Could teacher-perceived parental interest be an important factor in understanding how education relates to later physiological health? A life course approach.

Education is associated with later health, and notably with an indicator of physiological health measuring the cost of adapting to stressful conditions, named allostatic load. Education is itself the result of a number of upstream variables. We examined the origins of educational attainment through the lens of interactions between families and school i.e. parents' interest in their child's education as perceived by teachers. This study aims to examine whether parental interest during a child's educational trajectory is associated with subsequent allostatic load, and whether education or other pathways mediate this relationship. We used data from 9 377 women and men born in 1958 in Great Britain and included in the National Child Development Study to conduct secondary data analyses. Parental interest was measured from questionnaire responses by teachers collected at age 7, 11 and 16. Allostatic load was defined using 14 biomarkers assayed in blood from a biosample collected at 44 years of age. Linear regression analyses were carried out on a sample of 8 113 participants with complete data for allostatic load, missing data were imputed. Participants whose parents were considered to be uninterested in their education by their teacher had a higher allostatic load on average in midlife in both men (β = 0,41 [0,29; 0,54]) and women (β = 0,69 [0,54; 0,83]). We examined the role of the educational and other pathways including psychosocial, material/financial, and behavioral variables, as potential mediators in the relationship between parental interest and allostatic load. The direct link between parental interest and allostatic load was completely mediated in men, but only partially mediated in women. This work provides evidence that parents' interest in their child's education as perceived by teachers is associated with subsequent physiological health in mid-life and may highlight a form of cultural dissonance between family and educational spheres.

1. Introduction

One of the most consistent findings in the field of social epidemiology is that educational attainment is associated with health. Across countries, and over time, lower educational attainment has been associated with poorer health outcomes [1]. These associations are often explained by the fact that better educated people are less likely to experience the disadvantaged material conditions or psychosocial distress caused by economic hardship and tend to have healthier lifestyles compared to the less educated [2]. However, Gallo et al. showed that health behaviors and lifestyle factors explained only a part of the educational inequalities in total mortality of members of European cohorts [3]. Beyond these groups of mechanisms others have been examined, such as environmental exposures and material conditions [4, 5], calling for further studies to examine the mechanisms through which education relates to health.

The concept of embodiment rests upon a key set of mechanisms likely to underlie the relationship between education and health. This concept “refers to how we, like any living organism, literally incorporate, biologically, the world in which we live, including our societal and ecological circumstances" [6]. Embodiment can occur through several mechanisms including the internal physiological response to social exposures, among which feature intra-familial relationships and social interactions. Such interactions are perceived through the senses, interpreted by the central nervous system leading to peripheral physiological responses [7]. These physiological responses are adaptive processes which maintain physiological stability in response to environmental challenges [8]. The repeated activation of compensatory physiological mechanisms as a result of chronic exposure to stress can lead to physiological wear-and-tear, termed allostatic load. Allostatic load measures the consequence of a prolonged activation of the stress response system by external challenges, leading to physiological imbalances across systems [9]. Previous research has shown that allostatic load is associated with physical and functional decline, cardiovascular events, and mortality [10, 11]. A few studies have examined the association between education and allostatic load [12, 13] indicating that a higher level of education is associated with a lower allostatic load, only partly explained by different potential mediators (including health, behavioral and psychosocial factors) suggesting that an important part of this association remains to be explained.

Life course research indicates that the dynamic processes of adaptive allostasis most likely begin in early life [14, 15]. One recent study documented that disadvantaged early life socioeconomic conditions are associated with an increased risk of having a higher allostatic load in midlife, mainly through educational pathways [16]. Elsewhere, Hamdi et al. reported that the relationship between education and allostatic load may be partly explained by family influences [17]. As such, educational attainment “is an excellent marker of the ‘healthfulness’ of accumulated childhood experience” [18] as the social environment in early life may therefore have lasting effects on different social, biological, and behavioral factors that might act as mechanisms connecting education to repeated stress, which in turn affects health [19]. Because educational attainment is the end point of a complex process where a wide array of factors (institutional, interpersonal, individual) shape trajectories of schooling, there is a need to look upstream in order to capture education as a long-term process grounded in a broader social context [19]. However, it remains as yet unclear which elements of the early life environment upstream of education are likely to be involved in the embodiment dynamic, leading to physiological wear-and-tear.

An analysis of the 1958 British Birth Cohort Study suggested that parental interest in their child’s education could be an early life factor contributing to self-rated health at 33 years of age [20]. More broadly parental interest in their child’s education has also been reported to have positive effects on psychosocial adjustment [21] later mental health [22] and metabolic outcomes in middle-age [23]. Moreover, parental interest in their offspring’s schooling has been identified as a determinant of educational success [24-27]. This parental involvement in children’s academic socialisation, has been shown to influence academic success over and above child’s social class and cognitive abilities [28]. Consequently, parental interest in their child’s education could be a variable to consider, when examining how the relationship between the home and school environments in early life affects physiological wear-and-tear, through educational attainment.

Parental interest in their child’s education can be reported by the parents themselves, reflecting their perception of their own involvement in the upbringing of their children. It can also be reported by teachers. Teacher’s assessments provide one perspective of the situation which may partly reflect the position and viewpoint of the educational institution in terms of children’s compliance with academic requirements and potentially capture the tension between the home and school environments experienced by some children [29]. This dissonance that children may experience when exposed to different family and school environments, may lead to challenges when adapting to school. Such an early life stressor may lead to the chronic solicitation of children’s physiological stress response systems, in turn affecting subsequent physiological health.

A previous study in a Swedish cohort examined the association between parental academic involvement as perceived by teachers and allostatic load in midlife. Rather than parental social class or availability of practical academic support, parental interest in their children’s studies during the last year of school was found to predict adult allostatic load, largely mediated by academic achievement, which the authors attributed to the potential consequence of physiological stress over the life course [30]. Our study aims to examine the association between parents’ interest in their child’s education as perceived by teachers, and multi-system physiological dysregulation as measured by allostatic load, and whether education and other pathways mediate this relationship.

We hypothesize that parents’ interest in their child’s education as perceived by teachers is an indicator of academic socialisation that may capture early life stressors and thus be related to later physiological wear-and-tear, partly through education and other pathways. In this study, we take a life course approach to (i) test whether parental interest is associated with allostatic load, and (ii) explore the educational and other pathways through which parental interest may be differentially embodied during childhood, adolescence and early adulthood, leading to physiological wear-and-tear, as measured by allostatic load.

2. Materials and methods

2.1. Study population

Our study is based on secondary analyses of data from the 1958 National Child Development Study (NCDS), an observational prospective population cohort study, which included all live births in Great Britain during one week in 1958 (n = 18 555). The NCDS has been described in detail elsewhere [31]. Data collection on health, economic, social and developmental factors was carried out on cohort members from birth until now at age 7, 11, 16, 23, 33, 42, 44/45, 46, 50, 55 and 62 years and conducted by the Centre for Longitudinal Studies. Written informed consent was obtained from parents for childhood measurements and ethical approval for the adult data collection was obtained from the National Research Ethics Advisory Panel. When cohort participants were 44 and 45 years of age, a biomedical survey was conducted including a self-reported questionnaire, blood and saliva samples as well as anthropometric measurements with data available for 9 377 individuals. Ethical approval for the age 45 survey was given by the South East Multicentre Research Ethics Committee. Participants in this survey were found to be representative of the general cohort [32]. A total of 1 264 participants were excluded from our analyses, including pregnant women and those from whom blood was not obtained, leaving 8 113 participants. Our sample selection flow-chart is presented in Fig 1. NCDS data are open access datasets available to non-profit research organizations through the UK Data Service.

Fig 1

Diagram of inclusion and exclusion criteria for the analysis from the biomedical survey of the NCDS 58.

2.2. Allostatic load at age 44–45

The allostatic load score was constructed based on previous work using the NCDS the same initial definition of allostatic load [33]: in order to represent four physiological systems, 14 available biomarkers were used: the neuroendocrine system (salivary cortisol t1, salivary cortisol t1–t2); the immune and inflammatory system (insulin-like growth factor-1 (IGF1), C-reactive protein (CRP), fibrinogen, Immunoglobulin E (IgE)); the metabolic system (high-density lipoprotein (HDL), low-density lipoprotein (LDL), triglycerides, glycosylated hemoglobin (HbA1C)); the cardiovascular and respiratory systems: (systolic blood pressure (SBP), diastolic blood pressure (DBP), heart rate, peak expiratory flow). Using sex-specific quartiles, each biomarker was dichotomized into "high" (coded as 1) and "low" (coded as 0) risk. The sum of these 14 dichotomized biomarkers resulted in an overall allostatic load score ranging from 0 to 14, where a higher score represented a higher allostatic load, and a greater health risk. We also recoded allostatic load score into a 3 category variable where a score of 0–2 was considered to be “low”, 3–4 as “middle”, and 5–14 as “high” based on tertiles in the sample [34].

2.3. Parents’ interest in their child’s education as perceived by teachers

Parental interest was measured at ages 7, 11 and 16 using information provided by the child’s teachers. The teacher was asked to report the level of interest of each parent in their child’s education through closed questions with four possible answers: Overly concerned; Very interested; Some interest; and Little interest. Based on this, we created a new binary variable for parental interest aiming to identify parents “interested” or with”low/no interest” in their child’s education, according to the teacher. We grouped the “overly concerned” and “very interested” categories together to represent the “interested” category, while grouping the “some interest” with “little interest” categories together to represent “low/no interest”. We hypothesized that interest from both parents at any one age belongs to the category “interested”. However, if only one of the parents was considered to be interested or if neither were, we considered this to belong to the category “low/no interest”. We conducted a sensitivity analysis to examine the stability of parental interest, using a series of regression analyses to identify whether changes to the categories (Overly concerned; Very interested; Some interest; and Little interest) had an effect on the association with allostatic load. We observed no change to the results (S1 Table).

2.4. Prior confounders

To examine the relationship between parental interest and allostatic load, prior confounding variables potentially associated with both parental interest and allostatic load were added to the multivariable models. We selected variables most likely to be social or biological confounding factors from questionnaires completed during childhood by parents of cohort members, based on the literature.

Socioeconomic confounders were included using parental socioeconomic classification of occupations (SEC), via a questionnaire completed at birth (I-professional occupations & II-intermediate occupations/III-skilled occupations (non-manual)/III-partly skilled occupations (manual)/IV-partly skilled occupations & V-unskilled occupations) and using information on material living conditions, collected at ages 7, 11 and 16 (advantaged/disadvantaged).

Confounding factors related to parental education were also included using parental educational attainment at birth for the mother and at 7 year for the father (left school <14 year/ lefts school ≥15 year), and parenting practices including reading to the child and outdoor activities, measured at age 7 (“Frequent/Occasionally, Hardly ever) [35].

Other prior confounding variables were also selected. At ages 7, 11, and 16, a binary adverse childhood experiences variable (ACEs) was constructed, as well as a binary childhood pathologies variable. Using data collected at age 7, a birth order variable was created (Single child/Eldest/2nd place or more), and an assessment of the child’s cognitive ability at age 7 (Copy-a-Design test where scores range between 0 and 12). See S1 File for more information about early life confounder variables.

