Lipid Metabolism and Survival Across the Frontotemporal Dementia-Amyotrophic Lateral Sclerosis Spectrum: Relationships to Eating Behavior and Cognition

Abstract

Background:

Patients with frontotemporal dementia (FTD) and amyotrophic lateral sclerosis (ALS) exhibit changes in eating behavior that could potentially affect lipid levels.

Objective:

This study aimed to document changes in lipid metabolism across the ALS-FTD spectrum to identify potential relationships to eating behavior (including fat intake), cognitive change, body mass index (BMI), and effect on survival.

Methods:

One hundred and twenty-eight participants were recruited: 37 ALS patients, 15 ALS patients with cognitive and behavioral change (ALS-Plus), 13 ALS-FTD, 31 behavioral variant FTD, and 32 healthy controls. Fasting total cholesterol, low density lipoprotein cholesterol (LDL), high density lipoprotein cholesterol (HDL) and triglyceride levels were measured and correlated to eating behavior (caloric, fat intake), cognitive change, and BMI; effect on survival was examined using cox regression analyses.

Results:

There was a spectrum of lipid changes from ALS to FTD with increased triglyceride (p < 0.001), total cholesterol/HDL ratio (p < 0.001), and lower HDL levels (p = 0.001) in all patient groups compared to controls. While there was no increase in total cholesterol levels, a higher cholesterol level was found to correlate with 3.25 times improved survival (p = 0.008). Triglyceride and HDL cholesterol levels correlated to fat intake, BMI, and measures of cognition and disease duration.

Conclusion:

A spectrum of changes in lipid metabolism has been identified in ALS-FTD, with total cholesterol levels found to potentially impact on survival. These changes were mediated by changes in fat intake, and BMI, and may also be mediated by the neurodegenerative process, offering the potential to modify these factors to slow disease progression and improve survival.

Keywords

Amyotrophic lateral sclerosis cholesterol eating frontotemporal dementia hypothalamus metabolism,neurodegeneration neuroendocrine

INTRODUCTION

The significance of disturbances in lipid levelsand metabolism in frontotemporal dementia (FTD) and amyotrophic lateral sclerosis (ALS) remains a source of debate, specifically whether hyperlipi-demia may exert an effect on disease incidence, progression, and survival, and whether changes are related to cognitive changes and changes in eating behavior.

Epidemiological studies have shown that treatment with statins to lower cholesterol levels is associated with an increased incidence of ALS [1, 2]. Unlike other hyperlipidemic populations, ALS patients are typically characterized by low body mass index (BMI), complicated by weight loss with disease progression [3], making the concept of increased lipid levels in ALS perhaps counterintuitive. However, in a French ALS cohort, two thirds of patients were found to have increased low density lipoprotein cholesterol (LDL), decreased high density lipoprotein cholesterol (HDL) concentration or a combination of the two [4]. In a separate German cohort, elevated triglyceride and total cholesterol levels were associated with a positive effect on survival [5]. Other studies have suggested that increased cholesterol levels may be associated with slower functional decline and increased survival in ALS [6]. Not all studies have replicated these findings, with several studies suggesting that dyslipidemia does not occur in ALS [7 –9], and is not associated with a benefit to survival [8], while in other studies, although patients were not dyslipidemic, having a higher HDL/LDL ratio was correlated with improved survival [8, 9] These studies suggest variations in lipid levels occur in ALS cohorts and that determining the correlates may provide important healthoutcomes.

One factor yet to be studied is whether lipid metabolism in ALS may be associated with behavioral changes in eating, as has been described for FTD [10]. In FTD, changes in triglyceride and HDL levels have been reported and related to abnormal eating behavior and changes in BMI [11], with a complex neural network controlling eating behavior that include the hypothalamus, autonomic, and reward pathways [12]. At present, it remains unknown if these changes also exist along the ALS-FTD spectrum, and whether lipid levels are related to eating behaviors, cognitive change, caloric intake, and BMI. As such, the present study aimed to investigate the relationship between lipid metabolism, BMI, and eating behavior across the ALS-FTD spectrum, and whether any potential changes in lipid levels may be associated with differences insurvival.

METHODS

All patients were recruited from the ForeFront clinics, Sydney, Australia. ALS patients with enteral feeding via PEG tube, or where a carer was not available, were excluded from the study. Carers completed all surveys at a single visit, and at this visit cognitive measures, fasting lipid levels, and BMI were measured. Length of survival time was the total number of years from estimated symptom onset by carer until death or last follow-up.

Standard protocols and approvals and role of funding source

This study was approved by the University of New South Wales human ethics committees. Written informed consent was obtained.

All patients met current clinical diagnostic criteria for ALS [13], ALS-FTD [14], and bvFTD [15]. Presence of abnormalities in the C9orf72, SOD1, TDP43, FUS, GRN, and Tau genes was examined in all patients. Motor function was assessed using the ALS Functional Rating Scale (ALSFRS-R) [16] and patients were subclassified as limb or bulbar predominant based on their initial presentation. Patients with a diagnosis of ALS-plus (ALS-P) had additional behavioral or cognitive features that did not meet criteria for ALS-FTD, and diagnosis was based on previous validated criteria [17 –19]. Specifically, the presence of cognitive features was demonstrated by showing abnormalities on two validated tasks of executive function, with patients scoring below the fifth percentile. The following tests were used: excluded letter fluency, the Hayling test, letter fluency (P), animal fluency, and Trails (time B-A) [18]. The presence of behavioral features in the ALS-P group was demonstrated by impairment in at least two non-overlapping behaviors assessed by carer questionnaire (Cambridge Behavioral Inventory-CBI) [18, 20], and corroborated with caregiver interview. The CBI has been validated as sensitive to behavioral changes in ALS [21, 22].

Healthy controls were recruited from a panel of volunteers and were age- and education-matched and scored above 88/100 on the Addenbrooke’s Cognitive Examination-Revised (ACE-R) [23].

