Parental Age Effects on Odor Sensitivity in Healthy Subjects and Schizophrenia Patients

Abstract

A schizophrenia phenotype for paternal and maternal age effects on illness risk could benefit etiological research. As odor sensitivity is associated with variability in symptoms and cognition in schizophrenia, we examined if it was related to parental ages in patients and healthy controls. We tested Leukocyte Telomere Length (LTL) as an explanatory factor, as LTL is associated with paternal age and schizophrenia risk. Seventy-five DSM-IV patients and 46 controls were assessed for detection of PEA, WAIS-III for cognition, and LTL, assessed by qPCR. In healthy controls, but not schizophrenia patients, decreasing sensitivity was monotonically related to advancing parental ages, particularly in sons. The relationships between parental aging and odor sensitivity differed significantly for patients and controls (Fisher’s R to Z: χ²=6.95, P=0.009). The groups also differed in the association of odor sensitivity with cognition; lesser sensitivity robustly predicted cognitive impairments in patients (<0.001), but these were unassociated in controls. LTL was unrelated to odor sensitivity and did not explain the association of lesser sensitivity with cognitive deficits.Parental aging predicted less sensitive detection in healthy subjects but not in schizophrenia patients. In patients, decreased odor sensitivity strongly predicted cognitive deficits, whereas more sensitive acuity was associated with older parents. These data support separate risk pathways for schizophrenia. A parental age-related pathway may produce psychosis without impairing cognition and odor sensitivity. Diminished odor sensitivity may furthermore be useful as a biomarker for research and treatment studies in schizophrenia.

INTRODUCTION

The risk for schizophrenia is consistently linked to advancing paternal age (APA), with up to a threefold increase in risk to the offspring of older fathers compared to those of younger fathers [Malaspina et al., 2001; Brown et al., 2002a]. Some studies furthermore show that maternal age is related to an increased offspring risk after controlling for the paternal age risk, with increased risks associated with both younger mothers (<20 years) [El-Saadi et al., 2004] and older mothers (>35 years) [Hultman et al., 1997]. Some portion of the risk related to paternal age is likely to be conveyed by de novo mutations in the male germ line, which accumulate with paternal age-related risk [Kong et al., 2012; Kranz et al., 2015], although such events do not account for the entire paternal age-related attributable risk [Zammit et al., 2003; Jaffe et al., 2014]. Social factors and parental susceptibility genes are often proposed, but these would still entail biological mechanisms that could influence the phenotype. Our work in the current volume shows that paternal aging is also associated with psychosis in probands with bipolar disorder, showing that the risk pathway for psychosis is not restricted to schizophrenia [Lehrer et al., 2015].

Knowledge of how parental ages influence the phenotype of schizophrenia might provide information on the mechanisms that link these exposures to psychosis. Odor sensitivity is an interesting phenotype to study in schizophrenia; it is associated with variability in both symptoms and cognition [Turetsky et al., 2007; Malaspina et al., 2012a, b], but it has not been examined with respect to parental ages. Odor sensitivity is easily assessed using well-standardized tasks that assess the concentration threshold at which a subject can detect phenyl ethyl alcohol (PEA) [Doty et al., 1995]. If odor sensitivity is related to a particular etiology or neurobiology associated with psychosis, it could prove useful as a biomarker for individualized treatment approaches. Odor sensitivity is additionally related to personality attributes in healthy people [Croy et al., 2011], so influences of parental ages on odor sensitivity in control subjects may also be of interest.

As APA and a positive family history of psychosis are associated with elongated leukocyte telomere length (LTL) in schizophrenia patients [Malaspina et al., 2014], LTL assessments may be related to a phenotype that is associated with parental ages in the disease. It is possible that longer telomeres may be protective against insults to olfactory regions [Watabe-Rudolph et al., 2011]. Thus, odor sensitivity could plausibly be influenced by telomere length. In this study, we considered if parental ages at parturition influenced odor sensitivity in healthy controls and schizophrenia patients and conducted a pilot evaluation to examine the associations of odor sensitivity with LTL and cognition.