2.5. Intermediate life course variables

In order to determine whether any observed associations between parental interest and allostatic load were due to subsequent adult intermediate factors, we selected four intermediate groups of life course variables as pathways between parental interest and health, based on epidemiological evidence and on empirical studies. The following mediating factors were then added to the models: i) Parental interest could promote a positive accumulation of academic and social success [36], reducing the impact of economic difficulties on educational attainment [37]. We therefore added the respondent’s: educational attainment at age 23 (A level-12 years of education/O levels-10 years of education/ No qualifications), occupational social class at 33 y (I-professional occupations & II-intermediate occupations/III-skilled occupations (non-manual)/III-partly skilled occupations (manual)/IV-partly skilled occupations & V-unskilled occupations) and wealth at age 33 using a wealth variable based on information about home ownership and the price of the house adjusted for economic inflation of the year of purchase (owner highest price/owner high price/owner median price/owner lowest price/not owner]; ii) Tension between home and school environments experienced by some children may provide an insecure social environment, contributing to health through psychological processes whereby resulting negative emotions through psycho-neuroendocrine mechanisms could lead to a worse health status [38]. To capture this, we added psychological/psychosocial status [malaise inventory at age 23 (No psychological distress/ psychological distress); sense of personal control at age 33 (Internal/external)]; iii) The adoption of health behaviors is closely related to education, whereby people with a higher education level are more likely to have protective health behaviors, impacting physiological functioning [16, 39]. Health behaviors at age 42 were considered as a proxy for behavioral patterns in adulthood included self-reported physical activity, alcohol consumption and smoking status.

We defined the sets of mediators according to the temporal and causal assumptions of the life course framework. As such, we made a pragmatic methodological assumption about the temporal ordering of variables in order to examine the mediating pathways whereby the exposure variable (parental interest at age 7, 11 and 16) preceded the respondent’s education, which preceded the other intermediate variables (psychological distress, socioeconomic position and health behaviors).

2.6. Statistical analysis

Our analyses were stratified by sex since behavior at school differs between girls and boys. Girls tend to be more compliant with institutional rules, facilitating the teacher’s task, boys are more frequently in conflict between academic expectations and their socially recognized particularities. Also, this allows us to consider sex/gender differences in health.

First, descriptive and bivariate statistics were carried out, using the chi-squared test for categorical variables and Wilcoxon signed-rank test or Student t-test for continuous variables. We considered allostatic load as a categorical variable in three groups, in order to ascertain any association between the covariates and allostatic load. Second, to study the association between parental interest and allostatic load, regression coefficients and 95% confidence intervals (CI) were estimated using linear regressions where allostatic load was entered as a continuous variable. We compared regression coefficients across nested models to observe the change in effect according to subsequent adjustments.

To study the link between parental interest and allostatic load

Model 1: Linear regression between parental interest and allostatic load

Model 2: Model 1 plus baseline confounders (parental SEC, material living conditions, level of parental education, reading and outdoor activities, ACEs, health problems in childhood, sibling order, cognitive ability).

To qualify pathways mediating the relationship between parental interest and allostatic load

In these models, we considered that the last set of mediators had an important mediating role if the change of the regression coefficient characterizing the association between parental interest and allostatic load was large and if these mediators were associated to allostatic load.

Model 3: Model 2 plus educational level

Model 4: Model 3 plus psychosocial/psychological variables (Sense of personal control, malaise)

Model 5: Model 4 plus socioeconomic status/financial variables (Occupational social class and wealth)

Model 6: Model 5 plus health behaviors variables (smoking, alcohol consumption, physical activity)

Third and finally, to disentangle and quantify the direct and indirect effect for parental interest on allostatic load, we carried-out different linear regression steps in both men and women, to analyze mediation, applying the method of VanderWeele and Stijn Vansteelandt [40]. See S2 File for more information on the different steps of the analysis of estimated direct and indirect effects.

In order to control for potential biases due to missing data, multivariable analyses were conducted on the multiply imputed data using the ICE method available in Stata ®v14. Twenty imputations were performed assuming that the data were missing at random (MAR). Comparisons were then made between complete-case multivariable analyses and multivariable analyses based on imputation estimates (S2 Table).

3. Results

Descriptive statistics for men and women on the complete-case data are provided in S3 Table. The majority of our sample (78% in men and 75% in women) had a low [0-2] or medium [3-4] allostatic load in midlife. Additionally, during childhood, 47% of the cohort members’ parents were perceived by the teacher as taking interest in their child’s education, while 42% were described as uninterested or not very interested.

The bivariate analyses by allostatic load group on the complete-case data are provided for men in Table 1A and women in Table 1B. In both sexes, we observed a higher proportion of cohort members whose parents were deemed uninterested in their child’s education by the teacher in the high allostatic load group, compared to the medium and low allostatic load group. Regarding the early life social environment, both men and women with a high allostatic load at age 44 were more likely to have had parents with low SEC, lived in disadvantaged material conditions, had a father or a mother who left school before the age of 14, been exposed to adversity and to have had a lower score on the copy-a-design test, compared to those with medium and low allostatic load. In addition, men with a high allostatic load at age 44 were more likely to have been an only child, to have engaged occasionally or rarely in outdoor activities and to have had health problems, than those with medium and low allostatic load. Regarding intermediate variables in adulthood, both men and women with a high allostatic load at 44y were more likely to have had a low level of education, an external sense of personal control, a lower social position (occupational social class and wealth), to smoke heavily, to be heavy drinkers or abstainers and to have had a low level of physical activity, than those with medium and low allostatic load. In addition, women with a high allostatic load at 44y were more likely to report a psychological distress, than those with medium and low allostatic load.

Table 1

Bivariate statistics on the complete-case sample (n = 8 113): Allostatic load categories according to confounding and intermediate variables in (A) men and in (B) women.

	Allostatic load at 44y
	A. Men n(%) n = 4 075 (50%)				B. Women n(%) n = 4 056 (50%)
	Low n(%)	Medium n(%)	High n(%)	P-value*	Low n(%)	Medium n(%)	High n(%)	P-value*
	1793 (44.20%)	1386 (34.16%)	878 (21.64%)		1824 (44.97%)	1235 (30.45%)	997 (24.58%)
Parental interest (7-16y)
Both interested	930 (51.87%)	634 (45.74%)	348 (39.64%)	<0.001	989 (54.22%)	555 (44.94%)	362 (36.31%)	<0.001
Low/No interest	698 (38.93%)	578 (41.70%)	427 (48.63%)		654 (35.86%)	524 (42.43%)	488 (48.95%)
Missing	165 (9.20%)	174 (12.55%)	103 (11.73%)		181 (9.92%)	156 (12.63%)	147 (14.74%)
Prior confounders
Parental SEC (birth)
I & II	405 (22.59%)	240 (17.32%)	103 (11.73%)	<0.001	414 (22.70%)	184 (14.90%)	108 (10.83%)	<0.001
IIINM	196 (10.93%)	117 (8.44%)	67 (7.63%)		201 (11.02%)	110 (8.91%)	80 (8.02%)
IIIM	805 (44.90%)	666 (48.05%)	442 (50.34%)		789 (43.26%)	613 (49.64%)	509 (51.05%)
IV&V	295 (16.45%)	262 (18.90%)	211 (24.03%)		309 (16.94%)	248 (20.08%)	240 (24.07%)
Missing	92 (5.13%)	101 (7.29%)	55 (6.26%)		111 (6.09%)	80 (6.48%)	60 (6.02%)
Material living conditions (7y)
Advantaged	1191 (66.42%)	804 (58.01%)	500 (56.95%)	<0.001	1158 (63.49%)	735 (59.51%)	570 (57.17%)	<0.001
Disadvantaged	337 (18.80%)	360 (25.97%)	241 (27.45%)		399 (21.88%)	308 (24.94%)	300 (30.09%)
Missing	265 (14.78%)	222 (16.02%)	137 (15.60%)		267 (14.64%)	192 (15.55%)	127 (12.74%)
Father’s level of education (7y)
Left school ≥15y	451 (25.15%)	287 (20.71%)	135 (15.38%)	<0.001	487 (26.70%)	252 (20.40%)	167 (16.75%)	<0.001
Left school <14y	1101 (61.41%)	875 (63.13%)	598 (68.11%)		1067 (58.50%)	795 (64.37%)	677 (67.90%)
Missing	241 (13.44%)	224 (16.16%)	145 (16.51%)		270 (14.80%)	188 (15.22%)	153 (15.35%)
Mother’s level of education (birth)
Left school ≥15y	516 (28.78%)	325 (23.45%)	182 (20.73%)	<0.001	552 (30.26%)	300 (24.29%)	173 (17.35%)	<0.001
Left school <14y	1189 (66.31%)	971 (70.06%)	640 (72.89%)		1172 (64.25%)	858 (69.47%)	766 (76.83%)
Missing	88 (4.91%)	90 (6.49%)	56 (6.38%)		100 (5.48%)	77 (6.23%)	58 (5.82%)
Reading activities (7y)
Every week	920 (51.31%)	687 (49.57%)	423 (48.18%)	0.263	893 (48.96%)	590 (47.77%)	483 (48.45%)	0.832
Occasionally	488 (27.22%)	355 (25.61%)	240 (27.33%)		548 (30.04%)	389 (31.50%)	304 (30.49%)
Hardly ever	173 (9.65%)	142 (10.25%)	94 (10.71%)		168 (9.21%)	98 (7.94%)	90 (9.03%)
Missing	212 (11.82%)	202 (14.57%)	121 (13.78%)		215 (11.79%)	158 (12.79%)	120 (12.04%)
Outdoor activities (7y)
Most weeks	1402 (78.19%)	1017 (73.38%)	657 (74.83%)	0.032	1449 (79.44%)	951 (77.00%)	772 (77.43%)	0.41
Occasionally/hardly ever	182 (10.15%)	169 (12.19%)	101 (11.50%)		161 (8.83%)	130 (10.53%)	105 (10.53%)
Missing	209 (11.66%)	200 (14.43%)	120 (13.67%)		214 (11.73%)	154 (12.47%)	120 (12.04%)
Place in the sibling (7 y)
Single child	115 (6.41%)	103 (7.43%)	76 (8.66%)	0.024	119 (6.52%)	107 (8.66%)	79 (7.92%)	0.344
Eldest	511 (28.50%)	347 (25.04%)	210 (23.92%)		481 (26.37%)	317 (25.67%)	273 (27.38%)
≥ 2	961 (53.60%)	741 (53.46%)	476 (54.21%)		1015 (55.65%)	661 (53.52%)	528 (52.96%)
Missing	206 (11.49%)	195 (14.07%)	116 (13.21%)		209 (11.46%)	150 (12.15%)	117 (11.74%)
ACEs (7-16y)
None	1285 (71.67%)	891 (64.29%)	545 (62.07%)	<0.001	1322 (72.48%)	827 (66.96%)	626 (62.79%)	<0.001
One or more	399 (22.25%)	374 (26.98%)	259 (29.50%)		389 (21.33%)	320 (25.91%)	298 (29.89%)
Missing	109 (6.08%)	121 (8.73%)	74 (8.43%)		113 (6.20%)	88 (7.13%)	73 (7.32%)
Health problems in childhood (7-16y)
No	1360 (75.85%)	1002 (72.29%)	626 (71.30%)	0.014	1431 (78.45%)	923 (74.74%)	748 (75.03%)	0.090
Yes	416 (23.20%)	376 (27.13%)	248 (28.25%)		383 (21.00%)	302 (24.45%)	244 (24.47%)
Missing	17 (0.95%)	8 (0.58%)	4 (0.46%)		10 (0.55%)	10 (0.81%)	5 (0.50%)
Copy-a-Design test (7y)
Score: med [p25-p75]	8[6–9]	7[6–9]	7[6–8]	<0.001	8[6–9]	7[6–8]	7[6–8]	<0.001
Missing	183 (10.21%)	180 (12.99%)	110 (12.53%)		188 (10.31%)	131 (10.61%)	109 (10.93%)
Intermediate life course variables
Education level (23y)
A level	468 (26.10%)	270 (19.48%)	117 (13.33%)	<0.001	465 (25.49%)	215 (17.41%)	137 (13.74%)	<0.001
O level	601 (33.52%)	453 (32.68%)	270 (30.75%)		746 (40.90%)	498 (40.32%)	354 (35.51%)
No level	471 (26.27%)	456 (32.90%)	353 (40.21%)		409 (22.42%)	366 (29.64%)	365 (36.61%)
Missing	253 (14.11%)	207 (14.94%)	138 (15.72%)		204 (11.18%)	156 (12.63%)	141 (14.14%)
Malaise inventory (23y)
No psychological distress	1497 (83.49%)	1134 (81.82%)	705 (80.30%)	0.156	1508 (82.68%)	951 (77.00%)	737 (73.92%)	<0.001
Psychological distress	42 (2.34%)	45 (3.25%)	33 (3.76%)		109 (5.98%)	127 (10.28%)	121 (12.14%)
Missing	254 (14.17%)	207 (14.94%)	140 (15.95%)		207 (11.35%)	157 (12.71%)	139 (13.94%)
Sense of personal control (33y)
Internal	1387 (77.36%)	1018 (73.45%)	628 (71.53%)	0.012	1440 (78.95%)	960 (77.73%)	740 (74.22%)	0.037
External	132 (7.36%)	114 (8.23%)	78 (8.88%)		187 (10.25%)	129 (10.45%)	135 (13.54%)
Missing	274 (15.28%)	254 (18.33%)	172 (19.59%)		197 (10.80%)	146 (11.82%)	122 (12.24%)
Occupational social class (33y)
I & II	713 (39.77%)	460 (33.19%)	245 (27.90%)	<0.001	609 (33.39%)	339 (27.45%)	225 (22.57%)	<0.001
IIINM	178 (9.93%)	113 (8.15%)	68 (7.74%)		553 (30.32%)	393 (31.82%)	306 (30.69%)
IIIM	442 (24.65%)	396 (28.57%)	265 (30.18%)		108 (5.92%)	77 (6.23%)	70 (7.02%)
IV&V	186 (10.37%)	163 (11.76%)	134 (15.26%)		289 (15.84%)	242 (19.60%)	217 (21.77%)
Missing	274 (15.28%)	254 (18.33%)	166 (18.91%)		265 (14.53%)	184 (14.90%)	179 (17.95%)
Wealth (33y)
Owner highest price	370 (20.64%)	203 (14.65%)	87 (9.91%)	<0.001	395 (21.66%)	186 (15.06%)	118 (11.84%)	<0.001
Owner high price	356 (19.85%)	220 (15.87%)	109 (12.41%)		350 (19.19%)	203 (16.44%)	129 (12.94%)
Owner median price	293 (16.34%)	232 (16.74%)	141 (16.06%)		329 (18.04%)	234 (18.95%)	156 (15.65%)
Owner lowest price	281 (15.67%)	224 (16.16%)	169 (19.25%)		281 (15.41%)	200 (16.19%)	182 (18.25%)
Not owner	273 (15.23%)	293 (21.14%)	231 (26.31%)		274 (15.02%)	270 (21.86%)	299 (29.99%)
Missing	220 (12.27%)	214 (15.44%)	141 (16.06%)		195 (10.69%)	142 (11.50%)	113 (11.33%)
Smoking (42y)
Non-smoker	894 (49.86%)	572 (41.27%)	297 (33.83%)	<0.001	885 (48.52%)	530 (42.91%)	398 (39.92%)	<0.001
Ex smoker	510 (28.44%)	328 (23.67%)	200 (22.78%)		506 (27.74%)	295 (23.89%)	184 (18.46%)
Smoker < 10 cig./day	130 (7.25%)	108 (7.79%)	51 (5.81%)		143 (7.84%)	85 (6.88%)	65 (6.52%)
Smoker 10 to 19 cig./day	90 (5.02%)	116 (8.37%)	99 (11.28%)		134 (7.35%)	166 (13.44%)	118 (11.84%)
Smoker more than 20 cig./day	116 (6.47%)	203 (14.65%)	200 (22.78%)		106 (5.81%)	129 (10.45%)	193 (19.36%)
Missing	53 (2.96%)	59 (4.26%)	31 (3.53%)		50 (2.74%)	30 (2.43%)	39 (3.91%)
Alcohol consumption (42y)
Moderate	1078 (60.12%)	730 (52.67%)	375 (42.71%)	<0.001	1249 (68.48%)	769 (62.27%)	542 (54.36%)	<0.001
Abstinent	237 (13.22%)	233 (16.81%)	192 (21.87%)		416 (22.81%)	364 (29.47%)	346 (34.70%)
High	425 (23.70%)	365 (26.33%)	280 (31.89%)		109 (5.98%)	72 (5.83%)	70 (7.02%)
Missing	53 (2.96%)	58 (4.18%)	31 (3.53%)		50 (2.74%)	30 (2.43%)	39 (3.91%)
Physical activity (42y)
Active	1253 (69.88%)	877 (63.28%)	487 (55.47%)	<0.001	1252 (68.64%)	793 (64.21%)	557 (55.87%)	<0.001
Moderate	168 (9.37%)	122 (8.80%)	90 (10.25%)		124 (6.80%)	103 (8.34%)	71 (7.12%)
Inactive	318 (17.74%)	329 (23.74%)	270 (30.75%)		398 (21.82%)	308 (24.94%)	330 (33.10%)
Missing	54 (3.01%)	58 (4.18%)	31 (3.53%)		50 (2.74%)	31 (2.51%)	39 (3.91%)