Assessment of physical measurements

Height and weight were measured (shoes removed). Body mass index (BMI) was calculated: weight (kg)/height (m²).

Concomitant diseases and medications

Patient records from their local medical officer were obtained to ascertain presence of hypercholesterolemia and medications.

Blood measurements

Blood samples were obtained from all participants following a 10-h fast.

Measurement of cholesterol levels

Total cholesterol, triglyceride, and HDL levels were measured using the enzymatic colorimetric method using Cobas 8000, supplied by Roche diagnostics, Indianapolis, USA. LDL levels were calculated using the Friedewald formula: LDL =[Cholesterol –(Triglyceride/2.2)] –HDL

Assessment of eating behavior and food intake

The APEHQ comprises 34 questions examining changes in eating behaviors over the preceding 6 months (swallowing, appetite, eating habits, food preference, and other oral behaviors). Carers rated frequency on a 5-point Likert scale, ranging from 0 (never) to 4 (daily or continuously) and severity on a 4-point Likert scale, ranging from 0 (not applicable) to 3 (marked) for each behavior. A composite frequency × severity score was calculated for each question, as well as an overall score [24]. A normal control score on this questionnaire is considered as 0. The four questions from the Cambridge Behavioral Inventory (CBI) that are specifically related to eating behavior were also analyzed [24].

Assessment of daily food intake

Information on overall caloric intake, macronutrient composition, and food preferences was obtained using the Dietary Questionnaire for Epidemiologi-cal Studies (DQES) (http://www.cancervic.org.au/about-our-research/cancer-statistics/nutritional_assessment_services), a questionnaire completed by carers. Output provides comprehensive information on food and drink intake (e.g., water, kilojoule, total fat, total protein, carbohydrates, and sugars).

Data analysis

Data were analyzed using IBM SPSS statistics (version 21.0) (p < 0.05 regarded as significant). Kolmogorov-Smirnov tests were run to determine suitability of variables for parametric analyses. To confirm diagnostic group membership, analyses of variance (ANOVA), followed by Tukey post hoc tests, were used to confirm cognitive (ACE-R-scores), behavioral (total CBI score) and motor (ALSFRS scores) age, disease duration, and survival. Differences in frequency patterns of categorical variables (e.g., sex) and subtype of ALS were examined with Chi squared tests.

Initial group analyses were conducted to determine whether lipid metabolism and measures of eating behavior differed between groups using Kruskal-Wallis tests followed by post hoc Mann Whitney tests corrected for multiple comparisons (p≤0.01 regarded as significant) for the following variables: CBI eating score, total product of frequency and severity of APEHQ, total cholesterol, LDL, HDL, cholesterol/HDL ratio, triglyceride levels, BMI values, and total caloric intake and macronutrient intake. To identify relationships to alterations in lipid levels, a number of tests were used. Spearman correlations were carried out between lipid levels and BMI, eating questionnaire scores, total caloric intake, fat intake, ACE-R and ALSFRS scores, and disease duration. Separate Cox regression modeling was used to assess the effect of total cholesterol and triglyceride levels on survival, covarying for age, cognitive status (ACE-R score), disease duration, and diagnosis. The effect of BMI and fat intake on survival was also examined using univariate analysis. In order to address the question whether changes in triglyceride levels were simply related to BMI or were affected by fat intake and diagnosis, the relationships between triglyceride, HDL, and total cholesterol with fat intake and BMI and diagnosis were further explored using multiple linear regression analyses, using theEnter method.

RESULTS

Assessment of group differences

Demographics and diagnostic variables

The study cohort consisted of sixty-five ALS patients (37 ALS, 15 ALS-P, and 13 ALS-FTD), 31 bvFTD, and 32 healthy controls. Patient demographics confirmed representative disease cohorts (Table 1) with ALS, ALS-P, and ALS-FTD patient groups exhibiting a shorter disease duration than the bvFTD patient group (Table 1, all p values<0.001), and C9orf72 repeat expansions concentrating in ALS-FTD (38%) but also in bvFTD (13%) and in ALS-P (1/15 or 7%) (Table 1). No other genetic abnormalities were observed in the cohorts. On the ACE-R, the ALS-FTD and bvFTD groups scored lower than the ALS and ALS-P patients and controls (Table 1, all p values <0.001). The ALS-P group also scored lower than the ALS (p = 0.049) and control groups (p = 0.030). On the CBI (total score) the ALS-FTD and bvFTD groups scored higher than the ALS and ALS-P groups (Table 1, all p values <0.001). No group differences were detected between the ALS groups in terms of predominant ALS subtype (bulbar versus limb) or the ALSFRS. Significant group differences were also noted for BMI with all ALS groups with cognitive impairment (ALS-P, ALS-FTD, and bvFTD) having increased BMI compared to controls, and ALS-FTD and BvFTD compared to ALS. No significant group differences were noted for statin usage (Chi-square = 1.8, p = 0.778) with 16% ALS, 30% ALS-P, 25% ALS-FTD, 28% bvFTD, and 29% of controls ontreatment.

Table 1

Demographic characteristics and cognitive scores for the ALS, FTD, and control groups

