Associative Memory and Memory Complaints in People with First Episode of Depression: Use of the Face-Name Associative Memory Exam (FNAME)

1. Introduction

Depression is a common disorder worldwide, affecting 280 million people [1], and it is among the top ten causes of disability in the world [2]. It is presented as a set of predominantly affective symptoms impacting on people's functional status [3,4].

Importantly, major depressive disorder (MDD) can lead to cognitive dysfunction [5,6,7,8,9,10,11,12,13]. Memory complaints are amongst the most frequent complaints [12] and there is evidence that patients with depression have lower memory performance compared to healthy controls [6,14,15,16].

Research on memory complaints (i.e., the subjective perception of memory decline) in depression has gained significant attention recently [17,18,19]. Studies conducted in the general population found a relationship between depression and memory complaints [20,21,22,23,24,25]. Also, other studies with young adults show that depression is considered the only variable that explains memory complaints [26,27]. Furthermore, it has been shown that people with depression and severe anxiety report more memory complaints than people diagnosed with mild cognitive impairment (MCI) [28,29].

Another important aspect is that people's beliefs about their own memory abilities, known as memory self-efficacy, can influence how they actually use their memory in everyday situations [30,31]. Strongly negative beliefs about memory performance may have an important implication for cognitive biases [32,33], with a decrease in self-efficacy judgements observed in depression [34,35,36,37].

The relationship between memory complaints and objective memory has also sparked interest. Some studies showed that people with depression had a greater number of memory complaints and poorer performance on objective memory tests [38,39,40] showed that people with depression presented more cognitive complaints in daily life than in neuropsychological tests, probably due to the associated negative self-perception. In addition, a low correlation between objective and subjective assessment of memory has been observed in people with more severe depression [41].

In a recent review [42] concluded that the number of depressive episodes experienced might determine the type of cognitive impairment observed. While first episode of depression (FED) subjects performed significantly worse than controls on processing speed and executive/working memory, patients with recurrent depression performed significantly worse, with verbal learning and memory being the most impaired domain. Another recent review [43], assessing neurocognitive deficits in depression in the acute and remitted state, concluded that in the acute phase, there is strong support for impairment in processing speed, learning, and memory. Follow-up studies and direct comparisons revealed less pronounced deficits in remission, although deficits were still present in attention, learning and memory. A positive correlation between the number of episodes and cognitive deficits as well as depression severity and cognitive deficits was also reported.

In recent years, associative memory impairment has attracted interest in the field of neuropsychology [44,45,46,47,48] as a sensitive marker of cognitive decline. The focus of this study was to assess episodic memory using an ecologically valid and demanding test, namely, the adapted Face Name Associative Memory Exam (FNAME) [49,50,51,52,53,54,55,56,57] to detect differences in memory performance in people with FED.

In this study we use the adapted FNAME [54] because it can evaluate associative memory in a more ecological way. Furthermore, it is important to quantify memory complaints to obtain both subjective and objective measures of memory functioning in FED patients. Finally, MoCA was used as a reliable measure of general cognitive function that is sensitive to early changes in cognitive ability across domains but robust to depression symptoms within healthy cohorts [58].

Our prediction was that FED patients would present more memory complaints and lower memory performance than healthy participants.

2. Materials and Methods

2.1. Study Design

Analytical observational design. The primary study variables were FED and associative memory. Other secondary variables were memory complaints, global cognitive status, and symptoms and intensity of depression.

2.2. Participants

In this study, 30 participants were recruited and 15 of them met the criteria for FED. This group was recruited from the Mental Health Unit of the Psychiatric Hospital in Palma and the Mental Health Unit of the Regional Hospital of Inca. The other 15 participants were healthy controls (HCtrl) recruited from the general population. The recruitment lasted between March 2021 to October 2022.

FED participants met the following inclusion criteria: 1) they were between 18 and 65 years old and, 2) they had a diagnosis of FED fulfilling DSM-5^® criteria. Exclusion criteria for both FED and HCtrl participants included: 1) substance abuse, such as alcohol (>24 g/day in women, 40 g/day in men) and other drugs, 2) people with other mental health or neurological disorders. HCtrl participants were volunteers with similar sociodemographic characteristics to FED participants. They reported no neurological or psychological disorders, and they did not take psychotropic drugs.