Bivariate analyses by parental interest are reported in Table 2. Parental interest was associated with all childhood variables, and all intermediate variables. In relation to our hypothesis on the educational and others pathways, women and men who had a low level of education, psychological distress at age 23, an external sense of personal control at age 33, a low social position at age 33 (occupational class and wealth), and who were smokers, heavy alcohol drinkers or abstainers and had low physical activity, were more likely to have had parents who were described as uninterested or not very interested in their children’s education by the teacher.

Table 2

Bivariate statistics on the complete-case sample (n = 8 113): Parental interest categories according to confounding and intermediate variables in (A) men and in (B) women.

	Parental interest (7-16y)
	A. Men n(%) n = 4 075 (50%)				B. Women n(%) n = 4 056 (50%)
	Both interested n(%)	Low/No interest n(%)	Missing n(%)	P-value*	Both interested n(%)	Low/No interest n(%)	Missing n(%)	P-value*
	1912 (47.13%)	1703 (41.98%)	442 (10.89%)		1906 (46.99%)	1666 (41.07%)	484 (11.93%)
Allostatic load (44y)
Low	930 (48.64%)	698 (40.99%)	165 (37.33%)	<0.001	989 (51.89%)	654 (39.26%)	181 (37.40%)	<0.001
Medium	634 (33.16%)	578 (33.94%)	174 (39.37%)		555 (29.12%)	524 (31.45%)	156 (32.23%)
High	348 (18.20%)	427 (25.07%)	103 (23.30%)		362 (18.99%)	488 (29.29%)	147 (30.37%)
Prior confounders
Parental SEC (birth)
I & II	564 (29.50%)	135 (7.93%)	49 (11.09%)	<0.001	526 (27.60%)	132 (7.92%)	48 (9.92%)	<0.001
IIINM	223 (11.66%)	128 (7.52%)	29 (6.56%)		238 (12.49%)	119 (7.14%)	34 (7.02%)
IIIM	801 (41.89%)	905 (53.14%)	207 (46.83%)		788 (41.34%)	876 (52.58%)	247 (51.03%)
IV&V	229 (11.98%)	441 (25.90%)	98 (22.17%)		250 (13.12%)	445 (26.71%)	102 (21.07%)
Missing	95 (4.97%)	94 (5.52%)	59 (13.35%)		104 (5.46%)	94 (5.64%)	53 (10.95%)
Material living conditions (7y)
Advantaged	1375 (71.91%)	930 (54.61%)	190 (42.99%)	<0.001	1342 (70.41%)	902 (54.14%)	219 (45.25%)	<0.001
Disadvantaged	292 (15.27%)	528 (31.00%)	118 (26.70%)		315 (16.53%)	548 (32.89%)	144 (29.75%)
Missing	245 (12.81%)	245 (14.39%)	134 (30.32%)		249 (13.06%)	216 (12.97%)	121 (25.00%)
Father’s level of education (7y)
Left school ≥15y	655 (34.26%)	178 (10.45%)	40 (9.05%)	<0.001	657 (34.47%)	198 (11.88%)	51 (10.54%)	<0.001
Left school <14y	1056 (55.23%)	1268 (74.46%)	250 (56.56%)		1023 (53.67%)	1237 (74.25%)	279 (57.64%)
Missing	201 (10.51%)	257 (15.09%)	152 (34.39%)		226 (11.86%)	231 (13.87%)	154 (31.82%)
Mother’s level of education (birth)
Left school ≥15y	705 (36.87%)	244 (14.33%)	74 (16.74%)	<0.001	725 (38.04%)	218 (13.09%)	82 (16.94%)	<0.001
Left school <14y	1111 (58.11%)	1376 (80.80%)	313 (70.81%)		1084 (56.87%)	1359 (81.57%)	353 (72.93%)
Missing	96 (5.02%)	83 (4.87%)	55 (12.44%)		97 (5.09%)	89 (5.34%)	49 (10.12%)
Reading activities (7y)
Every week	1126 (58.89%)	747 (43.86%)	157 (35.52%)	<0.001	1076 (56.45%)	715 (42.92%)	175 (36.16%)	<0.001
Occasionally	459 (24.01%)	514 (30.18%)	110 (24.89%)		509 (26.71%)	586 (35.17%)	146 (30.17%)
Hardly ever	150 (7.85%)	224 (13.15%)	35 (7.92%)		134 (7.03%)	183 (10.98%)	39 (8.06%)
Missing	177 (9.26%)	218 (12.80%)	140 (31.67%)		187 (9.81%)	182 (10.92%)	124 (25.62%)
Outdoor activities (7y)
Most weeks	1615 (84.47%)	1202 (70.58%)	259 (58.60%)	<0.001	1599 (83.89%)	1257 (75.45%)	316 (65.29%)	<0.001
Occasionally/hardly ever	125 (6.54%)	283 (16.62%)	44 (9.95%)		128 (6.72%)	223 (13.39%)	45 (9.30%)
Missing	172 (9.00%)	218 (12.80%)	139 (31.45%)		179 (9.39%)	186 (11.16%)	123 (25.41%)
Place in the sibling (7 y)
Single child	155 (8.11%)	99 (5.81%)	40 (9.05%)	<0.001	172 (9.02%)	89 (5.34%)	44 (9.09%)	<0.001
Eldest	612 (32.01%)	377 (22.14%)	79 (17.87%)		596 (31.27%)	377 (22.63%)	98 (20.25%)
≥ 2	975 (50.99%)	1017 (59.72%)	186 (42.08%)		965 (50.63%)	1021 (61.28%)	218 (45.04%)
Missing	170 (8.89%)	210 (12.33%)	137 (31.00%)		173 (9.08%)	179 (10.74%)	124 (25.62%)
ACEs (7-16y)
None	1534 (80.23%)	1022 (60.01%)	165 (37.33%)	<0.001	1531 (80.33%)	1037 (62.24%)	207 (42.77%)	<0.001
One or more	287 (15.01%)	566 (33.24%)	179 (40.50%)		267 (14.01%)	541 (32.47%)	199 (41.12%)
Missing	91 (4.76%)	115 (6.75%)	98 (22.17%)		108 (5.67%)	88 (5.28%)	78 (16.12%)
Health problems in childhood (7-16y)
No	1415 (74.01%)	1260 (73.99%)	313 (70.81%)	<0.001	1495 (78.44%)	1254 (75.27%)	353 (72.93%)	<0.001
Yes	494 (25.84%)	435 (25.54%)	111 (25.11%)		409 (21.46%)	406 (24.37%)	114 (23.55%)
Missing	3 (0.16%)	8 (0.47%)	18 (4.07%)		2 (0.10%)	6 (0.36%)	17 (3.51%)
Copy-a-Design test (7y)
Score: med [p25-p75]	8[6–9]	7[6–8]	7[6–8]	<0.001	8[6–9]	7[6–8]	7[6–8]	<0.001
Missing	145 (30.66%)	183 (38.69%)	145 (30.66%)		151 (35.28%)	157 (36.68%)	120 (28.04%)
Intermediate life course variables
Education level (23y)
A level	685 (35.83%)	120 (7.05%)	50 (11.31%)	<0.001	662 (34.73%)	96 (5.76%)	59 (12.19%)	<0.001
O level	687 (35.93%)	515 (30.24%)	122 (27.60%)		811 (42.55%)	621 (37.27%)	166 (34.30%)
No level	289 (15.12%)	807 (47.39%)	184 (41.63%)		239 (12.54%)	724 (43.46%)	177 (36.57%)
Missing	251 (13.13%)	261 (15.33%)	86 (19.46%)		194 (10.18%)	225 (13.51%)	82 (16.94%)
Malaise inventory (23y)
No psychological distress	1623 (84.88%)	1374 (80.68%)	339 (76.70%)	<0.001	1610 (84.47%)	1237 (74.25%)	349 (72.11%)	<0.001
Psychological distress	36 (1.88%)	67 (3.93%)	17 (3.85%)		103 (5.40%)	201 (12.06%)	53 (10.95%)
Missing	253 (13.23%)	262 (15.38%)	86 (19.46%)		193 (10.13%)	228 (13.69%)	82 (16.94%)
Sense of personal control (33y)
Internal	1513 (79.13%)	1218 (71.52%)	302 (68.33%)	<0.001	1581 (82.95%)	1222 (73.35%)	337 (69.63%)	<0.001
External	116 (6.07%)	169 (9.92%)	39 (8.82%)		146 (7.66%)	231 (13.87%)	74 (15.29%)
Missing	283 (14.80%)	316 (18.56%)	101 (22.85%)		179 (9.39%)	213 (12.79%)	73 (15.08%)
Occupational social class (33y)
I & II	926 (48.43%)	378 (22.20%)	114 (25.79%)	<0.001	745 (39.09%)	303 (18.19%)	125 (25.83%)	<0.001
IIINM	207 (10.83%)	119 (6.99%)	33 (7.47%)		593 (31.11%)	505 (30.31%)	154 (31.82%)
IIIM	363 (18.99%)	608 (35.70%)	132 (29.86%)		98 (5.14%)	122 (7.32%)	35 (7.23%)
IV&V	123 (6.43%)	288 (16.91%)	72 (16.29%)		220 (11.54%)	439 (26.35%)	89 (18.39%)
Missing	293 (15.32%)	310 (18.20%)	91 (20.59%)		250 (13.12%)	297 (17.83%)	81 (16.74%)
Wealth (33y)
Owner highest price	422 (22.07%)	178 (10.45%)	60 (13.57%)	<0.001	447 (23.45%)	189 (11.34%)	63 (13.02%)	<0.001
Owner high price	400 (20.92%)	223 (13.09%)	62 (14.03%)		401 (21.04%)	219 (13.15%)	62 (12.81%)
Owner median price	318 (16.63%)	287 (16.85%)	61 (13.80%)		366 (19.20%)	261 (15.67%)	92 (19.01%)
Owner lowest price	250 (13.08%)	351 (20.61%)	73 (16.52%)		239 (12.54%)	335 (20.11%)	89 (18.39%)
Not owner	287 (15.01%)	404 (23.72%)	106 (23.98%)		268 (14.06%)	458 (27.49%)	117 (24.17%)
Missing	235 (12.29%)	260 (15.27%)	80 (18.10%)		185 (9.71%)	204 (12.24%)	61 (12.60%)
Smoking (42y)
Non-smoker	952 (49.79%)	631 (37.05%)	180 (40.72%)	<0.001	1000 (52.47%)	619 (37.15%)	194 (40.08%)	<0.001
Ex smoker	468 (24.48%)	457 (26.83%)	113 (25.57%)		480 (25.18%)	384 (23.05%)	121 (25.00%)
Smoker < 10 cig./day	154 (8.05%)	107 (6.28%)	28 (6.33%)		133 (6.98%)	134 (8.04%)	26 (5.37%)
Smoker 10 to 19 cig./day	113 (5.91%)	159 (9.34%)	33 (7.47%)		138 (7.24%)	228 (13.69%)	52 (10.74%)
Smoker more than 20 cig./day	168 (8.79%)	286 (16.79%)	65 (14.71%)		106 (5.56%)	249 (14.95%)	73 (15.08%)
Missing	57 (2.98%)	63 (3.70%)	23 (5.20%)		49 (2.57%)	52 (3.12%)	18 (3.72%)
Alcohol consumption (42y)
Moderate	1143 (59.78%)	821 (48.21%)	219 (49.55%)	<0.001	1312 (68.84%)	975 (58.52%)	273 (56.40%)	<0.001
Abstinent	232 (12.13%)	320 (18.79%)	110 (24.89%)		421 (22.09%)	538 (32.29%)	167 (34.50%)
High	480 (25.10%)	500 (29.36%)	90 (20.36%)		124 (6.51%)	101 (6.06%)	26 (5.37%)
Missing	57 (2.98%)	62 (3.64%)	23 (5.20%)		49 (2.57%)	52 (3.12%)	18 (3.72%)
Physical activity (42y)
Active	1305 (68.25%)	1036 (60.83%)	276 (62.44%)	<0.001	1282 (67.26%)	1026 (61.58%)	294 (60.74%)	<0.001
Moderate	207 (10.83%)	134 (7.87%)	39 (8.82%)		155 (8.13%)	112 (6.72%)	31 (6.40%)
Inactive	342 (17.89%)	471 (27.66%)	104 (23.53%)		420 (22.04%)	475 (28.51%)	141 (29.13%)
Missing	58 (3.03%)	62 (3.64%)	23 (5.20%)		49 (2.57%)	53 (3.18%)	18 (3.72%)