MATERIALS AND METHODS

Data were collected under NIMH funded and IRB approved studies at New York State Psychiatric Institute and NYU/Bellevue Psychiatric Center and all subjects were part of the Genomic Psychiatry Cohort (GPC). For inclusion in this study, patients with a primary diagnosis of schizophrenia or schizoaffective disorder were recruited from clinical settings. All patients had been on stable medication doses for at least 1 month and healthy controls were recruited from local advertisements and internet postings. All subjects provided written informed consent. Diagnosis was determined through a structured interview with the Diagnostic Interview for Genetic Studies (DIGS), which included all Diagnostic Interview for Psychosis and Affective Disorders (DI-PAD) criteria as well [Nurnberger et al., 1994]. For this pilot study, inclusion criteria for control subjects required the absence of any Axis I diagnoses and a personal or family history of psychosis. Patients were clinically stable and on unchanging medication regimens for the duration of the study.

Odor Sensitivity

The study employed the Smell Threshold Test (STT; Sensonics, Inc.), measuring ability to detect various concentrations of a single odor, PEA, which has a rose odor. Subjects were asked not to smoke, use cosmetics or perfume on the day of testing and did not eat or drink for at least 2 hr before testing; furthermore, they were screened prior to testing to rule out colds or allergies. For the STT, we used a single staircase forced-choice procedure, applying the odorant to each nostril separately, whereas occluding the opposite nostril with foam tape. The first dilution to be tested was the −6.00 log concentration; then, we applied the odorant in increasingly higher concentrations in full log steps until five consecutive correct detections occurred in a given trial. The staircase was then reversed and moved down or up in half log increments with two pairs of trials at each concentration. The mean of the last four staircase reversal points was used as the measure of threshold. Higher raw STT scores were transformed to absolute values, such that the higher score represents greater sensitivity. Testing of odor sensitivity was conducted for the right and left nostril; however, our results indicated only small differences between them, so the data for both sides were combined into a mean odor sensitivity used in these analyses.

Leukocyte Telomere Length (LTL)

DNA was extracted from lymphocytes to determine relative Telomere Length in a subgroup of healthy controls (N 16) and schizophrenia patients (N 35). TL was analyzed by quantitative polymerase chain reaction (qPCR) as previously described [Cawthon, 2002], using an iCycler real-time PCR system (Bio-Rad Laboratories, Hercules, CA) with the following modifications. For the PCR reaction, each well contained: 4 μl of sterile water, 5 μl (35 ng) of genomic DNA, 0.5 μl of each primer, and 10 μl of iQ SYBR Green Supermix Bio-Rad (reaction buffer with dNTPs, iTaq DNA polymerase, 6 mM MgCl2, SYBR Green I, fluorescein, and stabilizers), and 7 μg/ml of DNA. The telomerictelomeric primers were Telo-F (CGG TTT GTT TGG GTT TGG GTT TGG GTT TGG GTT TGG GTT) and Telo-R (GGC TTG CCT TAC CCT TAC CCT TAC CCT TAC CCT TAC CCT). The primers for the reference control gene, 36B4 single copy gene, were 36B4-F (CAG CAA GTG GGA AGG TGT AAT CC) and 36B4-R (CCC ATT CTA TCA TCA ACG GGT ACA A). The concentration of all four primers was 10 mM and PCR cycle conditions consisted of an initial denaturation step at 95°C for 10 min, followed by 40 cycles of 95°C for 15 sec and 60°C for 60 sec. The next steps were: 1 min at 95°C, 1 min at 55°C, and 81 repeats of 10 sec at 55°C. The reactions were set-up in triplicate in 96-well plates. Each plate included five DNA quantity standards (serial dilutions of a reference DNA Hela cells giving final DNA quantities of between 28 and 1.75 ng/μ per reaction), one negative control, and one internal control represented by 1,301 line cells. Each sample was run twice on different plates. If the difference between the mean of the three values on one plate differed by >15% from the mean of the three samples of the 2nd plate, the sample was run a third time and the mean of the two closest measurements was used for the analysis. A standard curve was made by serial dilutions of known amounts of DNA from Hela cells. The amount of each subject’s DNA sample was determined relative to the single reference DNA sample by the standard curve method. The telomere signal was normalized to the signal from the single-copy gene to generate a T/S ratio indicative of relative TL. The inter-assay coefficient of variation was 1.4%.