2.3. Material and Procedure

This study was an analytical, cross-sectional, prospective, observational design. Primary care physicians and/or psychiatrists from the different health centers in Palma initially assessed the participants considering the inclusion criteria. FED participants were then asked to voluntarily participate in the study and referred to the researchers to take a complete medical history and confirm the inclusion criteria. The approximate duration of each clinical and neuropsychological evaluation session was 50 minutes. On visit 1 (day 1), the psychiatrist explained the protocol to the participants and gave them the information sheet and the informed consent. The selection criteria were reviewed, and the clinical interview was conducted using the International Neuropsychiatric Interview (MINI)[59] to exclude other comorbid psychiatric disorders. Additionally, the interviewer-administered Montgomery-Asberg Depression Scale (MADRS)[60,61] was used to assess depressive severity. On visit 2 (day 7), the psychologist collected the sociodemographic data and used the Self-Reported Rapid Depressive Symptom Scale (QIDS-SR16)[62,63] to assess depressive severity of the last week.

The extended and modified Face-Name Associative Memory Exam (FNAME, Rentz et al., 2011) was then administered as shown in Figure 1 [54,64].

The administration of the FNAME was carried out in 12 steps: 1) Familiarization: participants underwent a trial where they only looked at faces, 2) Learning I: participants underwent the trial to look at the 12 face-name pairs, 3) Immediate recall I: they were asked to provide the name associated with each face, 4) Learning II: they looked at the face-name pairs they previously associated incorrectly, 5) Immediate recall II: again they were asked to provide the name associated with each face, 6) Re-learning: participants were presented again with the face-name pairs they did not remember, 7) Thirty minutes delay, 8) Spontaneous name recall: participants were asked to freely recall all the names they had learned in 2 minutes, 9) Face recognition: participants performed a face recognition and matching test, 10) Delayed recall: participants were asked again to say aloud the name associated with each face, 11) Name recognition: participants were asked to select the name associated with the face from among four. Finally, 12) Matching, where participants were asked to match the correct name with the corresponding face. Participants responded verbally and the responses were recorded on a scoring sheet.

The daily life memory failures questionnaire (MFE-30)[65] was administered during the 30-minutes delay within the administration of the FNAME. We used the version of the test validated in the Spanish population [66], which consists of 30 questions about daily situations and activities. Participants answered these using a Likert scale from 1 (never or almost never) to 5 (always or almost always), with a maximum score of 150.

The Montreal Cognitive Assessment test (MoCA)[67], a 30-items screening test that evaluates global cognitive status through different cognitive domains (memory, visuospatial ability, executive function, attention, concentration, working memory, language and orientation) was administered to observe whether participants had impaired global cognitive function. A cut-off score of 26 or higher out of 30 indicates normal cognitive state. One point was added to this score if the participant had 12 years or less of formal education when the total score was below 30.

2.4. Statistical Analysis

Statistical analyses were performed using IBM SPSS Statistics 20.1. Normality and homogeneity of variances were verified with the Shapiro-Wilk test and the Levene’s test, respectively. Descriptive statistics were carried out for the different sociodemographic, clinical, and cognitive variables of the FED and HCtrl groups, with means and standard deviation, and inferential statistics carried out by means of the Mann-Whitney (U) test and mixed model ANCOVA. Effect sizes were measured using rrb, rank biserial correlation, and Eta squared (η²). Moreover, correlations were calculated with Spearman's Rho (r) with the total sample (with p-value adjustments).

2.5. Ethical Aspects and Data Security

This study was conducted in compliance with the principles of the Declaration of Helsinki [68]. The processing, communication and transfer of personal data was in accordance with the provisions of Organic Law 3/2018, on the 5 December, on the protection of personal data and guarantee of digital rights. This study was approved by the Research Committee of the Son Espases University Hospital (HUSE) with dictum number CI-481-20 (approval date: 14^th January 2021) and the Research Ethics Committee of the Balearic Islands (CEI-IB) with dictum number IB 4429/21 PI (approval date: 24^th February 202). All participants signed the informed consent.