Abbreviations and symbols: n = number of people; med = median; p25 = 25e percentile; p75 = 75e percentile; statistically significant results at the 5% threshold are in bold. Values corresponding to the categories of allostatic load: Low: [0–2]; Medium: [3–4]; High: [5–12].

*P-values were calculated using the chi-squared test for categorical variables and Wilcoxon signed-rank test for the continuous variable.

The multivariable results of the association between parental interest and allostatic load, examining the a priori set of confounding and intermediate factors are presented in Tables 3 and 4, for men and women respectively. Men with parents perceived as uninterested or not very interested in their child’s education had higher allostatic load scores at 44 years compared to those perceived as interested by the school teacher (Model 1, β = 0.41 [0.29; 0.54]). After adjustment for prior confounders, the link between parental interest and allostatic load was weakened (Model 2, β = 0.18 [0.03; 0.32]) partly attributable to parental SEC, ACEs, health problems in childhood and cognitive skills. The associations between parental interest and allostatic load were rendered insignificant after including educational attainment in Model 3 (Model 3, β = 0.06 [-0.09; 0.21]). Further adjusting for psychological status at age 23 (Model 4) and occupational social class and wealth (Model 5) and health behaviors (Model 6) did not change the result patterns.

Table 3

Life course multivariable linear regression between allostatic load and parental interest using data obtained from multiple imputation for men (n = 4 057).

		Model 1		Model 2		Model 3		Model 4		Model 5		Model 6
		Coeff. [CI 95%]	P-value	Coeff. [CI 95%]	P-value	Coeff. [CI 95%]	P-value	Coeff. [CI 95%]	P-value	Coeff. [CI 95%]	P-value	Coeff. [CI 95%]	P-value
Parental interest
Both interested		ref		ref		ref		ref		ref		ref
Low/No interest		0.41 [0.29; 0.54]	<0.001	0.18 [0.03; 0.32]	0.015	0.06 [-0.09; 0.21]	0.439	0.05 [-0.10; 0.21]	0.475	0.02 [-0.13; 0.17]	0.823	-0.01 [-0.16; 0.14]	0.866
Parental SEC (birth)
I & II				ref		ref		ref		ref		ref
IIINM				0.04 [-0.2; 0.28]	0.732	0.02 [-0.22; 0.26]	0.867	0.02 [-0.22; 0.26]	0.845	0.05 [-0.19; 0.29]	0.693	0.02 [-0.21; 0.26]	0.836
IIIM				0.32 [0.12; 0.51]	0.001	0.26 [0.06; 0.45]	0.009	0.26 [0.07; 0.46]	0.008	0.26 [0.07; 0.45]	0.008	0.22 [0.03; 0.41]	0.022
IV&V				0.40 [0.17; 0.63]	0.001	0.32 [0.09; 0.55]	0.006	0.32 [0.10; 0.55]	0.005	0.29 [0.06; 0.51]	0.013	0.25 [0.03; 0.47]	0.029
Material living conditions (7y)
Advantaged				ref		ref		ref		ref		ref
Disadvantaged				0.13 [-0.04; 0.30]	0.121	0.13 [-0.04; 0.29]	0.132	0.13 [-0.04; 0.29]	0.139	0.10 [-0.07; 0.27]	0.227	0.10 [-0.06; 0.26]	0.236
Father’s level of education (7y)
Left school ≥15y				ref		ref		ref		ref		ref
Left school <14y				0.08 [-0.09; 0.26]	0.366	0.01 [-0.16; 0.19]	0.872	0.01 [-0.16; 0.19]	0.875	0.002 [-0.17; 0.18]	0.983	0.04 [-0.14; 0.21]	0.689
Mother’s level of education (birth)
Left school ≥15y				ref		ref		ref		ref		ref
Left school <14y				0.11 [-0.04; 0.27]	0.156	0.06 [-0.10; 0.22]	0.448	0.06 [-0.09; 0.22]	0.443	0.05 [-0.11; 0.20]	0.572	0.06 [-0.10; 0.21]	0.465
Reading activities (7y)
Every week				ref		ref		ref		ref		ref
Occasionally				-0.02 [-0.17; 0.13]	0.771	-0.04 [-0.19; 0.11]	0.575	-0.04 [-0.19; 0.11]	0.584	-0.03 [-0.18; 0.12]	0.732	-0.02 [-0.17; 0.12]	0.770
Hardly ever				0.08 [-0.15; 0.31]	0.475	0.08 [-0.15; 0.31]	0.502	0.08 [-0.15; 0.31]	0.506	0.09 [-0.14; 0.32]	0.454	0.06 [-0.16; 0.29]	0.597
Outdoor activities (7y)
Most weeks				ref		ref		ref		ref		ref
Occasionally/hardly ever				-0.12 [-0.33; 0.09]	0.261	-0.13 [-0.34; 0.07]	0.209	-0.13 [-0.34; 0.07]	0.204	-0.15 [-0.36; 0.06]	0.151	-0.14 [-0.34; 0.06]	0.169
Place in the sibling (7 y)
Single child				ref		ref		ref		ref		ref
Elder				-0.19 [-0.44; 0.07]	0.152	-0.20 [-0.46; 0.05]	0.118	-0.20 [-0.45; 0.05]	0.123	-0.2 [-0.45; 0.05]	0.119	-0.19 [-0.44; 0.06]	0.139
≥ 2				-0.15 [-0.38; 0.09]	0.215	-0.19 [-0.42; 0.05]	0.119	-0.19 [-0.42; 0.05]	0.117	-0.2 [-0.43; 0.03]	0.095	-0.20 [-0.43; 0.03]	0.092
ACEs (7-16y)
None				ref		ref		ref		ref		ref
One or more				0.20 [0.06; 0.34]	0.005	0.17 [0.03; 0.31]	0.021	0.16 [0.02; 0.30]	0.026	0.13 [-0.01; 0.27]	0.068	0.06 [-0.08; 0.20]	0.432
Health problems in childhood (7-16y)
No				ref		ref		ref		ref		ref
Yes				0.19 [0.06; 0.33]	0.006	0.18 [0.04; 0.32]	0.010	0.17 [0.04; 0.31]	0.013	0.14 [0.002; 0.28]	0.046	0.13 [-0.005; 0.26]	0.059
Copy-a-Design test (7y)
Score: med [p25-p75]				-0.06 [-0.09; -0.02]	0.001	-0.04 [-0.08; -0.01]	0.022	-0.04 [-0.08; -0.01]	0.023	-0.03 [-0.07; 0.0002]	0.051	-0.03 [-0.06; 0.004]	0.081
Education level (23y)
A level						ref		ref		ref		ref
O level						0.26 [0.09; 0.43]	0.002	0.26 [0.09; 0.43]	0.003	0.21 [0.03; 0.39]	0.022	0.12 [-0.05; 0.29]	0.179
No level						0.54 [0.34; 0.74]	<0.001	0.53 [0.33; 0.73]	<0.001	0.41 [0.19; 0.63]	<0.001	0.24 [0.02; 0.46]	0.029
Malaise inventory (23y)
No psychological distress								ref		ref		ref
Psychological distress								0.09 [-0.25; 0.43]	0.610	0.06 [-0.28; 0.40]	0.730	-0.03 [-0.37; 0.31]	0.851
Sense of personal control (33y)
Internal								ref		ref		ref
External								0.13 [-0.09; 0.35]	0.257	0.06 [-0.17; 0.29]	0.605	-0.02 [-0.24; 0.21]	0.873
Occupational social class (33y)
I & II										ref		ref
IIINM										-0.12 [-0.33; 0.09]	0.253	-0.11 [-0.32; 0.09]	0.265
IIIM										0.01 [-0.17; 0.18]	0.932	-0.05 [-0.22; 0.12]	0.571
IV&V										0.03 [-0.19; 0.24]	0.818	-0.04 [-0.26; 0.18]	0.745
Wealth (33y)
Owner highest price										ref		ref
Owner high price										0.06 [-0.14; 0.26]	0.546	0.06 [-0.14; 0.26]	0.548
Owner median price										0.29 [0.06; 0.51]	0.012	0.23 [0.01; 0.45]	0.038
Owner lowest price										0.35 [0.12; 0.57]	0.003	0.24 [0.02; 0.46]	0.030
Not owner										0.65 [0.44; 0.86]	<0.001	0.45 [0.23; 0.66]	<0.001
Smoking (42y)
Non-smoker												ref
Ex smoker												0.07 [-0.08; 0.22]	0.338
Smoker < 10 cig./day												0.08 [-0.16; 0.31]	0.532
Smoker 10 to 19 cig./day												0.68 [0.44; 0.92]	<0.001
Smoker more than 20 cig./day												0.93 [0.73; 1.12]	<0.001
Alcohol consumption (42y)
Moderate												ref
Abstinent												0.33 [0.16; 0.5]	<0.001
High												0.25 [0.11; 0.39]	0.001
Physical activity (42y)
Active												ref
Moderate												0.10 [-0.10; 0.30]	0.334
Inactive												0.30 [0.16; 0.45]	<0.001