Data Analysis

Data were entered and verified using the SIR Database Management Software (SIR 2002, SIR Pty Ltd). IBM SPSS (Statistics 20, Armonk, NY) was used for analyses. Descriptive statistics (means and standard deviations) and distributions of all continuous and categorical measures were examined to identify key features (e.g., non-normal distribution, outliers, skewness) that might impact inferential methods. Demographic characteristics, age, education, and WAIS-III scores were examined using univariate and multivariate ANOVA to examine effects of group (healthy controls versus schizophrenia patients) and sex (male vs. female), and to test for interactions between the two.

Raw odor sensitivity scores were transformed to absolute values, such that higher scores represented more sensitivity to the olfactory environment. Testing of odor sensitivity was conducted for the right and left nostril. Preliminary analyses of our results indicated negligible differences between left and right nostril sensitivity; thus, the two were combined into a single mean score. Parental ages, odor sensitivity and LTL were examined using univariate ANOVAs to assess the main effects of diagnosis and sex. We examined the associations between parental ages, sensitivity, and LTL using Pearson correlation coefficients. Multiple regression procedures were used to examine the association between odor sensitivity and parental ages, covarying subjects’ age, diagnosis, and sex. As the analyses included a large number of exploratory tests, significance was set at α<0.01.

RESULTS

Age of the subjects did not differ by diagnosis or sex for the 75 patients and 46 controls. A significant interaction term between sex and group indicated that male patients were older than male controls, whereas female patients and controls had similar ages. Stratifying education by less than high school, high school, some college/trade school, BA/BS degree, or graduate degrees showed significantly less educational attainment for patients (X²=25.54, df=5, P<0.001). Patient and control groups had similar sex composition (X²=0.06, df=1, P<0.815), race/ethnicity ([White, Black, Hispanic, Asian/Pacific Islander, American Native]; X²=2.97, df=4, P=0.563), and marital status ([single/never married, divorced/separated, married]; X²=2.26, df=2, P=0.323). Patients were more likely to be “current” smokers (χ² P<0.0001). Sex was unrelated to smoking status in either group (χ²=1.72, df=2, P=0.423; χ²=1.78, df=2, P=0.412). Paternal and maternal ages, odor sensitivity, and LTL (Table I) were similar across groups and by sex. WAIS testing demonstrated significantly lower intelligence for patients on full scale IQ, verbal IQ, and performance IQ and on the verbal comprehension, processing speed, and working memory indices, although not on the Perceptual Organization Index. Male and female patients had similar ages of onset.

TABLE I.

Comparisons of Case and Control Group Characteristics

	Healthy controls (mean, SD)		Schizophrenia patients (mean, SD)		Statistics F test
	Males	Females	Males	Females	Diagnosis	Sex	Diagnosis by Sex
	n = 24	n = 22	n = 41	n = 34
Age	32.3 (9.3)	36.8 (13.1)	39.2 (9.5)	35.8 (10.1)	2.32	0.09	4.15*
Age of onset			21.5 (6.7)	22.1 (6.8)		t = 0.37
Odor sensitivity	n = 24	n = 22	n = 41	n = 34
	4.51 (1.4)	4.64 (.96)	3.93 (1.3)	4.67 (2.0)	1.00	2.45	1.18
Paternal age	n = 21	n = 21	n = 38	n = 30
	30.9 (7.6)	31.1 (6.8)	32.0 (8.5)	32.8 (8.8)	0.79	0.09	0.04
Maternal age	n = 23	n = 22	n = 36	n = 31
	28.8 (6.9)	28.2 (5.9)	27.7 (7.4)	28.1 (6.8)	0.20	0.01	0.13
WAIS III (n’s)a	n = 24	n = 20	n = 38	n = 30
Full scale IQ	106.0 (12.3)	100.2 (12.5)	89.0 (19.4)	92.7 (15.4)	15.59***	0.12	2.37
			Multivariate Wilks’ λ →		7.93**	0.10	2.69
Verbal IQ	109.4 (12.2)	102.6 (14.0)	90.7 (18.4)	96.4 (17.2)	15.68***	0.03	3.99*
Performance IQ	100.6 (13.8)	97.6 (13.6)	88.9 (19.3)	89.5 (13.4)	10.43**	0.17	0.35
			Multivariate Wilks’ λ →		17.42***	3.70**	2.82*
Verbal comp.	109.8 (11.8)	106.3 (15.3)	92.3 (18.5)	101.2 (18.2)	12.02**	0.68	3.64
Perceptual org.	98.9 (14.3)	96.7 (14.5)	92.0 (19.1)	89.4 (14.0)	5.22*	0.60	0.00
Processing speed	104.6 (13.5)	102.6 (13.0)	83.2 (13.1)	87.6 (12.1)	52.57***	0.20	1.63
Working memory	109.1 (15.5)	97.9 (13.4)	88.8 (17.3)	89.9 (14.4)	21.86***	2.85	4.20*
Leukocyte telomere length	n = 10	n = 6	n = 19	n = 16
	1.94 (0.79)	1.58 (0.59)	2.17 (0.97)	1.77 (0.72)	0.68	2.31	0.00