3. Results

Patients with FED obtained scores indicating that FED patients had moderate depression with MADRS = 26.63 (5.54) points and QIDS-SR16 = 15.3 (5.05) points.

Socio-demographic data and results from the different general cognitive tests are shown in Table 1.

There were more women than men in the FED group, but this was not the case in the control group. There was no significant difference between groups for age (U(29) = 66.5, p = .06), but a significant difference between groups was observed for years of education (U(29) = 19, p < .01).

Given the significant difference between groups at the level of years of education, the comparison of groups for the cognitive measures was carried out using ANCOVAs controlling this variable as a covariate. Results showed significant differences between groups for the MFE [F (1, 29) = .06, MSE = 1690.34, p = .05, η² = 0.14], but not for the MoCA [F (1, 29) = .08, MSE = .06, p =.78, η² = 0.01].

Table 2 presents the results of the different FNAME subtests.

The ANCOVA controlling for the effect of education revealed no significant effect of group on any of the FNAME subtests: Immediate Recall I [F (1,29) = .14, MSE = .89, p = .71, η² = 0.01], Immediate Recall II [F (1,29) = .01 , MSE = .04, p = .93, η² = 0.01], total immediate recall [F (1,29) = .03 , MSE = .54 , p = .87 , η² = 0.01], Spontaneous Name Recall [F (1,29) = .58, MSE = 2.86, p = .45, η² = 0.02], Face Recognition [F (1,29) = .003, MSE = .001, p = .96, η² = 0.01], Delayed Recall [F (1,29) = .26, MSE = 1.79, p = .61, η² = 0.01], Name Recognition [F (1,29) = .47, MSE = .06, p = .50, η² = 0.02], Matching [F (1,29) = .54, MSE = 2.74 , p = .47, η² = 0.02], and the total score [F (1,29) = .03, MSE = 3.60, p = .86, η² = 0.01].

Recall means (subtest) (mixed model ANCOVA)

The immediate and delayed recall subtests were analysed using a 2 (groups) x 3 ANCOVA (subtest: immediate recall I, recall immediate II, delayed recall) controlling for education as a covariate. In this analysis, no significant group effect was observed [F (1, 26) = .09, MSE = 1.43, p = .76, η² = 0.003]. Therefore, the HCtrl group did not recall significantly more face-name associations than the FED group (see Table 3). For the type of recall subtest, a significant effect was not observed either [F (1, 26) = 1.90, MSE = 8.64, p = .17, η² = 0.13], and the interaction group by recall type (F (1,26) = .35, MSE = 1.29, p = .70, η² = .13) was not significant.

Correlations between FNAME, MFE-30 and MoCA of all participants (FED and healthy controls) are presented in Table 3.

The score obtained on the subjective memory complaints questionnaire (MFE-30) did not significantly correlate with the FNAME subtests (p > .05). The score obtained from the MoCA was significantly correlated (p < .01) with the different FNAME subtests (total immediate recall, spontaneous name recall, delayed recall, association, total score). The total immediate recall measure and the rest of the FNAME subtests were significantly correlated with each other (p < .01).

4. Discussion

The main aim of this study was to assess possible associative episodic memory deficits and subjective memory complaints in FED by using the FNAME and the MFE-30, respectively.

Major depressive disorder (MDD) is a heterogeneous mental disorder that affects people throughout his life [70,71,72]. This mental disorder affects neurocognitive functions, including processing speed, attention and executive functions [73]. It is also important to explore associative memory, scarcely investigated in individuals with depression, especially FED.

The results did not show statistically significant differences between groups in relation to their performance on the different FNAME subtests. This result is aligned with the observation made by Alegret et al. (2015b), that worse performance on the S-FNAME was not related to depression symptoms. Additionally, other studies that have assessed episodic memory in FED patients found no differences in performance compared to healthy controls’s [74,75,76,77,78]. Therefore, these findings do not support the hypothesis that patients with FED have memory problems due to poor learning [5,7,11,12], consolidation [79,80] or recall [81,82,83].