	ALS	ALS-P	ALS-FTD	bvFTD	controls	F value	Post hoc test
Sex (F:M)	11 : 26	4 : 11	3 : 10	9 : 22	12 : 20	1.2^#	NA
ALS onset	28 limb, 9 bulbar	9 limb, 6 bulbar	11 limb, 2 bulbar	N/A	NA	2.3^#	NA
Age (y)	55.9±11.4	60.8±11.8	65.1±6.8	63.4±9.2	64.7±15.1	3.2^*	ALS<controls
Disease duration (y)	1.7±1.1	2.3±1.9	3.3±2.5	6.2±2.9	NA	26.6^***	ALS, ALS-P, ALS-FTD<bvFTD
Number deceased (% of cohort)	7 (19%)	1 (6%)	4 (31%)	5 (16%)	NA	2.8^#	NA
Mean survival time of deceased patients from symptom onset	2.6±2.0	1.2	2.8±1.9	9.8±3.4	NA	15.8^***	ALS, ALS-FTD<bvFTD
C9orf72 expansion positive	0	1	5	4	NA
ALS-FRS	38.5±7.2	40.9±5.6	41.8±5.9	NA	NA	1.4	NA
ACE-R total (100)	94.1±3.1	88.0±9.8	74.8±9.8	79.9±10.3	94.6±3.6	33.4^***	ALS, ALS-P, controls>ALS-FTD, bvFTD; ALS, controls>ALS-P
CBI-Total	17±10.0	35.3±26.0	63.6±35.9	70.9±28.5	NA	33.2^***^∧	ALS, ALS-P<bvFTD, ALS-FTD
BMI	25.7±4.2	29.5±3.3	28.7±4.6	29.5±4.8	24.9±3.2	117.2^***^∧	ALS<ALS-FTD, bvFTD; controls < ALS-P, ALS-FTD, bvFTD
Statin Y:N	5 : 32	3 : 10	3 : 12	7 : 25	7 : 24	1.8^#	NA

^*p < 0.05, ^**p < 0.01, ^***p < 0.001; NA, not applicable; Y, yes; N, no. Data presented as mean±standard deviation. ^#Chi-square value. ^∧Kruskal-Wallis H value.

Lipid levels (Fig. 1)

No group differences were found for total cholesterol levels (H(4) = 6.5, p = 0.163) or for LDL levels (H(4) = 6.1, p = 0.188) (Fig. 1). In contrast, a significant group difference was found for HDL levels (H(4) = 19.4, p = 0.001), with the bvFTD group, ALS-FTD, ALS-P and ALS groups exhibiting a significantly lower HDL cholesterol level compared to controls. Triglyceride levels differed across groups (H(4) = 21.9, p < 0.001). Post hoc tests showed that the bvFTD, ALS-FTD, ALS-P and ALS groups had increased triglyceride level compared to the control groups. A group difference was also found for the total cholesterol to HDL ratio (H(4) = 20.2, p < 0.001), with bvFTD, ALS-FTD, ALS-P and ALS groups showing an increased ratio compared to controls. Changes in triglyceride and HDL levels, and in the total cholesterol to HDL ratio were uncovered along the ALS-FTD spectrum, with the most prominent changes in lipid levels seen as patients developed progressive cognitive and behavioralchanges.

Fig.1

Lipid levels across the ALS-FTD diagnostic spectrum. A) Total cholesterol and LDL cholesterol. B) Triglyceride and HDL cholesterol levels. A significant group difference was found for HDL levels (H(4) = 19.4, p = 0.001), with the bvFTD group (mean = 1.4 mmol/L; U = 199, p < 0.001), ALS-FTD (mean = 1.3 mmol/L; U = 83, p = 0.003), ALS-P (mean = 1.5 mmol/L; U = 114, p = 0.01), and ALS groups (mean = 1.5 mmol/L; U = 305, p = 0.002) exhibiting a significantly lower HDL cholesterol level compared to controls (mean = 1.9 mmol/L). Triglyceride levels differed across groups (H(4) = 21.9, p < 0.001). Post hoc tests showed that the bvFTD group (mean = 1.9 mmol/L; U = 214, p < 0.001), ALS-FTD (mean = 2.3 mmol/L; U = 52, p < 0.001), ALS-P (mean = 1.6 mmol/L; U = 103, p < 0.002), and ALS (mean = 1.9 mmol/L; U = 318, p < 0.002) had increased triglyceride level compared to the control groups (mean = 1.0 mmol/L).^* ALS, ALS-P, ALS-FTD, bvFTD significantly differed from control group; all p values <0.01. C) Total cholesterol/HDL cholesterol ratio. A group difference was found for the total cholesterol to HDL ratio (H(4) = 20.2, p < 0.001), with bvFTD (mean = 4.50 mmol/L; U = 219, p < 0.001), ALS-FTD (mean = 4.8 mmol/L; U = 119, p = 0.003), ALS-P (mean = 3.9 mmol/L; U = 114, p = 0.008), and ALS groups (mean = 4.4 mmol/L; U = 256, p < 0.001) showing an increased ratio compared to controls (mean = 2.9 mmol/L). ^*ALS, ALS-P, ALS-FTD, bvFTD >Controls, all p values <0.01.

Eating behavior

Significant group differences were observed on the total APEHQ score (Table 2), both FTD groups(ALS-FTD U = 48, p = 0.01; bvFTD U = 63.0, p < 0.001) scored higher than the ALS group, with worsening changes in ALS, as cognitive changes developed. The bvFTD group demonstrated the highest score followed by the ALS-FTD group, ALS-P group and finally the ALS group (Table 2). In terms of the individual subdomains no group differences were present for swallowing or other oral behaviors. The groups differed on measures of appetite, eating behavior, and food preference, with the ALS group differing from the bvFTD group in all of the above domains and the ALS-FTD group on food preference (Table 2). When the ALS groups were subdivided in terms of bulbar versus limb onset, as expected the bulbar onset group report more significant swallowing difficulties than the limb onset group (U = 49, p = 0.004), there were no other differences on the other subdomains of the APEHQ between limb and bulbar onset ALS. As with the APEHQ on the CBI eating questionnaire, the bvFTD group demonstrated the highest scores (Table 2), but changes were also present in the ALS groups with cognitive impairment.