Table 4

Life course multivariable linear regression between allostatic load and parental interest using data obtained from multiple imputation for women (n = 4 056).

	Model 1		Model 2		Model 3		Model 4		Model 5		Model 6
	Coeff. [CI 95%]	P-value	Coeff. [CI 95%]	P-value	Coeff. [CI 95%]	P-value	Coeff. [CI 95%]	P-value	Coeff. [CI 95%]	P-value	Coeff. [CI 95%]	P-value
Parental interest
Both interested	ref		ref		ref		ref		ref		ref
Low/No interest	0.69 [0.54; 0.83]	<0.001	0.40 [0.24; 0.56]	<0.001	0.28 [0.10; 0.45]	0.002	0.26 [0.08; 0.43]	0.004	0.21 [0.04; 0.39]	0.015	0.18 [0.01; 0.35]	0.041
Parental SEC (birth)
I & II			ref		ref		ref		ref		ref
IIINM			0.17 [-0.10; 0.44]	0.219	0.16 [-0.12; 0.43]	0.262	0.16 [-0.11; 0.43]	0.242	0.17 [-0.10; 0.43]	0.224	0.14 [-0.13; 0.40]	0.316
IIIM			0.42 [0.21; 0.63]	<0.001	0.38 [0.17; 0.59]	<0.001	0.38 [0.17; 0.59]	0.001	0.34 [0.13; 0.56]	0.001	0.32 [0.11; 0.53]	0.003
IV&V			0.50 [0.26; 0.75]	<0.001	0.44 [0.19; 0.69]	<0.001	0.44 [0.19; 0.68]	0.001	0.37 [0.12; 0.61]	0.004	0.32 [0.08; 0.57]	0.009
Material living conditions (7y)
Advantaged			ref		ref		ref		ref		ref
Disadvantaged			0.07 [-0.09; 0.23]	0.382	0.07 [-0.09; 0.23]	0.407	0.06 [-0.10; 0.22]	0.481	0.03 [-0.13; 0.19]	0.706	0.001 [-0.16; 0.16]	0.99
Father’s level of education (7y)
Left school ≥15y			ref		ref		ref		ref		ref
Left school <14y			0.07 [-0.11; 0.25]	0.447	0.02 [-0.16; 0.20]	0.801	0.02 [-0.16; 0.20]	0.827	-0.02 [-0.19; 0.16]	0.856	-0.03 [-0.21; 0.15]	0.747
Mother’s level of education (birth)
Left school ≥15y			ref		ref		ref		ref		ref
Left school <14y			0.28 [0.11; 0.45]	0.001	0.23 [0.06; 0.41]	0.008	0.24 [0.06; 0.41]	0.007	0.23 [0.06; 0.41]	0.008	0.23 [0.06; 0.40]	0.008
Reading activities (7y)
Every week			ref		ref		ref		ref		ref
Occasionally			-0.12 [-0.27; 0.03]	0.125	-0.12 [-0.27; 0.03]	0.116	-0.12 [-0.27; 0.03]	0.115	-0.12 [-0.27; 0.03]	0.123	-0.13 [-0.28; 0.02]	0.090
Hardly ever			-0.07 [-0.30; 0.17]	0.570	-0.08 [-0.31; 0.16]	0.523	-0.08 [-0.31; 0.15]	0.493	-0.08 [-0.31; 0.15]	0.491	-0.11 [-0.34; 0.12]	0.337
Outdoor activities (7y)
Most weeks			ref		ref		ref		ref		ref
Occasionally/hardly ever			0.09 [-0.14; 0.33]	0.432	0.08 [-0.15; 0.31]	0.521	0.06 [-0.17; 0.29]	0.594	0.04 [-0.18; 0.27]	0.700	0.04 [-0.18; 0.27]	0.703
Place in the sibling (7 y)
Single child			ref		ref		ref		ref		ref
Elder			-0.07 [-0.33; 0.20]	0.625	-0.08 [-0.35; 0.18]	0.531	-0.08 [-0.34; 0.18]	0.538	-0.08 [-0.34; 0.18]	0.539	-0.10 [-0.36; 0.15]	0.422
≥ 2			-0.20 [-0.47; 0.06]	0.130	-0.24 [-0.51; 0.02]	0.072	-0.25 [-0.51; 0.01]	0.064	-0.25 [-0.51; 0.01]	0.063	-0.26 [-0.52; -0.004]	0.047
ACEs (7-16y)
None			ref		ref		ref		ref		ref
One or more			0.19 [0.03; 0.35]	0.017	0.16 [0.002; 0.31]	0.046	0.14 [-0.012; 0.30]	0.071	0.10 [-0.06; 0.26]	0.219	0.05 [-0.11; 0.21]	0.544
Health problems in childhood (7-16y)
No			ref		ref		ref		ref		ref
Yes			0.14 [-0.02; 0.29]	0.086	0.12 [-0.04; 0.27]	0.134	0.11 [-0.04; 0.26]	0.164	0.10 [-0.05; 0.25]	0.198	0.08 [-0.07; 0.23]	0.300
Copy-a-Design test (7y)
Score: med [p25-p75]			-0.09 [-0.13; -0.05]	<0.001	-0.07 [-0.11; -0.03]	0.001	-0.06 [-0.10; -0.02]	0.002	-0.06 [-0.09; -0.02]	0.005	-0.05 [-0.09; -0.01]	0.008
Education level (23y)
A level					ref		ref		ref		ref
O level					0.15 [-0.05; 0.35]	0.145	0.14 [-0.05; 0.34]	0.154	0.08 [-0.14; 0.29]	0.482	0.05 [-0.16; 0.26]	0.658
No level					0.52 [0.28; 0.75]	<0.001	0.49 [0.26; 0.73]	<0.001	0.31 [0.05; 0.56]	0.019	0.20 [-0.06; 0.45]	0.139
Malaise inventory (23y)
No psychological distress							ref		ref		ref
Psychological distress							0.37 [0.13; 0.61]	0.003	0.35 [0.11; 0.59]	0.004	0.26 [0.02; 0.50]	0.032
Sense of personal control (33y)
Internal							ref		ref		ref
External							0.09 [-0.14; 0.32]	0.429	-0.03 [-0.25; 0.20]	0.819	-0.10 [-0.32; 0.12]	0.387
Occupational social class (33y)
I & II									ref		ref
IIINM									0.12 [-0.06; 0.31]	0.182	0.14 [-0.04; 0.32]	0.128
IIIM									0.08 [-0.21; 0.37]	0.578	0.04 [-0.24; 0.32]	0.784
IV&V									0.13 [-0.09; 0.35]	0.245	0.11 [-0.11; 0.33]	0.324
Wealth (33y)
Owner highest price									ref		ref
Owner high price									0.05 [-0.17; 0.27]	0.635	0.05 [-0.16; 0.27]	0.624
Owner median price									0.14 [-0.09; 0.37]	0.236	0.12 [-0.10; 0.35]	0.291
Owner lowest price									0.33 [0.10; 0.57]	0.006	0.25 [0.01; 0.48]	0.039
Not owner									0.74 [0.51; 0.97]	<0.001	0.55 [0.31; 0.78]	<0.001
Smoking (42y)
Non-smoker											ref
Ex smoker											-0.20 [-0.36; -0.03]	0.018
Smoker < 10 cig./day											-0.14 [-0.40; 0.12]	0.286
Smoker 10 to 19 cig./day											0.30 [0.07; 0.53]	0.009
Smoker more than 20 cig./day											0.86 [0.62; 1.09]	<0.001
Alcohol consumption (42y)
Moderate											ref
Abstinent											0.31 [0.16; 0.46]	<0.001
High											0.05 [-0.21; 0.32]	0.698
Physical activity (42y)
Active											ref
Moderate											0.14 [-0.11; 0.39]	0.267
Inactive											0.27 [0.12; 0.43]	0.001

A similar pattern was observed for women: women whose parents were perceived as uninterested or not very interested by the teacher had higher allostatic load scores at 44 years (Model 1, β = 0.69 [0.54; 0.83]). The association between parental interest and allostatic load was attenuated after controlling for early life confounder (Model 2, β = 0.40 [0.24; 0.56]) partly explained by parental SEC, mother’s level of education and ACEs and cognitive skills. Further adjustment for educational attainment reduced the strength of the association (Model 3, β = 0.28 [0.10; 0.45]). When psychological status was accounted for, the association between parental interest and allostatic load was marginally affected (Model 4, β = 0,26 [0.08; 0.43]) but was explained by the malaise inventory. Further adjustment for occupational social class and wealth lightly attenuated the association with income affecting it more strongly (Model 5, β = 0.21 [0.04; 0.39]) as well as health behaviors (Model 6, β = 0.18 [0.01; 0.35]). When all potential mediators were controlled for, parental interest remained significantly associated with allostatic load score (Model 6, β = 0.18 [0.01; 0.35]).

The analyses of the direct and indirect effects of parental interest on allostatic load are presented in Figs 2 and 3. For men, the direct link between parental interest and allostatic load was completely mediated, mainly by the educational pathway (67% of the total indirect effect) but also through other intermediate factors (33% of the total indirect effect).