Deficits in associative memory have been investigated in other clinical populations such as Alzheimer's disease [84,85] and Parkinson's disease [64,86,87], and in non-clinical populations such as ageing [53,54]. In the present study, associative memory performance in FED was analysed using supposedly more sensitive and cognitively demanding subtests, such as matching, which was shown to be sensitive to Parkinson's disease [64]. However, our results showed no significant differences between FED patients and healthy controls, even though FED participants reported a higher number of complaints than control participants in daily life.

These results support the findings from previous studies [22,24,25,28,38,88], showing that although patients with MDD had relatively preserved memory tests they under-estimated their memory functioning, a pattern distinct from HCtrl.

In our study, subjective memory complaints did not show a statistically significant correlation either with the different subtests of the FNAME. This lack of concordance between objective cognitive deficits and subjective memory complaints has also been shown in previous studies [72,83,89,90,91,92,93] in patients with heterogeneous referral depression diagnoses (severity, onset time, number of episodes, etc.).

Therefore, our results support the notion that patients with FED present a dissociation between objective memory (FNAME) and subjective memory (MFE-30). Beliefs about memory self-efficacy as an estimate of actual memory abilities may have an influence on the way people use their memory in everyday situations [30,31]. In this sense, when beliefs about memory performance are negative [32,33,94], this may result in a sort of self-fulfilling prophecy and, consequently, a confirmation bias. Therefore, our results suggest that FED patients may present a negative cognitive predisposition for their memory performance, which may affect subjective memory leading to a greater number of memory complaints in daily life [32,33] without objective memory being really affected. More specifically, self-efficacy moderates the relationship between self-rated memory function and depressive symptoms. Higher self-efficacy may buffer against the impact of subjective memory difficulty on one's mood and thereby mitigating the effect of depressive symptoms on memory [95,96].

The MoCA score was significantly correlated with the scores on all FNAME subtests. A stronger correlation was observed with matching, delayed recall, and spontaneous name recall, which have proven to be most sensitive to cognitive decline [64]. Noteworthy is that although FED participants and controls did not differ significantly on the MoCA score, FED presented on average a score below the cut-off score of 26 points [97,98]. This may be because the MoCA test evaluates attention and executive functions, while the FNAME evaluates associative episodic memory.

It is also important to refer to the limitations of this study. Firstly, we encountered the difficulty of recruiting patients with depression who met the inclusion criteria. In this sense, a relatively small sample was obtained that offered initial results and an approach to the objective of the study that requires broader research. Therefore, statistical analyses were carried out with a non-parametric approach. This is a pilot study that suggests new clues, but future research with larger samples should be carried out to obtain more consistent results. And secondly, this research began to be carried out during the COVID-19 pandemic, which led to significant fluctuations (non-recruitment periods) with participants. This is a pilot study that suggests new indications, but future research with larger samples should be carried out to obtain more consistent results. The difficulty in recruitment also determined that a 'one-to-one' match was not carried out between the participants in the experimental and control groups in relation to age, gender, and educational level. This gave rise to differences in educational level between participants in both groups, so it was controlled as a covariate. In addition, the control participants performed adequately on the cognitive tests, but a validated test was not used to evaluate mood; information was only collected through the participant's form about their general health status (they reported that they were active in the labour market, and self-employed). Therefore, in future research this would also be an aspect to consider.

Even considering the above limitations, the present study is relevant because: 1) FNAME is a more cognitively demanding test using sensitive measures in the assessment of associative memory, 2) patients with depression that underestimation in memory performance would be a premorbid state of functioning not identified in objective cognitive assessments, and 3) this assessment could have an important implication in the treatment of patients with a negative thought pattern. This could be incorporated for a section in cognitive restructuring therapy.

5. Conclusions

Although FED patients and healthy controls may present similar associative memory performance, FED tend to overestimate their memory difficulties, showing a dissociation between objective (FNAME) and subjective (MFE-30) memory.