Table 2

Appetite and Eating Habits Questionnaire, CBI eating subscores and macronutrient intake

	ALS	ALS-P	ALS-FTD	bvFTD	Control	Kruskal –Wallis H	Post Hoc
Total APEHQ	21.1±22	37.7±27.8	46.4±29.7	65.7±55.9	NA	21.8^***	bvFTD, ALS-FTD>ALS
APEHQ- swallowing	6.7±10.7	13.4±15.5	6.7±10.5	5.1±6.5	NA	2.1
APEHQ- appetite	4.6±5.1	10.0±8.5	7.2±7.3	20.1±19.4	NA	13.7^**
APEHQ- eating habits	6.9±11.0	5.3±4.5	13.6±9.7	14.7±12.4	NA	12.9^**
APEHQ- food preference	2.5±4.8	8.4±9.1	15.4±13.5	20.4±14.9	NA	25.8^***
APEHQ- other oral behaviors	0.5±1.3	0.7±1.0	3.5±7.5	5.6±7.9	NA	9.9
CBI total eating	0.4±0.6	1.9±2.6	6.2±6.0	7.6±3.2	NA	39.4^***	ALS<ALS-FTD; ALS, ALS-P<bvFTD
Total caloric intake (KJ)	8283±3069	7689±2090	10977±6191	9735±4314	6672±1805	12.3^**	bvFTD, ALS-FTD>controls
Protein (g)	103.6±40.8	96.5±26.5	117.4±69.2	106.7±25.7	81.2±23.6	3.7	NA
Carbohydrate (g)	195.0±84.4	169.6±62.9	281.2±151.3	251.8±98.2	175.3±55.6	16.4^**	controls<ALS-FTD, bvFTD
Sugar (g)	79.8±37.2	75.9±30.4	130.5±72.5	113.8±54.2	80.7±28.5	14.1^**	controls<ALS-FTD, ALS < bvFTD
Fat (g)	88.2.0±31.7	86.3±20.2	115.8±69.4	100.6±50.6	63.7±21.5	12.1^***	ALS, ALS-P, ALS-FTD, bvFTD>controls
Monofat (g)	32.5±12.5	30.1±7.2	41.8±26.8	35.4±18.9	23.5±8.6	11.4	NA
Saturated fat (g)	36.2±14.6	38.3±10.7	49.3±30.2	43.1±22.2	23.7±8.6	20.1^***	ALS, ALS-P, ALS-FTD, bvFTD>controls
Polyfat (g)	11.6±4.5	10.6±4.0	14.5±7.8	13.9±8.6	10.4±4.7	2.0	NA

^*p < 0.05, ^**p < 0.01, ^***p < 0.001. Data presented as mean±standard deviation. NA, not applicable. Note: High scores denote increased abnormal features. The total APEHQ score reflect the combination of frequency and severity of each relevant feature investigated by the questionnaire.

Energy and food intake

Analyses of reported food intake identified group differences in total caloric intake (H = 12.3 p = 0.01; Table 2), whereby ALS-FTD (10,977 KJ; U = 63, p = 0.01) and bvFTD (9735.9 KJ; U = 142.0, p = 0.01) groups consumed more total calories per 24 hcompared to the control (6672.8 KJ) group. The ALS (8283.0 KJ) and ALS-P (7689.0 KJ) cohorts did not differ significantly in the amounts consumed (Table 2). In terms of macronutrient intake, the groups with FTD exhibited increased carbohydrate and sugar intake compared to controls (Table 2). The ALS, ALS-P, ALS-FTD, and bvFTD groups all had increased fat intake compared to controls (ALS: U = 121, p = 0.008; ALS-P: U = 42.0, p < 0.001; ALS-FTD: U = 47, p = 0.007; bvFTD: U = 124.0, p = 0.003), with saturated fat intake increased in all patient groups: ALS (U = 110, p = 0.003), ALS-P (U = 37, p < 0.001), ALS-FTD (U = 41, p = 0.003) and bvFTD (U = 94, p < 0.001) compared to controls (Table 2).

Changes relating to lipid metabolism

Correlations in all patient cohorts combined revealed associations between lipid levels andseveral measures of disease progression (disease duration and ACE-R scores), eating behavior, fat intake, and BMI (Table 3). Total triglyceridelevels were positively correlated with total fat and saturated fat intake, and BMI, but negatively correlated with ACE-R scores. Total cholesterol levels and the total cholesterol to HDL ratio were positively correlated with total caloric intake and fat intake. In addition, the cholesterol to HDL ratio was negatively associated with decreasing ACE-R scores. Finally, HDL cholesterol levels were negatively correlated with disease duration andBMI.

Table 3

Correlations between lipid levels and eating behavior BMI and measures of disease progression

Factor	Fat intake fat intake	Saturated	BMI	Total caloric intake	Disease duration	ACE-R score	APEHQ	CBI eating	CBI total
Triglyceride	0.291^*	0.285^*	0.371^**			–0.228^*
HDL cholesterol			–0.423^**		–0.274^*
Total cholesterol	0.343^**			0.274^*
Total:HDL cholesterol ratio	0.407^**	0.422^**		0.417^**		–0.228^*
Disease duration			0.315^*			–0.371^**	0.391^*	0.482^*	0.497^*
BMI						–0.270^**
ACE-R	–0.334^*			–0.280^*			–0.366^*	–0.516^**

Spearman coefficients shown. ACE-R, Addenbrooke’s Cognitive Examination; APEHQ, Appetite and Eating Health Questionnaire; CBI, Cambridge Behavioral Inventory. ^*p < 0.01, ^**p < 0.001.

Survival in relation to cholesterol and triglyceride levels

During the study period, 13% ALS patients died, with the largest majority from the ALS-FTD group. Using Cox multivariate regression analyses (Table 4), cholesterol was examined in a model with covariates (age, ACE-R score, disease duration, and diagnoses). After controlling for these variables, a lower total cholesterol level was found to be a significant predictor of shorter survival ([HR] 0.307, 95% confidence interval [CI] 0.144–0.714, p = 0.002; χ² =25.3, p = 0.001). In summary, having a lower cholesterol level increased the risk of death 3.25-fold (Table 4). There was no effect on survival of triglyceride levels ([HR] 0.820, 95% CI 0.484–1.390, p = 0.094; χ² = 12.7, p = 0.078). Body mass index was not included in the models as on univariate analysis it did not influence survival ([HR] 1.019, 95% CI 0.910–1.14, p = 0.743; χ² = 0.108, p = 0.743). Fat intake also on univariate analysis did not affectsurvival ([HR] 0.997, 95% CI 0.989–1.011, p = 0.675; χ² = 0.176, p = 0.675).

