Notice: This opinion is subject to formal revision before publication in the Atlantic and Maryland Reporters. Users are requested to notify the Clerk of the Court of any formal errors so that corrections may be made before the bound volumes go to press.

DISTRICT OF COLUMBIA COURT OF APPEALS No. 21-FM-134 JL.B., APPELLANT, V. L.B. and M.S., APPELLEES.

Appeal from the Superior Court of the District of Columbia (2011-DRB-627) (Hon. Steven N. Berk, Trial Judge) (Argued September 15, 2022 Decided December 15, 2022) Massiel Leiva, with whom Donald P. Salzman, Annamaria Kimball, and Alexis Alvarez were on the brief, for appellant Jl.B.

No brief was filed for appellees L.B. and M.S.

Melissa Colangelo, with whom Krystal Montgomery was on the brief, as guardian ad litem representing the interests of Ja.B.

Sasha Drobnick, Elizabeth Vogel, Joshua L. Richardson, and Caitlin M.

Kasmar filed a brief for the Domestic Violence Legal Empowerment and Appeals Project as amicus curiae in support of appellant Jl.B.

Before MCLEESE and DEAHL, Associate Judges, and RUIZ, Senior Judge.

DEAHL, Associate Judge: This custody dispute concerns Ja.B., who is now fourteen years old. The conflict is largely between Ja.B.’s mother, Jl.B., and his grandmother, L.B., though Ja.B.’s father, M.S., is also involved. Jl.B. was twenty years old when she had Ja.B. in 2008, and she enlisted her mother L.B. to help raise him. M.S. was largely out of the equation at that time. While the custody arrangement was somewhat variable in the years that followed, by 2016, Jl.B. and L.B. had reached an agreement under which they shared roughly equal physical custody of Ja.B.: (1) L.B. had Ja.B. on school days; (2) Jl.B. had Ja.B. on weekends and vacations, including the long summer vacation; and (3) M.S. had a weekly visit with Ja.B. That was their arrangement for several years, and Ja.B. seemed to thrive under it.

Jl.B. sought to modify that custody arrangement in 2019 and moved for sole legal and physical custody of her son. After several days of trial, the court denied that request and initially issued an order preserving the status quo, stressing Ja.B.’s success under the existing arrangement and his need for continuity. But shortly after the court issued that order, Jl.B.’s husband, J.H., physically attacked Jl.B. in Ja.B.’s presence. Ja.B.’s guardian ad litem asked the court to reopen the evidence and reconsider its order in light of that incident and its effects on Ja.B. The court granted that request and, after hearing an additional day of testimony, amended its order substantially, stripping Jl.B. of all custody save for two supervised visits with Ja.B. per month. The court also amended its order to grant joint legal custody to M.S. and L.B.

Jl.B. now appeals that decision. She argues that the trial court (1) erred in concluding that the statutory presumption favoring parental custody had been rebutted, see D.C. Code § 16-831.05(a); (2) violated the statutory prohibition against using the fact that a parent has been the victim of domestic violence in determining the parental presumption had been rebutted, see id. § 16-831.07(c); and (3) abused its discretion in applying the “best interests of the child” factors when fashioning the current custody arrangement, see id. § 16-831.08(a). We disagree with Jl.B. on the first point and conclude that any error as to the second point was harmless where the statutory presumption was rebutted prior to and independent of the domestic abuse incident and its aftermath. We agree with Jl.B. on her third point, however, and we therefore vacate the trial court’s custody order and remand for further proceedings.

I.

Ja.B. was born to Jl.B. and M.S. in 2008. For the first eight years or so of his life, Ja.B. and his mother both lived at L.B.’s house in Northwest D.C., though there were stretches where L.B. would kick Jl.B. out for being “disrespectful and belligerent.” While the trio lived together, L.B. and Jl.B. would both care for Ja.B., though the majority of parenting responsibilities seemed to fall on L.B. during the early years, as Jl.B. finished her schooling. In 2011, L.B. and Jl.B. agreed to share joint legal and physical custody of Ja.B., and L.B. sought a court order to that effect.