Fig 2

Direct and indirect effect results between parental interest and allostatic load obtained from multiple imputation for men (n = 4057).

Fig 3

Direct and indirect effect results between parental interest and allostatic load obtained from multiple imputation for women (n = 4 056).

For women, 55% of the link between parental interest and allostatic load was mediated, through the educational pathway (30% of the total indirect effect) and by other intermediate factors (25% of the total indirect effect). A direct effect of 45% persisted after adjustment for confounding factors and mediators.

For this calculation among men, we did not consider the direct effect parental interest on allostatic load because the estimation had a non-significant value close to 0.

4. Discussion

Parents’ interest in their child’s education as perceived by teachers measured when cohort members were school children, was associated with their physiological health in mid-life in both men and women. Cohort members whose parents were perceived as uninterested or not very interested in their child’s education, as reported by the children’s teachers, had a higher allostatic load compared to individuals whose parents were considered to be interested. The association between parents’ interest in their child’s education as perceived by teachers and cohort member’s physiological wear-and-tear operated through intermediate pathways over the life course. Among men, 67% of the association operated through the educational pathway, and 33% through the other variables including income, smoking, alcohol consumption and physical activity. Among women, only 30% of the association operated through the education pathway, 25% worked through the other variables in adulthood, including psychological distress. Much of the association (45%) was direct, and unexplained by the tested pathways. Our results are in line with other studies where parental interest in their offspring’s studies as perceived by teachers was found to predict adult allostatic load and may buffer against poor mental health [30, 41]. Our findings provide insight into understanding how educational attainment as a reflection of dynamic life course social processes relates to physiological health, but also underline that parental-interest in children’s education has not been given much attention in relation to health over the life course.

Our results may highlight an interplay between culture and biology [42, 43] whereby a tension between a child’s home and school cultural environments may lead to a physiological stress response partly mediated by the educational trajectory. When a child attends school, the family social sphere meets the educational social sphere and if families have been socialized outside of the normative educational structure, they may need to adapt to the school environment [44]. Children who experience dissonance, as a chronically stressful challenge, may solicit their biological resources, experience multi-system physiological dysregulation as measured by allostatic load, and this embodiment may represent the cost of adaptation for the children partly mediated by the educational trajectory. Our results are suggestive of this pathway, especially for boys/men.

We observed different associations in men and women. For men, the educational pathway had a significant and stronger effect on allostatic load, consistent with previously observed differences in mortality by educational level across age groups more pronounced in men than in women [3].

For women our results show that, after controlling for confounders and mediators, a sizable part of the initial effect remained unexplained. Based on our findings, we also hypothesize that dissonance affects the physiological health for girls/women directly, or through pathways that remain to be tested. The unexplained direct effect may represent other possible pathways, or different early life socialization and embodiment processes among girls. Intersecting domains of power including class, gender and others (race, disability, etc) are likely to be at play, and deserve further attention [45, 46].

Furthermore, it is possible that the behavior of teachers towards children whose parents they consider to be less involved, could be different. Teachers perceive families’ economic and cultural capital once children enter school, and may unconsciously show favoritism toward students from more socially advantaged classes [47]. It is possible that our findings reflect a bias or difference whereby some teachers behaved differently towards children on this basis, contributing to increasing the stress of some children and therefore impacting their allostatic load.

An important aim of our analysis was to grasp the role of other intermediate factors through which parents’ interest in their child’s education as perceived by teachers may affect physiological processes. Adult income captured a large portion of the association for women and explained the association between educational level and allostatic load for men. Our findings suggest that consonant relationships between family and school, captured partly by parental interest, could promote ascending social mobility and therefore act as a vehicle towards social advantage, that may "buffer" the effects of an initially disadvantaged socio-economic environment on allostatic load [37, 48]. Furthermore, health behavior pathways appeared to explain a part of the association between parental interest and allostatic load for women and effects of education level on allostatic load for men. A consonant educational socialization could promote the embodiment of a health-relevant capital, i.e. the resources for acting in favor of health. Such consonance refers to all the “health related values, behavioral norm, knowledge and operational skills” [49]. However psychological malaise was found to explain the association between parental interest for women and allostatic load. Further analysis should be conducted in other cohorts to explore this association and ascertain its potential contextual specificity.

The main weakness of this study is that our variable measuring parental interest is one-sided, reflecting only the teacher’s point of view. It would have been interesting to compare this measure with parents’ perceptions; however, such data were unavailable. Attrition, and selection bias, common features related to longitudinal studies also pose issue. We carried out multiple imputation of missing data, a recommended method to avoid the interpretation of biased results, allowing them to be redressed to some extent. Information and recall biases may also be present, related to the self-reported nature of the data. With regard to alcohol consumption, we must consider that people with pathologies, but also those prone to alcohol addiction, are probably part of the abstention group, thus biasing the results of this group. Several years passed between the data collection sweeps wherein numerous life events most likely took place, which we cannot account for. Some variables in our study were measured at one given point in time, because we had only one measure available (i.e. allostatic load, sense of personal control, malaise), or we selected single variables as proxies for trajectories over time (i.e. behavior, social position). It is a regret that there is no earlier measurement of allostatic load in order to analyze its dynamic changes over time. However, studies have shown that the kinetics of allostatic load as measured in adulthood remains generally constant over time [50]. Our choice of statistical models and included variables are based on a priori theoretical and conceptual considerations. We may have overlooked variables or assumptions contributing to the relationship between parental interest and allostatic load. Lastly, NCDS 58 is a UK cohort, with unique cultural and historical aspects. It is therefore necessary to take precautions when extrapolating our results.

Despite these limitations, this study has a number of strengths. It is a longitudinal population-based study containing prospectively collected data with great detail and breadth across the life span, allowing us to control for numerous confounding and mediating variables. Another important strength is in the sample size included in the biomedical survey, and the large number of biomarkers available.

Education is often used as a measure of social position, where higher educational attainment is associated with better health outcomes. Our findings suggest the importance of considering education as a product of early life interactions between family and school social spheres. The family cultural environment can be examined through the three dimensions of capital i.e. economic, social and cultural capital [51, 52]. Economic capital refers to the material resources and financial support, social capital concerns interpersonal support whereas cultural capital exists in three forms: embodied (e.g. values, skills), objectivized (e.g. cultural goods, books) and institutionalized (e.g. educational level). In the context of coexisting social spheres between family and the school environments, socio-cultural dissonance may occur. Indeed, ‘‘the standards of the school are not neutral; their requests for parental involvement may be laden with the cultural experiences of intellectual and economic elites” [53]. Among the socially disadvantaged, who may not necessarily possess a common language with which to negotiate with representatives of the school institution, educational success at school may indicate a conversion of their cultural capital, a phenomenon described as “acculturation”. Conversely, for more socially advantaged students, this progression would be the result of the mobilization of their cultural capital heritage [54]. Consequently, the dissonance between the family social sphere and the school environment may lead to an “educational acculturation”, requiring the family to assimilate to the new educational culture.

5. Conclusion

Parents’ interest in their child’s education as perceived by teachers measured during childhood was associated with physiological wear-and-tear in mid-life in both men and women. This may be due to a physiological stress response, induced from early life due to a possible dissonance between family and school cultural environments, which have lasting effects on health, through pathways including educational attainment, particularly in men. These results suggest that awareness of children’s socio-cultural environments and gender should be considered when developing school or educational policies. As such, understanding family educational culture, cultural capital and socioeconomic position may contribute to developing adapted public policies supporting early childhood environments to reduce social inequalities in health.

Sensitivity analyses parental interest in 4 categories on complete-case data for men and women.

(DOCX)

Click here for additional data file.

Sensitivity analyses imputing parental interest measurement vs complete-case parental interest measurement for men and women.

(DOCX)

Click here for additional data file.

Descriptive characteristic on the subsample for men and women of the non-imputed data (n = 8 113).

Abbreviations and symbols: n = number of people; med = median; p25 = 25e percentile; p75 = 75e percentile. Values corresponding to the categories of allostatic load: Low: [0-2]; Medium: [3-4]; High: [5-12]. *P-values were calculated using a chi-squared test for categorical variables and Student t-test for the continuous variable.

(DOCX)

Click here for additional data file.

Detail on variable constructions.

(DOCX)

Click here for additional data file.

Direct and indirect effect.

(DOCX)

Click here for additional data file.

9 Oct 2020

PONE-D-20-18411

Could teacher-perceived parental interest be an important factor in understanding how education relates to later physiological health? A life course approach

PLOS ONE

Dear Dr. Camille Joannès,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

All comments from reviewer #1 must be addressed.

Please submit your revised manuscript by November 19. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Mary Hamer Hodges, MBBS MRCP DSc

Academic Editor

PLOS ONE

Additional Editor Comments:

Please carefully revise this manuscript to enable the reader to better understand your hypothesis and how you can draw meaningful conclusions when the follow-up rate is so low.

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I think this is an interesting manuscript but I feel that there is a need for the author to carefully analyze and present the paper more clearly to make it suitable for readers.

Some other comments are:

The paper is a little hard to read and the ideas don’t seem to flow sequentially and the quality of the language needs to be improved.

Some technical terms need to be adequately explained from the start: AL, PI ‘wear and tear’

The author failed to write on the relevance of the study, what is the gap there s/he wants to fill, not a sentence on assumptions about knowledge/epistemology of the paper, there was no literature in the introduction section triangulating with other studies and what is the related implication to programs. The author seems to be focusing more on the sociological aspect instead of articulating and linking the social activities of PI and AL to the child’s healthy or future outcomes.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: PONE-D-20-18411_reviewer (1).pdf

Click here for additional data file.

9 Dec 2020

Reviewer 1

1. I think this is an interesting manuscript but I feel that there is a need for the author to carefully analyse and present the paper more clearly to make it suitable for readers.

Author response: We thank the reviewer for this positive feedback on our work. We hope that the revised version of the manuscript will address the reviewer’s concerns on its clarity. More specifically, in the introduction section we have articulated the different conceptual stages from education to physiological wear and tear to health, considering the origins of educational attainment through the lens of interactions between families and school. We have also clarified the variable names and categories used.

Notably, p. 3 (lines 54-56), we have added a section to develop the inconsistency of the findings about the mechanisms through which education relates to health: “Beyond these groups of mechanisms others have been examined, such as environmental exposures and material conditions (5,6), calling for further studies to examine the mechanisms through which education relates to health“. In addition, p.4 (lines 82-87) we have explained the association of early life exposures and physiological wear-and-tear, mediated by education: “the social environment in early life may therefore have lasting effects on different social, biological, and behavioral factors that might act as mechanisms connecting education to repeated stress, which in turn affects health (20). Because educational attainment is the end point of a complex process where a wide array of factors (institutional, interpersonal, individual) shape trajectories of schooling, there is a need to look upstream in order to capture education as a long-term process grounded in a broader social context (20)”. We finally focused p.5 (lines 92-102) on parental interest in their child's education as an early life factor to examine “how the relationship between the home and school environments in early life affects physiological wear-and-tear, through educational attainment” (lines 100-102).

2. The paper is a little hard to read and the ideas don’t seem to flow sequentially and the quality of the language needs to be improved. Some technical terms need to be adequately explained from the start: AL, PI ‘wear and tear’

Author response: In the revised version of the manuscript, we have completely reviewed the language. We have also explained AL, PI and ‘wear and tear’ from the start. Regarding AL and wear and tear (p.3 lines 63-70) we have explained that these concepts refer to the embodiment process: “interactions (which) are perceived through the senses, are interpreted by the central nervous system leading to peripheral physiological responses. These physiological responses are adaptive processes, which maintain physiological stability in response to environmental challenges (10). The repeated activation of compensatory physiological mechanisms as a result of chronic exposure to stress can lead to physiological wear-and-tear, termed allostatic load (AL). AL measures the consequence of a prolonged activation of the stress response system by external challenges, leading to physiological imbalances across systems”.