Table 4

Univariate and multivariate Cox regression analyses of effect of total cholesterol and triglyceride level on survival of ALS and bvFTD patients

	Univariate		95% CI	Multivariate^*		95% CI
Covariate	Hazard Ratio	p-value		Hazard Ratio	p-value
Triglyceride	0.776	0.221	0.487–1.22	0.820	0.461	0.484–1.390
Total Cholesterol	0.636	0.033	0.494–0.958	0.366	0.002	0.144–0.656
Age	1.1	0.927	0.970–1.047	–	–	–
ACE-R	1.2	0.347	0.972–1.047	–	–	–
Disease duration	0.209	0.001	0.089–0.493	–	–	–
Diagnoses		0.584			–	–
ALS-P^#	0.953	0.953	0.048–1.391	–	–	–
ALS-FTD^#	0.938	0.938	0.220–1.829	–	–	–
BvFTD^#	0.958	0.951	0.036–0.333	–	–	–

In the Cox model cholesterol levels and triglyceride levels were examined as separate models. Diagnosis was entered as a categorical variable ^#compared to reference category ALS. ^*controlling for age, ACE-R, disease duration, and diagnosis.

Predictors of triglyceride, HDL, and total cholesterol levels

To predict correlates of triglyceride levels, linear regression modelling considering all study participants found that total fat intake (β= 0.447, p < 0.001), BMI (β= 0.221, p = 0.036) and diagnosis (β= 0.45, p = 0.018) were predictors of triglyceride levels, explaining 34% of the score variance. Using a similar approach showed that BMI was the main predictor of HDL cholesterol (β= 0.452, p < 0.001), explaining 29% of variance, with fat intake and diagnosis not affecting HDL levels. Similarly, total fat intake (β= 0.365, p = 002) was a significant predictor of total cholesterol, which explained 17% of score variance, while BMI and diagnosis did not affect total cholesterol levels.

DISCUSSION

This study examined lipid levels across the ALS-FTD spectrum to investigate relationships between lipid levels, eating behavior, body mass index, and patient survival. Changes in triglyceride and HDL levels, and in the total cholesterol to HDL ratio were uncovered along the ALS-FTD spectrum, with the most prominent changes in lipid levels seen as patients developed progressive cognitive and behavioral changes in the cognitively affected ALS and bvFTD groups. These changes, however, were also present in the pure ALS group, although to a limited extent.

Lipid changes were associated with changes in caloric and fat intake suggesting a potential avenue to target in modifying disease outcome. Notably changes in eating behavior have been documented along the ALS- FTD spectrum [12] and may mediate lipid levels and disease progression. Linear regression analyses demonstrated that total reported fat intake, diagnosis, and BMI were significant predictors of triglyceride levels, indicating that triglyceride levels can potentially be mediated by changes in fat intake and diet and that there is a relationship with underlying diagnosis, particularly cognitive change. This was further confirmed by correlations which showed triglyceride levels also correlated with cognitive function (ACE-R scores). HDL levels were predicted by changes in BMI, while total cholesterol levels were predicted by fat intake and correlated with total caloric intake. While fat intake affected triglyceride and cholesterol levels, fat intake did not directly influence survival. There are likely complex interactions between cognitive and behavioral changes affecting eating behavior that potentially affect lipid levels and underlying pathology at both a peripheral and central level, and these need to be tested further in experimental animal models/interventionstudies.

It is not clear how changes in lipid metabolism and fat intake might affect the underlying pathology in FTD and ALS. It has recently been shown that changes in fatty acid metabolism specifically a higher proportion of monounsaturated fatty acids and a lower proportion of polyunsaturated fatty acids may also potentially mediate survival in ALS [25]. The brain has the highest concentration of lipids, second only to adipose tissue [26]. Cholesterol potentially plays a protective role by providing an energy source for neuronal cells and also has a critical role in cell membrane signaling and structure [26], while triglycerides are a source of energy storage for the body. It is likely that triglyceride levels are elevated in response to increased fat intake, which has been previously reported in ALS (both pure and those with cognitive changes) [12], presymptomatic ALS [27] and FTD cohorts [24, 28]. One possible explanation may be as a mechanism that provides a source of energy in a hypermetabolic starved state which has been documented in ALS [29 –31] and FTD [32]. Animal models of ALS have shown an increased postprandial clearance of triglycerides, which is corrected by a high fat diet [33].

In the current study, although an overall increase in total cholesterol levels was not found, elevated cholesterol levels predicted an improved survival, with a 3.25 reduction in the risk of death, independent of diagnosis, age, disease duration, and cognition. This finding fits with previous studies in pure ALS [5 , 9] that have reported a survival benefit of higher cholesterol levels in ALS patients, even when cholesterol levels are similar to controls, and extends this finding to ALS with cognitive changes and bvFTD. How cholesterol may mediate survival is also not clear, but it could potentially affect underlying pathology. In presymptomatic SOD-1 animal models, a change in muscle metabolism appears to take place, with a switch from glucose to lipid as a source of energy [34], and a high fat diet in these animals leads to prolonged survival and increased motor neuron numbers [35]. It is possible that in humans the increased fat intake may reflect a response to changes in the source of energy and lipid metabolism [36, 37] and possibly to mitochondrial dysfunction [38] and that the increased fat intake may play a role in mediating lipid levels (cholesterol and triglyceride) as a response to these changes.

