wellness, and by the National Institutes of Health under National Institute of Mental Health award numbers R3759105 (JN), R01MH117177 and R01MH124824 (ES and JN), K01MH120507 (HG), and R01MH117177-S1 (ES and JN). This material is the result of work supported with resources and the use of facilities at the VA Portland Health Care System and OHSU, Portland, OR.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Contents do not represent the views of the U.S. Department of Veterans Affairs or the United States Government.