^*P < 0.050,

^**P < 0.010,

^***P < 0.001.

^aTest Statistic for cognitive scores is the ANCOVA controlling for age, age of onset, and education.

Odor Sensitivity

Odor sensitivity did not differ across diagnostic groups or sex. Levene’s test of equality of variances showed similar distributions of odor sensitivity for the groups (F [3,117]=2.09, P=0.105) (Fig. 1). Nonetheless, more patients than controls had odor sensitivity in the lower two quintiles (65% vs. 38%) whereas more controls than cases (62% vs. 35%) had odor sensitivities in the upper two quintiles (χ²=11.22, df=4, P=0.024), consistent with lesser acuity in a larger portion of patients.

FIG. 1.

Boxplot of odor acuity among controls and schizophrenia patients.

Odor Sensitivity and Subject Ages

Odor sensitivity and patient ages were negatively correlated (r=−0.317, df=73, P=0.006) with no significant association in controls (r=−0.238, df=44, P=0.111). Sex-specific analyses revealed that females in both groups showed declining odor sensitivity based on their age (patients: r=−0.453, df=32, P=0.007, controls: r=−0.451, df=20, P=0.035), which was not true for males (patients: r=−0.086, df=39, P=0.592, controls: r=−0.112, df=22, P=0.603).

Odor Sensitivity and Smoking

Mean odor sensitivity was examined across three smoking categories for controls and patients, with values as follows, respectively: for never smoked 4.55 (1.2) and 4.67 (1.9), former smoker 5.56 (1.1) and 4.52 (2.0), and current smokers 3.99 (1.0) and 3.93 (1.2). Current smokers in both groups had less sensitive odor detection, but adjustments for age and sex revealed that there were no significant effects of diagnosis (F [1/113]=0.87, P=0.422), smoking status (F [2/113]=1.25, P=0.291), or their interaction (F [2/113]=0.89, P=0.413) on odor sensitivity. Comparing odor sensitivity from the combined categories of “never smoked” and “former smokers” to “current smokers” and adjustments for age and sex, also showed odor sensitivity was unassociated with cigarette smoking in either group.

Odor Sensitivity and Parental Ages

Figures 2 and 3 display scatterplots of paternal and maternal age, respectively, with odor sensitivities in the schizophrenia and control groups with regression lines drawn for each group. Figure 2 shows that odor sensitivity significantly declined with their paternal age in controls (r=−0.40, P=0.010), but not in patients (r=0.11, P=0.380). These correlations of odor sensitivity to paternal age were significantly different for patients and controls (Fisher’s R to Z comparison: χ²=6.95, P=0.009). Sex-stratified analyses demonstrated that increasing paternal age was particularly associated with declining odor sensitivity in male controls (r=−0.54, P=0.012), without significance in healthy females (r=−.25, P=0.271). The association of odor sensitivity with paternal age was flat for both male (r=0.13) and female patients (r=0.08).

FIG. 2.