It remains to be determined whether certain neural networks and regional brain areas potentially contribute to changes in lipid levels. Complex neural networks [39]— including the hypothalamus [11],striatal pathways [28, 40], and structures controlling autonomic function including the insula, mesial temporal cortex and anterior cingulate cortex [32]— control eating behavior in FTD, networks that are affected by the neurodegenerative process [41],and it is plausible through changes in eating behavior that these structures may also affect lipid levels. It is possible that the hypothalamus plays a critical role in energy homeostasis with hypothalamic atrophy identified in FTD and ALS related to eating behavior and BMI [11, 42], and pathological studies identifying increased TDP-43 pathology in the lateral hypothalamic area in ALS related to BMI [43], as well as changes in neuroendocrine peptides incorporating melanocortin pathways similar to those seen in FTD, also identified in ALS mouse models [44]. It is therefore plausible that changes in these central structures and the neurodegenerative process mediate lipid metabolism in ALS-FTD through changes in caloric and fatintake.

The role of lipids and the development of dementia and mild cognitive impairment has also received attention [45]. Several studies have found that hypertrigliceremia, and low HDL are associated with the development of mild cognitive impairment [45, 46]. The use of statins [47] has been associated with a lower incidence of dementia and cognitive impairment. However, evidence suggests that increased cholesterol and lipid levels may be protective and reduce mortality rates with aging particularly after the age of 70 [48, 49]. This raises questions about the role of lipids in neurodegeneration and if they may initially promote the process of neurodegeneration as suggested in Alzheimer’s disease animal models, but then later transition to take on a protective role in terms of providing an energy source for neuronal cells. Animal studies will likely prove useful in examining the interactions between diet, types of lipids (triglycerides versus cholesterols), genetic status and underling pathology and the neurodegenerativeprocess.

The present study uncovered both stable and more variable changes in lipid metabolism across the ALS-FTD diagnostic spectrum with increased total cholesterol levels correlating with increased survival. It is proposed that the variations in lipid levels develop as a result of changes in eating behavior, caloric intake, fat intake, and through the underlying neurodegenerative process. At this stage, it remains to be determined how these changes affect underlying disease pathology, how they may change with disease progression, and how site of onset and cognitive versus motor onset in ALS- FTD may affect lipid metabolism. Future research is required as to whether modification of lipid levels clinically may be harnessed to slow disease progression and thereby improve patient prognosis in both frontotemporal dementia and amyotrophic lateralsclerosis.

Footnotes

ACKNOWLEDGMENTS

We thank Heidi Cartwright for assistance with figure preparation.

This work was supported by funding to Forefront, a collaborative research group dedicated to the study of frontotemporal dementia and amyotrophic lateral sclerosis, from the National Health and Medical Research Council of Australia (NHMRC) program grant (#1037746 to GH, MK, and JH), the Australian Research Council Centre of Excellence in Cognition and its Disorders Memory Program (#CE110001021 to OP and JH), and other grants/sources (NHMRC project grant #1003139), The Brain foundation of Australia and The Royal Australasian College of Physicians (Vincent Fairfax Foundation research fellowship). We are grateful to the research participants involved with the Forefront research studies. RA is a NHMRC Early Career Fellow (#1120770). GH is a NHMRC Senior Principal Research Fellow (#1079679). OP is an NHMRC Senior Research Fellow (#1103258). SF is supported by the Wellcome Trust, the National Institute for Health Research Cambridge Biomedical Research Centre, and the Bernard Wolfe Health Neuroscience Fund.

Authors’ disclosures available online ().

References

Golomb

, Kwon

, Koperski

, Evans

(2009) Amyotrophic lateral sclerosis-like conditions in possible association with cholesterol-lowering drugs: An analysis of patient reports to the University of California, San Diego (UCSD) Statin Effects Study. Drug Saf 32, 649–661.

Colman

, Szarfman

, Wyeth

, Mosholder

, Jillapalli

, Levine

, Avigan

(2008) An evaluation of a data mining signal for amyotrophic lateral sclerosis and statins detected in FDA’s spontaneous adverse event reporting system. Pharmacoepidemiol Drug Saf 17, 1068–1076.

Reich-Slotky

, Andrews

, Cheng

, Buchsbaum

, Levy

, Kaufmann

, Thompson

(2013) Body mass index (BMI) as predictor of ALSFRS-R score decline in ALS patients. Amyotroph Lateral Scler Frontotemporal Degener 14, 212–216.

Dupuis

, Corcia

, Fergani

, Gonzalez De Aguilar

, Bonnefont-Rousselot

, Bittar

, Seilhean

, Hauw

, Lacomblez

, Loeffler

, Meininger

(2008) Dyslipidemia is a protective factor in amyotrophic lateral sclerosis. Neurology 70, 1004–1009.

Dorst

, Kuhnlein

, Hendrich

, Kassubek

, Sperfeld

, Ludolph

(2011) Patients with elevated triglyceride and cholesterol serum levels have a prolonged survival in amyotrophic lateral sclerosis. J Neurol 258, 613–617.

Ikeda

, Hirayama

, Takazawa

, Kawabe

, Iwasaki

(2012) Relationships between disease progression and serum levels of lipid, urate, creatinine and ferritin in Japanese patients with amyotrophic lateral sclerosis: A cross-sectional study. Intern Med 51, 1501–1508.

Yang

, Kim

, Park

, Kim

, Lee

, Sung

(2013) Hypolipidemia in patients with amyotrophic lateral sclerosis: A possible gender difference? J Clin Neurol 9, 125–129.

Chio

, Calvo

, Ilardi

, Cavallo

, Moglia

, Mutani

, Palmo

, Galletti

, Marinou

, Papetti

, Mora

(2009) Lower serum lipid levels are related to respiratory impairment in patients with ALS. Neurology 73, 1681–1685.

Sutedja

, van der Schouw

, Fischer

, Sizoo

, Huisman

, Veldink

, Van den Berg

(2011) Beneficial vascular risk profile is associated with amyotrophic lateral sclerosis. J Neurol Neurosurg Psychiatry 82, 638–642.

10.

Ahmed

, Irish

, Piguet

, Halliday

, Ittner

, Farooqi

, Hodges

, Kiernan

(2016) Amyotrophic lateral sclerosis and frontotemporal dementia: Distinct and overlapping changes in eating behaviour and metabolism. Lancet Neurol 15, 332–342.