The court—by all accounts, erroneously—awarded L.B. sole legal and physical custody of Ja.B. Jl.B. has since filed a series of motions to modify that erroneous custody order, to no avail. Her earliest request to modify the custody order, made mere months after the erroneous order issued, asked the court to give her joint custody of Ja.B. That request was denied because Jl.B. apparently failed to properly serve her motion on the requisite parties.

Jl.B. again tried to modify the custody order in 2016, by which point she had married J.H. and moved out of L.B.’s house, though Ja.B. continued to live with L.B.

This time Jl.B. sought sole legal and physical custody of Ja.B. After mediation, Jl.B. and L.B. agreed that Ja.B. would live with L.B. during the school year and with Jl.B. on weekends and school breaks (a roughly even split of the calendar year). M.S. did not take part in the mediation, but Jl.B. and L.B. agreed that he would continue his established practice of visiting Ja.B. about once a week.

Ja.B. was happy with this 2016 arrangement and thrived under it. He attended school near L.B.’s house in Northwest D.C., where he got good grades and was on his school’s track and field and cross country teams. He had many friends, and some of his teachers described Ja.B. as their “best student.” L.B. tended to Ja.B.’s education: she helped him with his homework, communicated with his teachers, and attended parent-teacher conferences. Ja.B. liked spending time with both his parents and his grandmother. He also enjoyed spending time with his maternal and paternal half-sisters. All in all, Ja.B. testified that he “like[d] the way that things are” under the arrangement.

But not everybody was content with it. Jl.B. still wanted sole custody of her son, and in 2019, she again sought a modification of the controlling custody order and requested sole legal and physical custody of Ja.B. Animating her desire for sole custody is the fact that Jl.B. has fraught relationships with both L.B. and M.S. and doubts their abilities as caregivers; M.S. was notably convicted of assaulting Jl.B. while she was pregnant with Ja.B. L.B. and M.S., who are on better terms with one another, both opposed the requested modification. The case went to trial, and the court appointed a guardian ad litem to represent Ja.B.’s interests.

The guardian ad litem called Jl.B., L.B., and M.S. as witnesses, and questioned each of them about their suitability as parents. In response to those questions, Jl.B. acknowledged that she has been diagnosed with bipolar disorder, but denied that she was currently affected by the disorder and noted that she had not been on any medication since 2017. Rather, she described herself as having mood swings that resulted from certain stressors. L.B. testified that she herself had lupus and depression, but believed that she could effectively manage those conditions and that neither affected her day-to-day functioning. She also explained that she suffered a heart attack in 2018 and took medication and regularly saw her cardiologist. M.S. testified that he had no relevant mental or physical health issues. All three testified that they had positive relationships with Ja.B., which Ja.B. confirmed in his own testimony.

A court-appointed psychologist, Dr. Seth King, also testified about the parties’ relationships with Ja.B. and their abilities to co-parent. Dr. King described Ja.B.’s relationships with Jl.B., L.B., and M.S. in generally positive terms and expressed “[n]o major concerns” about any of their abilities to care for Ja.B. Dr. King noted that the conflict between Jl.B., on the one hand, and L.B. and M.S., on the other, limited “their ability to work collaboratively together.” He further opined that if Jl.B. were given sole legal and physical custody of Ja.B., it was his impression that Jl.B. would cut off all contact between L.B. and Ja.B., which he thought would be detrimental to Ja.B.’s well-being. Dr. King also expressed concerns about Jl.B.’s diagnosed but untreated bipolar disorder, which he noted could lead to periods where her “emotional functioning” and “emotional stability” were compromised.

After the hearing, the court issued a detailed order in May 2020 that essentially maintained the 2016 custody arrangement. It ordered that while L.B. would retain sole legal custody of Ja.B., L.B. and Jl.B. would share joint physical custody—with Ja.B. spending a roughly equal number of days with each of them—and M.S. would retain his weekly visitation. In reaching that result, the court acknowledged the statutory presumption favoring custody with parents, D.C. Code § 16-831.05(a), a presumption that operated against granting L.B. any amount of custody. But it found the presumption had been overcome in two statutorily enumerated ways: (1) that excluding L.B. from the custody arrangement “would be detrimental to the physical or emotional well-being of” Ja.B., D.C. Code § 16-831.07(a)(2); and (2) “[t]hat exceptional circumstances . . . support[ed] rebuttal of the [parental] presumption,” id. § 16-831.07(a)(3). Having found the parental presumption rebutted, the court concluded that it was in Ja.B.’s best interests to maintain the status quo, keeping him in his current home with L.B. during the school year, but still allowing him to spend substantial time with both his parents. The court’s reasoning was largely based on Ja.B.’s considerable success and happiness under the existing arrangement.