With regard to PI (parents' interest in their child's education as perceived by teachers), p.8 (lines 174-175) it “was measured at age 7, 11 and 16 using information provided by the child’s teachers. The teacher was asked to report the level of interest of each parent in their child’s education”. Also, p. 5 (lines 105-108) “Teacher’s assessments [...] may partly reflect the position and viewpoint of the educational institution in terms of children’s compliance with academic requirements and potentially capture the tension between the home and school environments experienced by some children”.

3. The author failed to write on the relevance of the study, what is the gap there s/he wants to fill, not a sentence on assumptions about knowledge/epistemology of the paper, there was no literature in the introduction section triangulating with other studies and what is the related implication to programs. The author seems to be focusing more on the sociological aspect instead of articulating and linking the social activities of PI and AL to the child’s healthy or future outcomes.

Author response: Thank you for this constructive comment. In the revised version of the manuscript, we have amended the text to carefully highlight the gap our study aims to fill. More specifically, we have detailed by which mechanisms educational attainment is related to health (p.3 first paragraph) showing that part of the social gradient in health remains to be explained. We have added literature triangulating with other studies (p.5 lines 94-99): “parental interest in their child’s education has also been reported to have positive effects on psychosocial adjustment (22) later mental health (23) and metabolic outcomes in middle-age (24). Moreover, parental interest in their offspring’s schooling has been identified as a determinant of educational success (25–28). This parental involvement in children’s academic socialisation has been shown to influence academic success over and above child's social class and cognitive abilities (29).”

We also added a section regarding the knowledge gap we want to fill (p.5 lines 114-122) “To our knowledge, only one previous study examined the association between parental academic involvement as perceived by teachers and AL in midlife in a Swedish cohort. Parental interest in their children’s studies during the last year of school, rather than parental social class or availability of practical academic support, was found to predict adult AL largely mediated by academic achievement, which the authors attributed to the potential consequence of physiological stress over the life course (31). Our study aims to examine the association between parents' interest in their child's education as perceived by teachers, during a child’s educational trajectory, and multi-system physiological dysregulation measured using allostatic load, and whether education and other pathways mediate this relationship.”

Reviewer 2

ABSTRACT

1. Line 27: Which child education are you referring? Is it the teacher's child or the pupils?

Author response: We agree with the reviewer that the formulation ‘teacher-perceived parental interest in their child’s education’ was rather confusing. In the revised version of the manuscript, it has been replaced by ‘parents' interest in their child's education as perceived by teachers’.

2. Line 30: Who are these participants? You have to make it clear from the onset.

Author response: These participants are those who originally had AL scores available. We have modified the sentence to clarify this point p.2 (lines 30-31) “We used data from 9 377 women and men born during in 1958 in Great Britain and included in the National Child Development Study to conduct secondary data analyses” and in the section method p.7 (lines 135-137) “Our study is based on secondary analyses of data from the 1958 National Child Development Study (NCDS), an observational prospective population-based cohort study, which included all live births in Great Britain during one week in 1958 (n = 18 555)”.

3. Line 30: You are the data analysis, sample size and results in one sentence. i will suggest: Linear regression analyses for a 7850... Result revealed that people (be specific, who are these people?)...

Author response: We have amended the abstract on this sense p2 (lines 34-35). “Linear regression analyses were carried out on a sample of 8 113 participants with complete data for AL, missing data were imputed”.

4. Line 35: why do you use conditional word 'may', your findings should be able to give answers to this hypothesis.

Author response: Thank you for this suggestion. We have modified this section to provide clear answers to our hypothesis. However, in the case of our study, it seems slightly out of scope to do not use terms that nuance because this study is a piece of evidence, with data limitations, and it would require further research to confirm the conceptual theories mobilized.

5. Line 36: line 35 and 36 are unclear, please rephrase.

Author response: This comment is linked with the previous one. We have accordingly modified this section.

INTRODUCTION

6. Line 46: ...However, the mechanisms through which education relates to health, remain poorly understood...reference for this please.

Author response: Thank you, in line with this comment and reviewer 1’s, we have re-written the first paragraph of the introduction to give more details on the mechanisms by which education is associated with health and illustrating that this relationship is not entirely explained. In the literature, several mechanisms are highlighted by which education can affect health. However, when these variables are taken into account in different models, the association between education and health is affected but not fully explained, suggesting the existence of other pathways to be explored : “Gallo et al. showed that health behaviors and lifestyle factors explained only a part of the educational inequalities in total mortality (3). Similarly access to healthcare had a modest role in the educational gradient in health (4). Beyond these groups of mechanisms others have been examined, such as environmental exposures and material conditions (5,6), calling for further studies to examine the mechanisms through which education relates to health “(p.3 lines 51-56).

7. Line 60: How? please try to make your message reflective to the reader instead of being abstract. try to bring or collaboration in your thought... for eg: People with low socioeconomic conditions, i.e. as defined by low income, education and occupation, are at higher risk of infection compared to people with higher socioeconomic conditions.

Author response: We have added a sentence in the introduction (p.4 lines 77-79) in order to clarify this message “One recent study documented that disadvantaged early life socioeconomic conditions are associated with an increased risk of having a higher AL in midlife, mainly through educational pathways (17).”

8. Lines 61-65: sentence is too long. lots of repetition of 'wear and tear'...try to use other synonyms such as repeated stress or physiological consequence and more as appropriate.

Author response: We reformulated lines 80-84 to have a clearer and less repetitive sentence “As such, educational attainment “is an excellent marker of the ‘healthfulness’ of accumulated childhood experience’’(19) as the social environment in early life may therefore have lasting effects on different social, biological, and behavioral factors that might act as mechanisms connecting education to repeated stress, which in turn affects health”.

9. from line 48-80 are confusing. The early life socioeconomic conditions and associations have not been articulated well. this link might help: https://equityhealthj.biomedcentral.com/articles/10.1186/s12939-017-0553-7

Author response: Thank you for this useful reference which has helped us restructure our presentation of the paper in the introduction & methods to clarify what we mean by early life socioeconomic conditions and associations (p.4).

10. Line 88: the author fail to write on the relevant of the study, what are the gap there you want to fill, what is the related implication to policies and programs.

Author response: We have now incorporated a section p.4 to develop the relevance of the study and the gap in knowledge we want to fill. We aim to contribute to the scientific literature about relevant pathways from the early life environment towards health outcomes in adulthood. This scientific evidence may inform future research about intervention programs and policies however our study is not adapted to be directly interpreted for policy purposes.

11. Line 95: the introduction does not have literature on important factor in understanding how education relates health. Does not triangulate with other literatures across the region. The literature in unclear on the factors affecting PI, education and heath outcomes. See: https://www.unaids.org/sites/default/files/sub_landing/files/10_4-Intro-to-triangulation-MEF_0.pdf

Author response: We thank the reviewer for this helpful document. In the revised version of the manuscript, we extended the first paragraph to better illustrate how education is related to health. Additionally, most of the introduction has been re-written in order to better justify the use of PI as an upstream determinant of educational attainment that may affect the allostatic load and in-turn impact health.

METHOD

12. Line 110: The methods section does not have any structure. It is hard to identify the study design (desk review or quantitative or mix method?). What sampling strategy was employed in selecting participants for the study?

Author response: We have restructured and clarified the methods section. The sampling strategy and participant selection that was used to constitute the NCDS cohort study is provided in more details on p.7, section ‘Study population’.

13. Line 111: please state the criteria inclusion and exclusion

Author response: As detailed above, the study population section has been modified to give a detailed description of the participants included in the study. The flow chart (Figure 1) has been modified accordingly.

14. Line 128: which data are we using here 'previous' or current. Where is the previous one?

Author response: Thank you for pointing out this confusion. The National Child Development Study is a national cohort study which was set up in the UK in 1958 and has been made accessible and open for analysis by scientists since its inception. There are therefore many previous analyses and papers based on the data from this cohort. To avoid any confusion, we’ve replaced ‘as used previously within this cohort’ by ‘based on tertiles in the population’. We have also clarified that we are conducting secondary data analysis on the cohort in the first sentence of the methods section (p.8 lines 167-169)

15. Line 132: Did you interview the parent of these children? If not, I think it would have been good to make them your Key informant thereby increasing the validity and reliability of your findings. see: https://methods.sagepub.com/book/key-concepts-in-social-research/n28.xml

Author response: As mentioned above, we have clarified that we are carrying out secondary data analysis on a UK national cohort study set up in 1958, therefore we were not in charge of data collection. We have also added the following sentence: ‘Data collection on health, economic, social and developmental factors was carried out on cohort members from birth until now at age 7, 11, 16, 23, 33, 42, 44/45, 46, 50, 55 and 62 years and conducted by the Centre for Longitudinal Studies.’ in the Study population section.

We thank the reviewer to raise this point and agree that it would have been interesting to compare parents' interest in their child's education as perceived by teachers with a measure parental interest from the perspective of parents as key informants, this would have certainly increase the validity and reliability of our findings. Because no data on parental interest was collected from the parents themselves we could only focus on parental interest perceived by the teacher. This was mentioned as the main limitation of our study: “The main weakness of this study is that our variable measuring parental interest is one-sided, reflecting only the teacher’s point of view. It would have been interesting to compare this measure with parents' perceptions” (p.35 lines 450-452).

16. Line 134: Were these scaled answers or open-ended answers. If open-ended questions were used, what was the process used to determine theses categories?

Author response: The teachers were asked the following closed question about each parent “With regards to the child’s educational progress, do the father and mother appear:” The teacher could then provide one of the following closed responses: “Over concerned about the child’s progress and/or expecting too high a standard ; Very interested ; To show some interest ; To show little or no interest ; Can’t say ; Inapplicable (e.g. no father)”. We have modified the sentence on this on p.8 (lines 174-177).

17. Line 156: these can be summed in a table as 'Participant category'

Author response: We reformulated the section using “Prior confounders” and added information regarding the corresponding variables “We selected variables from questionnaires completed during childhood by parents of cohort members that were likely confounders between PI and AL, based on the literature” (p.9 lines 191-192)

18. Line 161: was any demographic data collected? how many children studied?

Author response: The NCDS study information has been clarified p7. We have a sample of 9 377 participants who were followed up from birth, throughout childhood and into adulthood. The cohort study contains a vast number of variables on each individual across their life course. More information can be found here https://cls.ucl.ac.uk/cls-studies/1958-national-child-development-study/

19. Line 179: You must include a section on assumptions about knowledge / epistemology in your paper. The epistemological approach should be stated and referenced. How and why the approach applies to the study should be present, it should be made clear why this approach was the most appropriate for the study. There should be a brief discussion of how the research will use the approach.

Author response: We have, accordingly, modified this section (p.10-11, lines 211-240) to emphasize this point. “In order to determine whether any observed associations between PI and AL were due to subsequent adult intermediate factors, we selected four intermediate groups of life course variables as pathways between PI and health, based on epidemiological evidence and on empirical studies. The following mediating factors were then added to the models: i) Lynch et al. (37) suggested that health is the result of an accumulation of experiences and exposures due to the material world. PI could promote a positive accumulation of academic and social success (38), reducing the impact of economic difficulties on educational attainment […] ii) Tension between home and school environments experienced by some children may provide an insecure social environment, contributing to health through psychological processes. Indeed, the resulting negative emotions through psycho-neuroendocrine mechanisms could lead to a poor health status […] iii) The adoption of health behaviors is closely related to education with people with a higher education level are more likely to have protective health behaviors, impacting physiological functioning (17,41) […] We defined the sets of mediators according to the temporal and causal assumptions of the life course approach, applying the general approach of Baron & Kenny (42). We assumed the set health behaviors (at 42y) came after and could be influenced by socioeconomic position (at 33y), which came after and could be influenced by psychosocial condition (at 23y), which came after and could be influenced by education (at 23y), which came after and could be influenced by PI.”