11.

Ahmed

, Latheef

, Bartley

, Irish

, Halliday

, Kiernan

, Hodges

, Piguet

(2015) Eating behavior in frontotemporal dementia: Peripheral hormones vs hypothalamic pathology. Neurology 85, 1310–1317.

12.

Ahmed

, Caga

, Devenney

, Hsieh

, Bartley

, Highton-Williamson

, Ramsey

, Zoing

, Halliday

, Piguet

, Hodges

, Kiernan

(2016) Cognition and eating behavior in amyotrophic lateral sclerosis: Effect on survival. J Neurol 263, 1593–603.

13.

EFNS Task Force on Diagnosis and Management of Amyotrophic Lateral Sclerosis, Andersen

, Abrahams

, Borasio

, de Carvalho

, Chio

, Van Damme

, Hardiman

, Kollewe

, Morrison

, Petri

, Pradat

, Silani

, Tomik

, Wasner

, Weber

(2012) EFNS guidelines on the clinical management of amyotrophic lateral sclerosis (MALS)–revised report of an EFNS task force. Eur J Neurol 19, 360–375.

14.

Strong

(2008) The syndromes of frontotemporal dysfunction in amyotrophic lateral sclerosis. Amyotroph Lateral Scler 9, 323–338.

15.

Rascovsky

, Hodges

, Knopman

, Mendez

, Kramer

, Neuhaus

, van Swieten

, Seelaar

, Dopper

, Onyike

, Hillis

, Josephs

, Boeve

, Kertesz

, Seeley

, Rankin

, Johnson

, Gorno-Tempini

, Rosen

, Prioleau-Latham

, Lee

, Kipps

, Lillo

, Piguet

, Rohrer

, Rossor

, Warren

, Fox

, Galasko

, Salmon

, Black

, Mesulam

, Weintraub

, Dickerson

, Diehl-Schmid

, Pasquier

, Deramecourt

, Lebert

, Pijnenburg

, Chow

, Manes

, Grafman

, Cappa

, Freedman

, Grossman

, Miller

(2011) Sensitivity of revised diagnostic criteria for the behavioural variant of frontotemporal dementia. Brain 134, 2456–2477.

16.

Kaufmann

, Levy

, Thompson

, Delbene

, Battista

, Gordon

, Rowland

, Levin

, Mitsumoto

(2005) The ALSFRSr predicts survival time in an ALS clinic population. Neurology 64, 38–43.

17.

Mioshi

, Caga

, Lillo

, Hsieh

, Ramsey

, Devenney

, Hornberger

, Hodges

, Kiernan

(2014) Neuropsychiatric changes precede classic motor symptoms in ALS and do not affect survival. Neurology 82, 149–155.

18.

Mioshi

, Lillo

, Yew

, Hsieh

, Savage

, Hodges

, Kiernan

, Hornberger

(2013) Cortical atrophy in ALS is critically associated with neuropsychiatric and cognitive changes. Neurology 80, 1117–1123.

19.

Strong

, Abrahams

, Goldstein

, Woolley

, McLaughlin

, Snowden

, Mioshi

, Roberts-South

, Benatar

, HortobaGyi

, Rosenfeld

, Silani

, Ince

, Turner

(2017) Amyotrophic lateral sclerosis –frontotemporal spectrum disorder (ALS-FTSD): Revised diagnostic criteria. Amyotroph Lateral Scler Frontotemporal Degener 18, 153–174.

20.

Wedderburn

, Wear

, Brown

, Mason

, Barker

, Hodges

, Williams-Gray

(2008) The utility of the Cambridge Behavioural Inventory in neurodegenerative disease. J Neurol Neurosurg Psychiatry 79, 500–503.

21.

Lillo

, Mioshi

, Zoing

, Kiernan

, Hodges

(2011) How common are behavioural changes in amyotrophic lateral sclerosis? Amyotroph Lateral Scler 12, 45–51.

22.

Lillo

, Savage

, Mioshi

, Kiernan

, Hodges

(2012) Amyotrophic lateral sclerosis and frontotemporal dementia: A behavioural and cognitive continuum. Amyotroph Lateral Scler 13, 102–109.

23.

Mioshi

, Dawson

, Mitchell

, Arnold

, Hodges JR (2006) The Addenbrooke’s Cognitive Examination Revised (ACE-R): A brief cognitive test battery for dementiascreening. Int J Geriatr Psychiatry 21, 1078–1085.

24.

Ahmed

, Irish

, Kam

, van Keizerswaard

, Bartley

, Samaras

, Hodges

, Piguet

(2014) Quantifying the eating abnormalities in frontotemporal dementia. JAMA Neurol 71, 1540–1546.

25.

Henriques

, Blasco

, Fleury

, Corcia

, Echaniz-Laguna

, Robelin

, Rudolf

, Lequeu

, Bergaentzle

, Gachet

, Pradat

, Marchioni

, Andres

, Tranchant

, Gonzalez De Aguilar

, Loeffler

(2015) Blood cell palmitoleate-palmitate ratio is an independent prognostic factor for amyotrophic lateral sclerosis. PLoS One 10, e0131512.

26.

Adibhatla

, Hatcher

(2007) Role of lipids in brain injury and diseases. Future Lipidol 2, 403–422.

27.

Huisman

, Seelen

, van Doormaal

, de Jong

, de Vries

, van der Kooi

, de Visser

, Schelhaas

, van den Berg

, Veldink

(2015) Effect of presymptomatic body mass index and consumption of fat and alcohol on amyotrophic lateral sclerosis. JAMA Neurol 72, 1155–1162.

28.

Ahmed

, Irish

, Henning

, Dermody

, Bartley

, Kiernan

, Piguet

, Farooqi

, Hodges

(2016) Assessment of Eating behavior disturbance and associated neural networks in frontotemporal dementia. JAMA Neurol 73, 282–290.

29.