About two weeks after the court issued that order maintaining the status quo, Jl.B. was physically beaten in her own home by her husband, J.H., while Ja.B. was staying with her. Following this incident, the guardian ad litem filed a motion to reopen trial and amend the judgment in light of new evidence. See Super. Ct. Dom.

Rel. R. 59(a)(2). The court held another evidentiary hearing, in which L.B., Jl.B., and M.S. testified about the battery and its aftermath. Ja.B. did not testify or otherwise indicate that his preference for maintaining the status quo had changed.

The credited testimony was as follows.

On the night of the battery, Ja.B. texted M.S. for help stating that “stepdad was beating on” Jl.B. M.S. immediately called Ja.B. and could hear J.H. and Jl.B. in the background, both yelling, screaming, and cursing. J.H. eventually snatched the phone from Ja.B.’s hand and ended the call. M.S. then called the police, and when they arrived, Ja.B. was physically unharmed, but appeared afraid and had “blood all over his jeans,” apparently from trying to break up the fight. L.B. and M.S. both arrived on the scene after the police, and Jl.B. either told Ja.B., or told L.B. within Ja.B.’s earshot, that Ja.B. was not welcome back into her house because he had called M.S. for help. She repeated that remark in a text to M.S. the following morning. This incident appeared to have shaken Ja.B. For a few weeks afterward, he seemed unusually “quiet and withdrawn,” and he told L.B. he wanted to enroll in therapy.

J.H. was arrested and criminally charged. The court initially ordered J.H. to stay away from Jl.B., though the following month that order was modified to simply direct him not to abuse or harass her. Jl.B. testified that J.H. was not welcome back into her home unless he completed a domestic violence program. At the time Jl.B. testified, J.H. had not yet begun that program. Yet L.B. testified that she had seen J.H. at Jl.B.’s house in the months that immediately followed the beating. L.B. also testified that this was not the first or last time J.H. had beaten Jl.B.: L.B. claimed that there had been “several” “physical altercations” between the two earlier in 2020 and that additional “acts of violence” had occurred since the incident, though she did not elaborate on the basis for her belief.

The court issued a new custody order in January 2021, superseding the May 2020 order. The court largely incorporated all of its findings of fact and reasoning from the May 2020 order into this new order, including its persistent themes that Ja.B. was thriving and happy under the status quo custody arrangement. The only material change in the court’s order was that it detailed the incident of domestic violence and its aftermath, which led the court to then drastically alter the order’s bottom line. Rather than maintaining the status quo, the court granted joint legal and physical custody to L.B. and M.S., and granted Jl.B. just two visits per month—or twenty-four per year—all of which were to be supervised. Jl.B. now appeals that order.

II.

In the District of Columbia, there is a strong statutory presumption, with “constitutional underpinnings,” that parents should have custody of their children.

W.H. v. D.W., 78 A.3d 327, 338 (D.C. 2013) (citation omitted). Under the District’s third-party custody statute, D.C. Code §§ 16-831.01 to .13, a non-parent may be awarded custody only when the parents consent to it or this statutory presumption is rebutted by clear and convincing evidence showing that (1) “the parents have abandoned the child,” (2) “custody with a parent is or would be detrimental to the physical or emotional well-being of the child,” or (3) other “exceptional circumstances . . . support rebuttal of the presumption favoring parental custody.” Id. § 16-831.07(a). If the statutory presumption is rebutted through one of those showings, then a court may award some degree of custody to non-parents only to the extent it finds that it “is in the child’s best interests.” Id. § 16-831.06(a)(2); see also id. § 16-831.04(a) (explaining that the ordered custody may include joint custody between parents and non-parents or any “other custody arrangement the court determines is in the best interests of the child”).