Odor acuity and maternal age in healthy controls (r=−0.46, n=45, P=0.001) and schizophrenia patients (r=−0.02, n=67, P=0.852).

FIG. 3.

Odor acuity and paternal age in healthy controls (r=−0.40, n=42, P=0.010) and schizophrenia patients (r=−0.11, n=68, P=0.380).

Figure 3 shows that advancing maternal age also predicted declining odor sensitivity in healthy controls (r=0.46, P=0.001), with odor sensitivity being unrelated to maternal age in the patients (r=−0.02, P=0.852). The correlations were significantly different for the patients and controls: Fisher’s R to Z: χ²=5.78, P=0.017. Sex-specific analyses showed that maternal age was particularly related to decreasing odor sensitivity in healthy males (r=−0.66, P=0.001) without a significant relationship in healthy females (r=−0.12, P=0.592). Both male and female schizophrenia groups showed “flat” associations between maternal age and odor sensitivity (males: r=−0.03, females r=−0.04).

As paternal and maternal ages both influenced odor sensitivity in healthy sons, we examined the correlation between maternal age and paternal age for the healthy male subjects, finding them to be highly associated (r 0.878, n 20, P<0.001). Independent effects of maternal versus paternal ages could not be further disentangled, but slightly stronger effects, as above, were observed for maternal than for paternal age to negatively influence odor sensitivity in sons.

Multivariate Evaluations of Parental Ages and Odor Sensitivity

Finally, regression analyses were undertaken with “odor sensitivity” as the dependent measures to probe if the relevant factors were independent and if parental age effects varied between the control and patient groups. Regression analyses were separately conducted for maternal and paternal age, considering effects of sex, age, diagnosis, the other parents’ age, and the interaction of diagnosis with the parental age variable (Table II). Better odor sensitivity was associated with younger age, being a healthy subject, younger maternal age and being female. Controlling for subjects’ age and sex and for the other parents’ age showed marginal effects that were consistent with an interaction between diagnosis and paternal age on odor sensitivity, supporting the observed different associations of paternal age with odor sensitivity in patients and controls. Similar effects were consistent with the different associations of maternal age with odor sensitivity in patients and controls. Although underpowered for this full regression analysis, the continuing trend significance for each variable in the models suggested their independence.

TABLE II.

Regression Results: Odor Sensitivity as the Dependent Measure and Subjects Age, Sex, Diagnosis and Paternal Age (above) and Maternal Age (below) as the Independent Covariates

	Standardized Coefficients	t	Sig.	95.0% Confidence Interval for B
Paternal Age Model	Beta			Lower Bound	Upper Bound
(Constant)		4.092	.000	4.455	12.832
Subjects Age	−.273	−2.973	.004	−.063	−.013
Diagnosis	−.687	−1.755	.082	−4.424	.270
Paternal Age	−.659	−1.886	.062	−.249	.006
Sex	.201	2.221	.029	.063	1.120
Interaction Paternal Age and Diagnosis	.992	1.849	.067	−.005	.140
	Standardized Coefficients	t	Sig.	95.0% Confidence Interval for B
Maternal Age Model	Beta			Lower Bound	Upper Bound
(Constant)		4.939	.000	6.319	14.794
Subjects Age	−.328	−3.716	.000	−.072	−.022
Diagnosis	−.804	−2.093	.039	−4.817	−.131
Maternal Age	−.785	−2.516	.013	−.314	−.037
Sex	.182	2.102	.038	.031	1.071
Interaction Maternal Age and Diagnosis	.942	1.978	.051	.000	.161

Odor Sensitivity and Cognitive Function

Examining cognition and odor sensitivity (Table III) demonstrated that lesser odor sensitivity was significantly related to cognitive impairments in schizophrenia patients, with the exception of verbal comprehension. Conversely, odor sensitivity was not related to cognition in healthy controls. Controlling for current smoking status (not tabled) did not influence the association of odor sensitivity and cognition in any group.

TABLE III.