20. Line 223: Please try to make your ideas flow succinctly from top to bottom...

Author response: We have modified the format of the presentation of the different stages of the statistical analysis and moved a theoretical paragraph upstream of the statistical analysis section (p.12). We hope that this allows a clearer reading.

RESULT

21. Line 236: can you describe the men and women in this study.

Author response: We have added an additional table (S5_Table C) in the Supplementary information that describes the characteristics of the men and women included in our study sample. In the revised version of the manuscript, a description of this new Supplementary Table has been added (p. 14 lines 290-296)

22. Line 256: the results need to be discussed and interpreted with each of the confounders, making linkages or association and correlations with other variables. The author need to do comparison with PI vs AL and correlated with other factors.

Author response: We have made changes throughout the text of the results p.14-15 to detail our results according to the statistical associations between AL or PI and the other variables.

23. Line 268: You should be able to explain and interpret your findings in this

Author response: In this section, we focussed on describing the most salient results relating to our hypotheses. We felt that the explanatory section interpreting our findings would be more appropriate at the beginning of the discussion and the interpretive section throughout the discussion text.

24. Line 419: It fails to provide guidance/suggestions about how national policy-makers and other concerned actors, including ministry of Basic education, social welfare, health personnel, nongovernmental organizations and international development agencies, should proceed.

Author response: Thank you for this suggestion. While we think our findings are a relevant contribution to the literature to understand how the early life environment contributes to health dynamics over the life course, we do not aim to directly inform guidance to policy-makers. Rather, this work may inform other types of research set-up to examine interventions which could, in turn, inform policy-making.

In addition to the above comments, all spelling and grammatical errors pointed out by the reviewers have been corrected. We look forward to hearing from you in due time regarding our submission and to respond to any further questions and comments you may have.

Sincerely,

Submitted filename: 1. Response to Reviewers.docx

Click here for additional data file.

7 Apr 2021

PONE-D-20-18411R1

Could teacher-perceived parental interest be an important factor in understanding how education relates to later physiological health? A life course approach

PLOS ONE

Dear Dr. Joannès,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

I have been assigned to help with this manuscript as the original editor is no longer available. One reviewer who had already reviewed the original submission has also commented on the resubmission, for which we are grateful. In contrast to the reviewer's opinion, I think the paper is well suited at PLOS ONE after all comments have been sufficiently addressed.

The paper deals with an important research question of how education and physiological health may be influenced by family characteristics such as parental interest in their child's education. The limitations, e.g. allostatic load is measured at only one point in time, parental interest is only assessed as perceived as teachers, have been appropriately commented on.

After careful reading of the paper, I have noticed a few additional issues that were not raised in the first round of revision, which I mention below, and which need to be addressed before moving forward with this manuscript.

==============================

Please submit your revised manuscript by May 22 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Anja K Leist, Professor Dr.

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Additional Editor Comments (if provided):

- Personally, I usually recommend to abstain from all unnecessary abbreviations as the reader needs to remember these throughout the manuscript. So I would suggest to increase readability to say, "parents' interest in their child's education as perceived by teachers, mentioned as parental interest in the following", and speak of parental interest instead of PI; and of allostatic load instead of AL.

- I was pleased to see VanderWeele and Vansteelandt's method applied to test the putative mediators. If not absolutely necessary, I would remove references to Baron and Kenny and instead define mediators etc. in line with VanderWeele and Vansteelandt.

- I suggest to add to the abstract that the direct link between parental interest and allostatic load was completely mediated in men, but only partially mediated in women.

- The main set of results for men and women need to be carefully checked. For example, I would use p<0.001 or a manually adjusted p level to indicate significance, as otherwise family wise error rate (alpha inflation) would be a problem. Equally importantly, on p. 36, line 339 ff., it seems by adjusting for confounders, the main association between parental interest and allostatic load is rendered insignificant in men (the confidence interval includes the null, the p's are above 0.05). If I am reading this correctly, the results need to be reformulated to reflect the fact that the main association was not significant anymore in the adjusted models.

Minor issues:

- Please also use . instead of , for the p values, confidence intervals (decimal points)

- p. 6 line 130 - there is an incomplete sentence (a few more missing periods throughout the manuscript, please check if ends of sentences are complete).

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This has improved compared to the first submission; however, I still think the authors need to improve on the sequential flow of the concept. I will also recommend that this paper is submitted to International Journal of Research and Scientific Innovation in Social sciences (RSIS) or a mental health journal as this is not a public health paper. It is kind of related to early childhood development or education and its answers some correlated questions to its impact on earning potential, midlife health and behavioural status but I still find it hard to conceptualize and put the jigsaw concisely.

Line 49-50 please remove the “harsh material condition”

Line 51-52 this sentence is unclear, …inequalities in total mortality of Who?

The study design/methodology is not well-presented to explain the results. Interpretation of the results and the discussion are too wordy and are difficult to understand. It needs more careful proofreading.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

5 May 2021

Editor

1. Personally, I usually recommend to abstain from all unnecessary abbreviations as the reader needs to remember these throughout the manuscript. So I would suggest to increase readability to say, "parents' interest in their child's education as perceived by teachers, mentioned as parental interest in the following", and speak of parental interest instead of PI; and of allostatic load instead of AL.

Author response: Thank you for this suggestion, we have amended the abbreviations throughout the text of the manuscript, to increase the readability, by using “parental interest” instead of PI, “allostatic load” instead of AL, and we have also added “Sense of personal control” instead if SOC.

2. I was pleased to see VanderWeele and Vansteelandt's method applied to test the putative mediators. If not absolutely necessary, I would remove references to Baron and Kenny and instead define mediators etc. in line with VanderWeele and Vansteelandt.

Author response: We have amended the manuscript to remove the reference to Baron and Kenny, and altered the paragraph on mediators in line with VanderWeele and Vansteelandt’s recommendation on page 11 line 240: “We defined the sets of mediators according to the temporal and causal assumptions of the life course framework. As such, we made a pragmatic methodological assumption about the temporal ordering of variables in order to examine the mediating pathways whereby the exposure variable (parental interest at age 7, 11 and 16) preceded the respondent’s education, which preceded the other intermediate variables (psychological distress, socioeconomic position and health behaviors)”.

3. I suggest to add to the abstract that the direct link between parental interest and allostatic load was completely mediated in men, but only partially mediated in women.

Author response: Thank you for this we have amended the abstract page 2 line 40.

4. The main set of results for men and women need to be carefully checked. For example, I would use p<0.001 or a manually adjusted p level to indicate significance, as otherwise family wise error rate (alpha inflation) would be a problem. Equally importantly, on p. 36, line 339 ff., it seems by adjusting for confounders, the main association between parental interest and allostatic load is rendered insignificant in men (the confidence interval includes the null, the p's are above 0.05). If I am reading this correctly, the results need to be reformulated to reflect the fact that the main association was not significant anymore in the adjusted models.

Author response: Thank you for pointing this out, we agree that the section highlighted was ambiguous. We have modified the formulation throughout the results section for the multivariate results for men to better reflect the non-significant association. Page 24 lines 345-347, after adjustment for prior confounder, the association between parental interest and allostatic load remained significant (Model 2, β= 0.18 [0.03; 0.32]; p-value= 0.015). However, from the model 3, the association was no longer significant (Model 3, β= 0.06 [-0.09; 0.21]; p-value=0.44). Therefore, we have added line 349 “which was no longer significant” and line 353 “After sequentially controlling for all time-ordered life course, SEP and health behaviors, the association between parental interest and allostatic load was attenuated and not statistically significant (Model 6, β= -0.01 [-0.16; 0.14]).”

5. Please also use . instead of , for the p values, confidence intervals (decimal points)

Author response: Thank you for alerting us to this error. We have replaced commas with decimal points in the main text and supplementary files.

6. p. 6 line 130 - there is an incomplete sentence (a few more missing periods throughout the manuscript, please check if ends of sentences are complete).

Author response: Thank you for pointing this omission out. We have carefully proofread the text for other omissions and for fluidity of language. Page 6 line 131: “PI is associated with allostatic load, and (ii) explore four pathways through which PI may be differentially embodied during childhood, adolescence and early adulthood, leading to physiological wear-and-tear, as measured by allostatic load” and page 13 line 285 “See S3 File for more information on the different steps of the analysis of estimated direct and indirect effects.”

Reviewer

1. This has improved compared to the first submission; however, I still think the authors need to improve on the sequential flow of the concept. I will also recommend that this paper is submitted to International Journal of Research and Scientific Innovation in Social sciences (RSIS) or a mental health journal as this is not a public health paper. It is kind of related to early childhood development or education and its answers some correlated questions to its impact on earning potential, midlife health and behavioural status but I still find it hard to conceptualize and put the jigsaw concisely.

Line 49-50 please remove the “harsh material condition”

Line 51-52 this sentence is unclear, …inequalities in total mortality of Who?

The study design/methodology is not well-presented to explain the results. Interpretation of the results and the discussion are too wordy and are difficult to understand. It needs more careful proofreading.

Author response: Thank you for your helpful comments. We think that understanding how early life conditions in both the family and school environment may relate to later health are considerations of interest to a broad scientific and health audience, therefore we hope to pursue publication in Plos One.

Line 51-52 We replaced “harsh material condition” by disadvantaged material condition

Line 54-55 We added the additional point that it was total mortality of members of European cohorts.

We also have clarified some part of the methods section. Page 9 line 194 “To examine the relationship between parental interest and allostatic load, prior confounding variables potentially associated with both parental interest and allostatic load were added to the multivariable models. We selected variables most likely to be social or biological confounding factors from questionnaires completed during childhood by parents of cohort members, based on the literature.” Then the types of confounding factors were added such as “Socioeconomic confounders” line 199 or “Confounding factors related to family education” line 204.

Parts of the discussion section have been reorganized to make it easier to understand.

We look forward to hearing from you in due course regarding our submission and to respond to any further questions and comments you may have.

Sincerely,

Submitted filename: 1. Response to Reviewers.docx

Click here for additional data file.

18 May 2021

Could teacher-perceived parental interest be an important factor in understanding how education relates to later physiological health? A life course approach

PONE-D-20-18411R2

Dear Dr. Joannès,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Anja K Leist, Professor Dr.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

The authors have appropriately addressed all conceptual comments. However, a few shortcomings need to be taken care of. Please check particularly point (3), the results need to be presented correctly before the manuscript can be accepted for publication.

(1)There are still a few language errors (e.g. in the discussion, "We observed two different scenarii for each gender respectively" instead of "We observed different associations in men and women").

(2) Allostatic load is abbreviated as AL in at least two instances in the manuscript.

(3) The results section still presents insignificant associations that are "reduced" or "attenuated" which doesn't make sense:

"Controlling for educational attainment reduced the

349 strength of the association between parental interest and allostatic load which was no longer

350 significant (Model 3, β= 0.06 [-0.09; 0.21]). The association was only marginally affected when

351 psychological status at age 23 were accounted for (Model 4, β= 0.05 [-0.10; 0.21]).

352 Occupational social class and wealth reduced the strength of the association further (Model 5,

353 β= 0.02 [-0.13; 0.17]) with wealth making a significant contribution. After sequentially

354 controlling for all time-ordered life course, SEP and health behaviors, the association between

355 parental interest and allostatic load was attenuated and not statistically significant (Model 6, β=

356 -0.01 [-0.16; 0.14])."

I suggest to rewrite "The associations between parental interest and allostatic load were rendered insignificant after including educational attainment in Model 3 (Model 3, β= 0.06 [-0.09; 0.21]). Further adjusting for psychological status at age 23 (Model 4) and occupational social class and wealth (Model5) did not change the result patterns". I would not comment on the significant contribution of wealth in line with the Table 2 fallacy.

Reviewers' comments:

9 Jun 2021

PONE-D-20-18411R2

Could teacher-perceived parental interest be an important factor in understanding how education relates to later physiological health? A life course approach

Dear Dr. Joannès:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Prof. Dr. Anja K Leist

Academic Editor

PLOS ONE