Jl.B. challenges the trial court’s ruling at both steps of its analysis. That is, she first argues that the statutory presumption favoring parental custody was not in fact rebutted, both because the evidence did not suffice to rebut it and because the trial court committed legal error when it used the fact that she was abused against her, contrary to a statutory prohibition against doing just that. See D.C. Code § 16-831.07(c). She next argues that, even if the parental presumption had been rebutted, the trial court abused its discretion in stripping her of virtually all of her custody rights. In evaluating these arguments, we generally will reverse a trial court’s custody order only if we determine that it abused its discretion, Macklin v. Johnson, 268 A.3d 1273, 1279 (D.C. 2022), though Jl.B.’s argument that the trial court violated D.C. Code § 16-831.07(c) by counting the fact that she had been abused against her is a claim of legal error, which we review de novo.

We consider each of Jl.B.’s arguments in turn.

A.

A parent in a custody dispute with a third party is entitled to a presumption that “custody with the parent is in the child’s best interests.” D.C. Code § 16-831.05(a). This presumption protects a parent’s “constitutionally-protected right to care for, have custody of, and manage their children.” W.H., 78 A.3d at 338. A third party who seeks some degree of custody “bear[s] the burden of rebutting the parental presumption by clear and convincing evidence.” D.C. Code § 16-831.06(b).

At the outset, we note that the parties all interpret this statutory presumption as an all-or-nothing inquiry, protecting parents’ right to sole custody, but with no residual effect once the presumption of sole custody in the parents is rebutted. That is not the only way to read the statutory presumption, however. It might be interpreted as having residual force even if a third party can overcome the presumption of sole custody in the parents, so that it effectively maximizes the degree of parental custody so long as the parents are able and willing to care for the child to that degree, that degree of custody is not detrimental to the child’s physical or emotional well-being, and exceptional circumstances do not counsel against that degree of parental custody. See D.C. Code § 16-831.07(a). Under this alternative view, every incremental intrusion into parents’ custody rights would need to be justified via rebuttal of the parental presumption. No party advances that view, however, and so we proceed with the parties’ shared understanding that the parental presumption has no such residual effect. 1

L.B. argues that the parental presumption of sole custody was rebutted in two statutorily enumerated ways: (1) that “custody with a parent is or would be detrimental to the physical or emotional well-being of the child,” and (2) that “exceptional circumstances . . . support rebuttal of the presumption favoring parental custody.” D.C. Code §§ 16-831.07(a)(2)-(3). The trial court made both of those determinations in its initial May 2020 order. The trial court reasoned that the parental presumption had been rebutted based on the fact that L.B. “has been [Ja.B.’s] primary caretaker for nearly a decade” and “[h]er house has been his anchor, providing him with safety and stability.” The court found that living with L.B. has provided Ja.B. with a stable home environment in which he has been able

Also, the parties agree that the parental presumption applies in this case despite the fact that L.B. has technically had sole legal custody of Ja.B. since 2011 and the parental presumption usually applies only in initial custody determinations.

See S.M. v. R.M., 92 A.3d 1128, 1129-30, 1137 (D.C. 2014). We likewise will not second-guess their agreement on that point, particularly in light of the unusual background in this case and the apparent error underlying that 2011 order. to flourish, and that L.B. was the parental figure most “involved in the school and known well by school personnel” and who tended to Ja.B.’s medical needs.

Meanwhile, the court found that Jl.B.’s untreated bipolar disorder had “interfered with her ability to work with others for [Ja.B.’s] wellbeing.” It noted that her “interactions with [Ja.B.’s] school have been combative and accusatory” and her interactions with M.S. and L.B. “are characterized by cursing, threatening, name- calling, and blaming.” Indeed, Jl.B. described M.S. and L.B. as her “triggers.”

Crucially, the trial court concluded that this “evidence raises concerns about whether and how [Jl.B.] would keep [M.S.] and [L.B.] involved in [Ja.B.’s] life if she were his full-time caregiver.” Those concerns were supported by Dr. King’s credited testimony that Jl.B. seemed inclined to cut L.B. out of Ja.B.’s life. The court concluded that this evidence constituted exceptional circumstances that supported rebuttal of the parental presumption, and that it also showed that sole custody with Jl.B. would be detrimental to Ja.B.’s emotional well-being. See D.C. Code § 16-831.07(a)(2), (3).