Correlations With Odor Sensitivity (Pearson’s r)

In schizophrenia patients	All n = 68	Males n = 38	Females n = 30
Full scale IQ	0.374**	0.426**	0.317
Performance IQ	0.397**	0.484**	0.345
Verbal IQ	0.318**	0.341*,a	0.216
Verbal comp. index	0.327**	0.321*	0.275
Perceptual org. index	0.415***	0.474**	0.444*
Working memory index	0.283*	0.359*	0.215
Processing speed index	0.295*	0.253	0.299
In healthy controls	All n = 44	Males n = 24	Females n = 20
Full scale IQ	−0.035	−0.084	0.082
Performance IQ	0.013	0.038	−0.009
Verbal IQ	−0.073	−0.199a	0.142
Verbal comp. index	−0.049	−0.267	0.247
Perceptual org. index	0.026	0.071	−0.032
Working memory index	−0.094	−0.140	0.048
Processing speed index	−0.003	0.101	−0.169

^*P < 0.050,

^**P < 0.010,

^***P < 0.001.

^aCorrelations differed between patients and controls.

Odor Sensitivity and Telomere Length

As shown in Table I, mean LTL did not differ between patients and controls. LTL was also not associated with odor sensitivity in all patients (−0.167, n=35), or in subgroups of male (0.235, n=19) or female patients (0.130, n=16). Likewise, LTL was unassociated with odor sensitivity in all controls (−0.110, n=16), or in male (−0.030, n=10) or female control subgroups (−0.206, n=6). LTL was furthermore not associated with parental age in controls (−0.366) or patients (0.036), and not in the subgroups of male controls and patients (−0.341; 0.277) or female controls and patients (−0.703; −0.376). Maternal age also was uncorrelated with LTL in these controls (−0.106) and patients (0.136), and in respective subgroups of males (−0.243; 0.290) or females (−0.022; 0.071). Multivariate regression analyses examining the predictors of LTL revealed no significant associations in healthy control and schizophrenia groups, respectively, to age (t=−0.364, P=0.724; t=0.434, P=0.667), sex (t=−0.654, P=0.528; t=−0.851, P=0.401), or odor sensitivity (t=−1.13, P=0.286; t=1.19, P=0.243).

DISCUSSION

Although these results are preliminary and must be replicated in larger samples, they demonstrate for the first time that odor sensitivity decreases in association with parental aging in healthy subjects, for both maternal and paternal age, particularly in sons. By contrast, odor sensitivity did not decline with parental aging in patients with schizophrenia. Furthermore, in schizophrenia patients, but not in controls, lesser sensitivity to odor detection was robustly associated with cognitive impairments. Contrary to predictions, telomere length measured as LTL, did not explain any of these associations.

These findings suggest that different pathways may be involved in some of the schizophrenia susceptibility related to being an offspring of an older parent, as this effect did not entail diminished odor sensitivity and impaired cognition. The confluence of lesser odor sensitivity and low-intelligence scores in other cases implicates shared neural substrates between cognition and olfaction for an important etiology for schizophrenia. Structural deficits in orbitofrontal regions in schizophrenia are a possible shared substrate, as this region is related to cognitive function in schizophrenia [Schobel et al., 2009] and is an olfactory processing region [Barbas, 2007]. The lack of association between odor sensitivity and verbal comprehension in these schizophrenia patients, unlike all other IQ scores and indices, suggests that any shared pathophysiology impacting cognition and odor sensitivity begins after verbal skills are established, perhaps in the prodromal period or early illness phase, or that the pathology involves separate brain regions from those involved in verbal comprehension. The only cognitive task that did not show group differences between patient and control groups, the Perceptual Organization Index, actually showed the most robust associations with odor sensitivity. The association of this task’s performance and odor sensitivity in schizophrenia patients is also consistent with an active process, as ultra-high risk subjects who are not yet psychotic did not show perceptual organization problems in one large study [Silverstein et al., 2006].

This work supports the heterogeneity of the origins of the disease. A factor that does not diminish sensitivity to the environment or produce cognitive deficits may be an etiology for psychosis related to later parental ages. Further studies can examine if odor sensitivity is a biomarker for environmental sensitivity in general, given its ease at administration. Impaired sensitivity to the environment is demonstrated in schizophrenia [Brown et al., 2002b] and autism [Robertson and Simmons, 2013] so such a measure could be of importance in research and treatment studies.