Dupuis

, Pradat

, Ludolph

, Loeffler

(2011) Energy metabolism in amyotrophic lateral sclerosis. Lancet Neurol 10, 75–82.

30.

Desport

, Torny

, Lacoste

, Preux

, Couratier

(2005) Hypermetabolism in ALS: Correlations with clinical and paraclinical parameters. Neurodegener Dis 2, 202–207.

31.

Kasarskis

, Mendiondo

, Matthews

, Mitsumoto

, Tandan

, Simmons

, Bromberg

, Kryscio

, ALS Nutrition/NIPPV Study Group (2014) Estimating daily energy expenditure in individuals with amyotrophic lateral sclerosis. Am J Clin Nutr 99, 792–803.

32.

Ahmed

, Landin-Romero

, Collet

, van der Klaauw

, Devenney

, Henning

, Kiernan

, Piguet

, Farooqi

, Hodges

(2017) Energy expenditure in frontotemporal dementia: A behavioural and imaging study. Brain 140, 171–183.

33.

Fergani

, Oudart

, Gonzalez De Aguilar

, Fricker

, Rene

, Hocquette

, Meininger

, Dupuis

, Loeffler

(2007) Increased peripheral lipid clearance in an animal model of amyotrophic lateral sclerosis. J Lipid Res 48, 1571–1580.

34.

Palamiuc

, Schlagowski

, Ngo

, Vernay

, Dirrig-Grosch

, Henriques

, Boutillier

, Zoll

, Echaniz-Laguna

, Loeffler

, Rene

(2015) A metabolic switch toward lipid use in glycolytic muscle is an early pathologic event in a mouse model of amyotrophic lateral sclerosis. EMBO Mol Med 7, 526–546.

35.

Dupuis

, Oudart

, Rene

, Gonzalez de Aguilar

, Loeffler

(2004) Evidence for defective energy homeostasis in amyotrophic lateral sclerosis: Benefit of a high-energy diet in a transgenic mouse model. Proc Natl Acad Sci U S A 101, 11159–11164.

36.

Schmitt

, Hussain

, Dupuis

, Loeffler

, Henriques

(2014) A plural role for lipids in motor neuron diseases: Energy, signaling and structure. Front Cell Neurosci 8, 25.

37.

Dodge

, Treleaven

, Pacheco

, Cooper

, Bao

, Abraham

, Cromwell

, Sardi

, Chuang

, Sidman

, Cheng

, Shihabuddin

(2015) Glycosphingolipids are modulators of disease pathogenesis in amyotrophic lateral sclerosis. Proc Natl Acad Sci U S A 112, 8100–8105.

38.

Muyderman

, Chen

(2014) Mitochondrial dysfunction in amyotrophic lateral sclerosis –a valid pharmacological target? Br J Pharmacol 171, 2191–2205.

39.

Ahmed

, Devenney

, Irish

, Ittner

, Naismith

, Ittner

, Rohrer

, Halliday

, Eisen

, Hodges

, Kiernan

(2016) Neuronal network disintegration: Common pathways linking neurodegenerative diseases. J Neurol Neurosurg Psychiatry 87, 1234–1241.

40.

Perry

, Sturm

, Seeley

, Miller

, Kramer

, Rosen

(2014) Anatomical correlates of reward-seeking behours in behavioural variant frontotemporal dementia. Brain 137, 1621–1626.

41.

Seeley

, Crawford

, Rascovsky

, Kramer

, Weiner

, Miller

, Gorno-Tempini

(2008) Frontalparalimbic network atrophy in very mild behavioral variant frontotemporal dementia. Arch Neurol 65, 249–255.

42.

Ahmed

, Farooqi

(2017) Hypothalamic atrophy is related to body mass index and age at onset in amyotrophic lateral sclerosis. J Neurol Neurosurg Psychiatry, doi: 10.1136/jnnp-2017-316382

43.

Cykowski

, Takei

, Schulz

, Appel

, Powell

(2014) TDP-43 pathology in the basal forebrain and hypothalamus of patients with amyotrophic lateral sclerosis. Acta Neuropathol Commun 2, 171.

44.

Vercruysse

, Sinniger

, El Oussini

, Scekic-Zahirovic

, Dieterle

, Dengler

, Meyer

, Zierz

, Kassubek

, Fischer

, Dreyhaupt

, Grehl

, Hermann

, Grosskreutz

, Witting

, Van Den Bosch

, Spreux-Varoquaux

, GERP ALS Study Group, Ludolph

, Dupuis

(2016) Alterations in the hypothalamic melanocortin pathway in amyotrophic lateral sclerosis. Brain 139, 1106–1122.

45.

Frisardi

, Solfrizzi

, Seripa

, Capurso

, Santamato

, Sancarlo

, Vendemiale

, Pilotto

, Panza

(2010) Metabolic-cognitive syndrome: A cross-talk between metabolic syndrome and Alzheimer’s disease. Ageing Res Rev 9, 399–417.

46.

Panza

, Frisardi

, Capurso

, Imbimbo

, Vendemiale

, Santamato

, D’Onofrio

, Seripa

, Sancarlo

, Pilotto

, Solfrizzi

(2010) Metabolic syndrome and cognitive impairment: Current epidemiology and possible underlying mechanisms. J Alzheimers Dis 21, 691–724.

47.

Cramer

, Haan

, Galea

, Langa

, Kalbfleisch

(2008) Use of statins and incidence of dementia and cognitive impairment without dementia in a cohort study. Neurology 71, 344–350.

48.

Kronmal

, Cain

, Ye

, Omenn

(1993) Total serum cholesterol levels and mortality risk as a function of age.A report based on the Framingham data. Arch Intern Med 153, 1065–1073.

49.

Piguet

, Grayson

, Creasey

, Bennett

, Brooks

, Waite

, Broe

(2003) Vascular risk factors, cognition and dementia incidence over 6 years in the Sydney Older Persons Study. Neuroepidemiology 22, 165–171.