We discern no abuse of discretion in those conclusions. The evidence supports the trial court’s findings that L.B. had parented Ja.B. throughout his life, that she had done so with great success, and that Jl.B. was not committed to keeping L.B. in Ja.B.’s life. It can fairly be described as exceptional circumstances, as the statute envisions, where a non-parent has essentially taken on the lead parenting role for more than a decade. And it is no abuse of discretion to conclude that cutting that primary and long-term caregiver out of the child’s life would be detrimental to the child’s physical and emotional well-being. The trial court acted within its discretion in finding that the parental presumption had been rebutted in its May 2020 order.

Jl.B. also argues that the trial court erred when, in its January 2021 order, it further considered the fact that she was abused as a mark against her in finding the parental presumption had been rebutted. We conclude that any error on that score was harmless, as the trial court had found the parental presumption rebutted independent of that abuse and its aftermath in its May 2020 order—reasoning that it repeated in its January 2021 order—and we have already held that the trial court acted within its discretion in doing so. Nonetheless, some aspects of the trial court’s revised order seemed to run afoul of the statutory prohibition that Jl.B. highlights, and because we have never previously discussed that statutory provision, we pause to explain why that is so.

A trial court in the District “shall not use the fact that a parent has been the victim of an intrafamily offense against [them] in determining whether the presumption favoring parental custody has been rebutted.” D.C. Code § 16-831.07(c). Jl.B. and amicus curiae, the Domestic Violence Legal Empowerment and Appeals Project (DV LEAP), make compelling arguments that the court violated this prohibition when it drastically reduced her custody in seeming response to her being abused.

For instance, while the court stated that it did not consider the “intrafamily offense as the sole basis for concluding that the parental presumption ha[d] been rebutted” (emphasis added), that statement clearly implies that the court did consider the fact of Jl.B.’s abuse as one factor among others in reducing her custody. The guardian ad litem defends the trial court’s ability to do just that and asks us to “interpret § 16-831.07(c)’s prohibition . . . to mean that a court cannot rely solely on evidence that the parent is a survivor of domestic violence” (emphasis added). That is not a sensible reading of the statutory language. The statute’s instruction that a court “shall not use” the fact that a parent is the victim of an intrafamily offense in its analysis, by its plain language, prohibits any use of that fact, whether as the sole factor in the analysis or as one factor among many. To be sure, the trial court might have given weight to the effects of the incident on Ja.B., or to the fact that Ja.B. was perhaps in harm’s way and exposed to physical abuse while in Jl.B.’s custody. But the trial court’s analysis gives us no assurance that it was carefully cabining its consideration in that manner, as opposed to treating Jl.B.’s status as a victim of domestic abuse against her in the custody calculus.

Jl.B. and DV LEAP also raise plausible arguments that the trial court drew a series of impermissible inferences from the sole fact that Jl.B. was abused without the necessary evidentiary support to ground those inferences. For instance, while it would certainly be proper for a court to consider past incidents of domestic violence as predictive evidence that it might recur, the trial court’s conclusory assertion that there was a “risk of future incidents that imperil [Ja.B.’s] physical or mental wellbeing” had little support in the record. The trial court made no findings about whether J.H. continued to live with Jl.B., whether Ja.B. had even been in the same vicinity as J.H. since the incident, or any number of other questions that would seem necessary to establish by clear and convincing evidence that Ja.B.’s physical and emotional well-being would be endangered by the potential for future abuse. See D.C. Code § 16-831.07(a). In short, the trial court’s findings and reasoning were too cursory to support its conclusion on this point.

Ultimately, because we have upheld the trial court’s reasoning that the parental presumption had been rebutted independent of the abuse and its aftermath, any error in violating the statutory prohibition in § 16-831.07(c) was harmless.

B.