Although LTL did not explain any of these relationships, other environmental circumstances or epigenetic mechanisms could be involved. Parental ages could be only indirectly related to odor sensitivity and future studies should examine birth order and family size as well. Humans are highly variable in odor sensitivity. Disruptions in the early environment that increase schizophrenia vulnerability, such as trauma or inflammation, could reduce odor sensitivity. In mice, postnatal experiences modulate lifelong olfactory receptor sensitivity [He et al., 2012], and such effects could even be heritable. Odor fear conditioning in naive male mice produced two subsequent generations of male offspring with increased sensitivity to the same fear odor without prior exposure [Dias and Ressier, 2014]. Similar results were found in male offspring of female mice who were conditioned prior to conception. Parts of the olfactory bulb increase in size when exposed to the conditioned odor (compared to control offspring), suggesting neuroanatomical factors in modulating odor sensitivity [Dias and Ressier, 2014; Szyf, 2014]. Although the sensitivity for specific odors is associated with haplotypes for respective olfactory receptor genes, a portion of the healthy population also has general hyperosmia [Menashe et al., 2007]. If the offspring of younger parents have enhanced sensitivity in larger studies, then these data may shed light on a portion of the “general factor” in olfactory perception, which is reported to account for the greatest amount of interindividual variability in human odor sensitivity [Menashe et al., 2007]. Much remains to be learned about the modulation of odor sensitivity and the suggestion of its association with the risk for schizophrenia is intriguing.

Together with our accompanying paper in this volume by Lehrer et al., demonstrating the association of paternal aging with psychosis in bipolar disorder, this preliminary study supports a paternal age-related psychosis mechanism that does not diminish odor sensitivity or produce cognitive deficits. Certain de novo mutations or epigenetic mechanisms could influence the phenotype and be related to parental ages. Epigenetic mechanisms can modulate gene expression based on the physical and social environments and parental aging may act as another such “environment.” Parent-of-origin imprinting is an intergenerational epigenetic mechanism that results in the monoallelic expression of only the maternal or paternal allele in offspring, depending on the gene and tissue [Badyaev and Uller, 2009]. Some genomic imprinting is based on both the sex of the offspring and the transmitting parent [Gregg et al., 2010], consistent with the sex and parental sex differences in odor sensitivity we observed. Even though only a hundred genes with monoallelic expression based on parent of origin are fully defined [Jirtle, 2012], some work shows that hundreds of genes may be imprinted. These are likely to be particularly relevant to brain regions related to olfactory processing, including feeding, motivational behavior, and the hypothalamus [Glenn et al., 1997].

Although this is a small sample and a preliminary study, robust group differences were demonstrated in the principal aims of the study. Group differences in education, cigarette smoking and cognition (P<0.001) were demonstrated between the case and control groups, as were differences in the associations of parental aging with odor sensitivity (P<0.009). Likewise, robust in the cases only (P<0.001) was the relationship of odor sensitivity to perceptual organization scale of the WAIS, with other cognitive measures being associated at P<0.01. By contrast, no associations of LTL with any measures were demonstrated at any level of significance, although LTL had been a promising mechanism to explore in light of its association with paternal age in schizophrenia [Malaspina et al., 2014]. This study may be underpowered to detect possible associations between LTL and these measures, so larger studies are indicated to absolutely exclude these associations. Another important limitation of this study is the use of only PEA to assess olfactory thresholds. Although this test odorant is a reliable measure of odor sensitivity and widely used for this purpose [Doty et al., 1995; Lotsch et al., 2004], future studies should employ an array of odorants, and on multiple occasions. All patients were taking stable doses of prescribed medications at the time of the study although no association has been demonstrated between odor sensitivity and medication in schizophrenia patients [Badyaev and Uller, 2009].

CONCLUSIONS

The novel finding in this study was that parental aging was unexpectedly associated with relatively intact odor sensitivity in schizophrenia patients whereas diminishing acuity accompanied parental aging in the controls. Lessened sensitivity, by comparison, was only related to the cognitive deficits in the patients. If confirmed, the susceptibility to cognitive impairment and decreased odor sensitivity may reflect a different risk pathway than the one associated with parental aging.