We now address Jl.B.’s argument that the trial court abused its discretion when, after concluding that the parental presumption had been rebutted, it reasoned that limiting her custody to twenty-four supervised visits per year was in Ja.B.’s best interests. In assessing that argument, “we first look to whether the trial court considered ‘all relevant factors and no improper factor,’ and then we ‘evaluate whether the decision is supported by substantial reasoning . . . drawn from a firm factual foundation in the record.’” Macklin, 268 A.3d at 1279 (quoting In re A.M., 589 A.2d 1252, 1257-58 (D.C. 1991)). The relevant factors that the trial court must consider when fashioning a custody arrangement are:

(1) The child’s need for continuity of care and caretakers, and for timely integration into a stable and permanent home, taking into account the differences in the development and the concept of time of children of different ages; (2) The physical, mental, and emotional health of all individuals involved to the degree that each affects the welfare of the child, the decisive consideration being the physical, mental, and emotional needs of the child; (3) The quality of the interaction and interrelationship of the child with his or her parent, siblings, relatives, and caretakers, including the third- party complainant or movant; and (4) To the extent feasible, the child’s opinion of his or her own best interests in the matter. D.C. Code § 16-831.08(a). There is no suggestion that this list of factors is exhaustive. Rather, the “relevant factors” in any particular case will depend on the facts of the specific custody dispute before the court. Cf. Johnson v. Washington, 756 A.2d 411, 418 (D.C. 2000) (“Child custody cases present complex factual situations, and we necessarily rely on the trial court’s careful balancing of the various factors that may impact the child.”). Because the trial court did not consider all relevant factors, and because its analysis of the factors that it did consider failed to support its conclusions, the court abused its discretion.

First, we note that the trial court seemed to give no weight to Ja.B.’s preference in its January 2021 order. Contra D.C. Code § 16-831.08(a)(4). The court, in both its May 2020 and January 2021 orders, found Ja.B. to be a “bright, articulate, and mature” child who was “well-spoken and seem[ed] wise beyond his age.” After hearing him testify, the court opined that he had his “head on straight” and commented about how he was “[s]o thoughtful.” In short, Ja.B. was clearly capable of expressing an informed view about what was in his best interests. And the only evidence in the record is that, when the court asked Ja.B. “what [he’d] like [the court] to do,” Ja.B. responded that he “like[d] the way that things are.” To be sure, that evidence came prior to the domestic abuse incident, and the trial court initially granted Ja.B.’s request to keep the arrangement as it had been for years. But when the trial court then drastically altered the custody arrangement, it had no evidence before it that Ja.B.’s preferences had changed, and it gave absolutely no explanation for discarding Ja.B.’s preference or failing to make any inquiry into whether it had changed. Of course, a court is free to disagree with a child as to what is in the child’s best interests. See P.F. v. N.C., 953 A.2d 1107, 1117 (D.C. 2008).

But when the child is older and seemingly mature enough to express informed preferences, a court cannot be said to have meaningfully taken their preferences into account when it so fundamentally deviates from them without so much as an explanation.

In addition to failing to fully consider Ja.B.’s views, the court’s discussion of the other factors does not support its conclusion. Though the court provided some support for reducing Jl.B.’s custody—in the form of concerns about Jl.B.’s mental health and the potential for recurring domestic violence in her house, 2 see D.C. Code § 16-831.08(a)(2)—its reasoning, taken as a whole, is internally contradictory. This is perhaps unsurprising, since the majority of the court’s reasoning in its January 2021 order is identical to its May 2020 order despite the two orders’ drastically different bottom lines. For instance, in both orders, the court repeatedly emphasized

As we noted above, the court’s analysis of the risk of future domestic violence occurring and harming Ja.B. is sparse. See supra Part II.A. that Ja.B.’s continuity of care and his loving relationship with Jl.B. supported “regular, frequent” visitation with her. See D.C. Code § 16-831.08(a)(1)-(3). This reasoning supported the court’s conclusion in May 2020 that “it [wa]s in [Ja.B.’s] best [interests] to maintain the current custodial arrangement,” which included “substantial parenting time” with Jl.B. But it is at odds with the court’s subsequent decision to reduce Jl.B.’s visitation to only two supervised visits per month. That is not “frequent” visitation and certainly does not provide “continuity” in Ja.B.’s relationship with Jl.B., particularly when all of the visits must be supervised so that Ja.B. can no longer spend any time alone with his mother.

Though the court expressed legitimate concerns about the impact of Jl.B.’s mental health on Ja.B., all of the evidence relating to Jl.B.’s mental health was in the record prior to May 2020, when the court concluded that Jl.B.’s “mental health . . . does not support [the conclusion that] [Ja.B.]’s time with his mother be limited further.” Jl.B.’s mental health thus cannot have formed the basis for the court’s revised order in January 2021.

Nor was any of the reasoning that the court added to the January 2021 order enough to cure the internal contradictions in the rest of the order. The only major difference in the January 2021 order was the addition of a single paragraph in which the court stated that “unsupervised visitation with [Jl.B.] poses a risk to [Ja.B.]’s physical and emotional health” because of “concern[s] about ongoing domestic violence in [Jl.B.’s] home and the potential for physical harm to [Ja.B.]” and “concern[s] about continuing emotional harm to [Ja.B.] . . . in light of [Jl.B.’s] decision to lash out at [Ja.B.] for seeking help.” Though these are legitimate concerns, the court provided no analysis of the likelihood that Ja.B. would be harmed, emotionally or physically, by domestic violence in the future. Supra Part II.A. Nor did it consider whether Jl.B’s “lashing out” at her son may have been mitigated by the fact that she herself had just been the victim of a violent attack. Nor did the court attempt to square its concerns for Ja.B’s emotional and physical wellbeing with the three other statutory factors which, by the court’s own judgment in May 2020, all supported maintaining the status quo. This minor addition to the court’s order cannot alone sustain its substantial modification of the custody agreement.

In addition to erroneously reducing Jl.B.’s physical custody, the court erroneously increased M.S.’s custody when it granted him joint legal and physical custody with L.B. This is so for two reasons: the court failed to consider M.S.’s conviction for simple assault against Jl.B., and its expansion of M.S.’s custody lacked sufficient supporting reasoning. First, the court failed to consider M.S.’s conviction for simple assault against Jl.B. while she was pregnant with Ja.B. Jl.B. argues that the court was bound by the statutory requirements governing custody disputes between parents, including that a court shall not award any custody to a parent who has committed an intrafamily offense without a written statement justifying that determination. See D.C. Code § 16-914(a-1). The guardian ad litem responds that the provisions governing custody disputes between parents do not apply to third-party custody disputes such as this one.

We need not resolve that technical dispute because the court was, in any case, bound to consider M.S.’s intrafamily offense as a “relevant factor[]” under the third- party custody statute. D.C. Code § 16-831.08(a). M.S.’s conviction for simple assault against Jl.B. while she was pregnant with Ja.B. is relevant here because, as a general principle, “[t]he law of this jurisdiction requires a judicial officer to exercise considerable caution before granting custody or visitation rights to a parent who has committed an intrafamily offense.” P.F., 953 A.2d at 1112. Though the court acknowledged that M.S. had been convicted of assaulting Jl.B., it never mentioned this fact in its analysis, let alone justified why joint custody with M.S. would be in Ja.B.’s best interests despite this conviction. That is not to say that the court could not grant joint custody to M.S.—only that it must explain its reasons for doing so despite his apparent past abusive conduct.

Second, the trial court’s grant of joint legal custody to M.S. lacked sufficient supporting reasoning. The court awarded joint legal custody to M.S. for the first time in January 2021, despite the fact that no new evidence relating to M.S. had been introduced since the May 2020 order, in which M.S. was not granted legal custody.

If the court wished to give M.S. joint legal custody, it was required, at a minimum, to explain its reasons for doing so. See Macklin, 268 A.3d at 1279 (custody order must be “supported by substantial reasoning . . . drawn from a firm factual foundation in the record”); D.C. Code § 16-831.04(b) (“An order granting relief under this chapter shall be in writing and shall recite the findings upon which the order is based.”).

Thus, even with the high standard of deference we afford the trial court in child custody cases, we cannot say that the trial court’s decision to reduce Jl.B.’s visitation to twenty-four supervised visits a year and to award M.S. joint legal and physical custody was supported by substantial reasoning. We conclude that the court’s January 2021 custody order was an abuse of discretion.

III.

For these reasons, while we do not disturb the trial court’s conclusion that the parental presumption was rebutted, we vacate the court’s January 2021 revised modified custody order and remand for further proceedings.

So ordered.